library(rstanarm)
Loading required package: Rcpp
Registered S3 methods overwritten by 'ggplot2':
method from
[.quosures rlang
c.quosures rlang
print.quosures rlang
Registered S3 method overwritten by 'dplyr':
method from
print.rowwise_df
Registered S3 methods overwritten by 'htmltools':
method from
print.html tools:rstudio
print.shiny.tag tools:rstudio
print.shiny.tag.list tools:rstudio
Registered S3 method overwritten by 'htmlwidgets':
method from
print.htmlwidget tools:rstudio
Registered S3 method overwritten by 'xts':
method from
as.zoo.xts zoo
rstanarm (Version 2.18.2, packaged: 2018-11-08 22:19:38 UTC)
- Do not expect the default priors to remain the same in future rstanarm versions.
Thus, R scripts should specify priors explicitly, even if they are just the defaults.
- For execution on a local, multicore CPU with excess RAM we recommend calling
options(mc.cores = parallel::detectCores())
- Plotting theme set to bayesplot::theme_default().
library(rstan)
Loading required package: ggplot2
Loading required package: StanHeaders
rstan (Version 2.18.2, GitRev: 2e1f913d3ca3)
For execution on a local, multicore CPU with excess RAM we recommend calling
options(mc.cores = parallel::detectCores()).
To avoid recompilation of unchanged Stan programs, we recommend calling
rstan_options(auto_write = TRUE)
library(glmnet)
Loading required package: Matrix
Loading required package: foreach
Loaded glmnet 2.0-16
library(caret)
Loading required package: lattice
Registered S3 method overwritten by 'data.table':
method from
print.data.table
Attaching package: ‘caret’
The following objects are masked from ‘package:rstanarm’:
compare_models, R2
randseed <- 12345
set.seed(randseed)
# load data
WORK_DIR <- "/Users/linyingzhang/LargeFiles/Hripcsak/deconfounder/"
DATA_PATH <- paste0(WORK_DIR, "data/simulation_multicause_data/")
# load causes
x_df <- read.table(paste0(DATA_PATH, "simulated_causes.txt"))
names(x_df) <- seq(1, ncol(x_df), 1)
# load substitute confounders
T_hat_pmf <- read.table(paste0(DATA_PATH, "x_post_np_PMF_k450.txt"))
x_t_df_pmf <- as.data.frame(cbind(x_df, T_hat_pmf))
T_hat_def <- read.table(paste0(DATA_PATH, "x_post_np_DEF_2_2.txt"))
x_t_df_def <- as.data.frame(cbind(x_df, T_hat_def))
# load true confounders
C <- read.table(paste0(DATA_PATH, "simulated_multicause_conf.txt"))
x_c_df <- as.data.frame(cbind(x_df, C))
# load outcome
ys <- read.table(paste0(DATA_PATH, "simulated_outcomes_1592223649.txt"))
ys <- t(ys)
# load true coefficients
betas <- read.table(paste0(DATA_PATH, "simulated_true_coeffs_1592223649.txt"))
betas <- t(betas)
n_causes <- dim(x_df)[2]+1
n_sims = dim(ys)[2]
# Run outcome models
summary_stats <- array(NA, dim=c(n_sims, 4, 4))
for (sim in seq(1, n_sims, 1)){
y <- ys[,sim]
beta <- betas[,sim]
#fit ridge models
fitridge_no_control = stan_glm(y~., data = x_df, family = gaussian(), prior = normal(),
algorithm = "meanfield", adapt_delta = NULL, QR = FALSE,
sparse = FALSE, seed = randseed)
fitridge_oracle = stan_glm(y~., data = x_c_df, family = gaussian(), prior = normal(),
algorithm = "meanfield", adapt_delta = NULL, QR = FALSE,
sparse = FALSE, seed = randseed)
fitridge_pmf = stan_glm(y~., data = x_t_df_pmf, family = gaussian(), prior = normal(),
algorithm = "meanfield", adapt_delta = NULL, QR = FALSE,
sparse = FALSE, seed = randseed)
fitridge_def = stan_glm(y~., data = x_t_df_def, family = gaussian(), prior = normal(),
algorithm = "meanfield", adapt_delta = NULL, QR = FALSE,
sparse = FALSE, seed = randseed)
no_control_coefs <- fitridge_no_control$coefficients[2:n_causes]
oracle_coefs <- fitridge_oracle$coefficients[2:n_causes]
pmf_coefs <- fitridge_pmf$coefficients[2:n_causes]
def_coefs <- fitridge_def$coefficients[2:n_causes]
rmse_no_control <- sqrt(mean((beta - no_control_coefs)**2))
rmse_oracle <- sqrt(mean((beta - oracle_coefs)**2))
rmse_pmf <- sqrt(mean((beta - pmf_coefs)**2))
rmse_def <- sqrt(mean((beta - def_coefs)**2))
# CI
ci95_no_control <- posterior_interval(fitridge_no_control, prob = 0.95)
ci95_oracle <- posterior_interval(fitridge_oracle, prob = 0.95)
ci95_pmf <- posterior_interval(fitridge_pmf, prob = 0.95)
ci95_def <- posterior_interval(fitridge_def, prob = 0.95)
# coverage: if the 95ci covers the true coefficients
nc_coverage <- (beta >=ci95_no_control[2:n_causes,1]) & (beta <= ci95_no_control[2:n_causes,2])
oracle_coverage <- (beta >=ci95_oracle[2:n_causes,1]) & (beta <= ci95_oracle[2:n_causes,2])
pmf_coverage <- (beta >=ci95_pmf[2:n_causes,1]) & (beta <= ci95_pmf[2:n_causes,2])
def_coverage <- (beta >=ci95_def[2:n_causes,1]) & (beta <= ci95_def[2:n_causes,2])
truth <- as.factor(ifelse(beta != 0, 1, 0)) # factor of positive / negative cases
oracle_all_coverage <- sum(oracle_coverage)/50
nc_all_coverage <- sum(nc_coverage)/50
pmf_all_coverage <- sum(pmf_coverage)/50
def_all_coverage <- sum(def_coverage)/50
oracle_causal_coverage <- sum(oracle_coverage[truth==1])/10
nc_causal_coverage <- sum(nc_coverage[truth==1])/10
pmf_causal_coverage <- sum(pmf_coverage[truth==1])/10
def_causal_coverage <- sum(def_coverage[truth==1])/10
oracle_noncausal_coverage <- sum(oracle_coverage[truth==0])/40
nc_noncausal_coverage <- sum(nc_coverage[truth==0])/40
pmf_noncausal_coverage <- sum(pmf_coverage[truth==0])/40
def_noncausal_coverage <- sum(def_coverage[truth==0])/40
summary_stats[sim,,] <- rbind(cbind(rmse_oracle, oracle_all_coverage, oracle_causal_coverage, oracle_noncausal_coverage),
cbind(rmse_no_control, nc_all_coverage, nc_causal_coverage, nc_noncausal_coverage),
cbind(rmse_pmf, pmf_all_coverage, pmf_causal_coverage, pmf_noncausal_coverage),
cbind(rmse_def, def_all_coverage, def_causal_coverage, def_noncausal_coverage))
}
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.008676 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 86.76 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49700.120 1.000 1.000
Chain 1: 200 -20897.885 1.189 1.378
Chain 1: 300 -18059.343 0.845 1.000
Chain 1: 400 -18870.918 0.645 1.000
Chain 1: 500 -16393.697 0.546 0.157
Chain 1: 600 -12632.725 0.505 0.298
Chain 1: 700 -14970.259 0.455 0.157
Chain 1: 800 -14506.094 0.402 0.157
Chain 1: 900 -11552.639 0.386 0.157
Chain 1: 1000 -19152.592 0.387 0.256
Chain 1: 1100 -18596.025 0.290 0.157
Chain 1: 1200 -11403.896 0.215 0.157
Chain 1: 1300 -13037.097 0.212 0.156
Chain 1: 1400 -11282.999 0.223 0.156
Chain 1: 1500 -11676.713 0.211 0.156
Chain 1: 1600 -11740.951 0.182 0.155
Chain 1: 1700 -12148.375 0.170 0.125
Chain 1: 1800 -11181.154 0.175 0.125
Chain 1: 1900 -11073.978 0.151 0.087
Chain 1: 2000 -12205.852 0.120 0.087
Chain 1: 2100 -11714.121 0.122 0.087
Chain 1: 2200 -11879.519 0.060 0.042
Chain 1: 2300 -18317.082 0.082 0.042
Chain 1: 2400 -10048.753 0.149 0.042
Chain 1: 2500 -10433.995 0.150 0.042
Chain 1: 2600 -10171.449 0.152 0.042
Chain 1: 2700 -10881.492 0.155 0.065
Chain 1: 2800 -17586.888 0.184 0.065
Chain 1: 2900 -11523.739 0.236 0.093
Chain 1: 3000 -17291.079 0.260 0.334
Chain 1: 3100 -10638.473 0.318 0.351
Chain 1: 3200 -9936.393 0.324 0.351
Chain 1: 3300 -10164.480 0.291 0.334
Chain 1: 3400 -13416.738 0.233 0.242
Chain 1: 3500 -14307.945 0.236 0.242
Chain 1: 3600 -10640.556 0.267 0.334
Chain 1: 3700 -9947.948 0.268 0.334
Chain 1: 3800 -11206.992 0.241 0.242
Chain 1: 3900 -14546.368 0.211 0.230
Chain 1: 4000 -10457.604 0.217 0.230
Chain 1: 4100 -11780.649 0.166 0.112
Chain 1: 4200 -11213.899 0.164 0.112
Chain 1: 4300 -9124.151 0.184 0.229
Chain 1: 4400 -8765.716 0.164 0.112
Chain 1: 4500 -9138.568 0.162 0.112
Chain 1: 4600 -8960.912 0.130 0.112
Chain 1: 4700 -15306.317 0.164 0.112
Chain 1: 4800 -9667.296 0.211 0.229
Chain 1: 4900 -11051.895 0.201 0.125
Chain 1: 5000 -10247.646 0.170 0.112
Chain 1: 5100 -9253.745 0.169 0.107
Chain 1: 5200 -9638.156 0.168 0.107
Chain 1: 5300 -12246.290 0.166 0.107
Chain 1: 5400 -8916.423 0.200 0.125
Chain 1: 5500 -9676.128 0.203 0.125
Chain 1: 5600 -10974.674 0.213 0.125
Chain 1: 5700 -14654.241 0.197 0.125
Chain 1: 5800 -9705.381 0.190 0.125
Chain 1: 5900 -14270.785 0.209 0.213
Chain 1: 6000 -11989.303 0.220 0.213
Chain 1: 6100 -9393.491 0.237 0.251
Chain 1: 6200 -11125.570 0.249 0.251
Chain 1: 6300 -9285.662 0.247 0.251
Chain 1: 6400 -13228.224 0.240 0.251
Chain 1: 6500 -9523.821 0.271 0.276
Chain 1: 6600 -9506.314 0.259 0.276
Chain 1: 6700 -8929.827 0.240 0.276
Chain 1: 6800 -9731.294 0.198 0.198
Chain 1: 6900 -10517.531 0.173 0.190
Chain 1: 7000 -8771.463 0.174 0.198
Chain 1: 7100 -13217.988 0.180 0.198
Chain 1: 7200 -10586.770 0.189 0.199
Chain 1: 7300 -9195.446 0.185 0.199
Chain 1: 7400 -11755.669 0.177 0.199
Chain 1: 7500 -9964.139 0.156 0.180
Chain 1: 7600 -9270.456 0.163 0.180
Chain 1: 7700 -8831.011 0.161 0.180
Chain 1: 7800 -8798.246 0.154 0.180
Chain 1: 7900 -8829.098 0.146 0.180
Chain 1: 8000 -8835.454 0.127 0.151
Chain 1: 8100 -9089.226 0.096 0.075
Chain 1: 8200 -8715.796 0.075 0.050
Chain 1: 8300 -8668.834 0.061 0.043
Chain 1: 8400 -9593.847 0.048 0.043
Chain 1: 8500 -11556.073 0.047 0.043
Chain 1: 8600 -8915.974 0.070 0.043
Chain 1: 8700 -8468.806 0.070 0.043
Chain 1: 8800 -8558.708 0.071 0.043
Chain 1: 8900 -9825.711 0.083 0.053
Chain 1: 9000 -9070.447 0.091 0.083
Chain 1: 9100 -10257.115 0.100 0.096
Chain 1: 9200 -10984.155 0.103 0.096
Chain 1: 9300 -9089.613 0.123 0.116
Chain 1: 9400 -8918.947 0.115 0.116
Chain 1: 9500 -9398.515 0.103 0.083
Chain 1: 9600 -10385.187 0.083 0.083
Chain 1: 9700 -9132.623 0.092 0.095
Chain 1: 9800 -9267.848 0.092 0.095
Chain 1: 9900 -11484.163 0.098 0.095
Chain 1: 10000 -8626.034 0.123 0.116
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001473 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.73 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -59086.121 1.000 1.000
Chain 1: 200 -18393.965 1.606 2.212
Chain 1: 300 -9062.464 1.414 1.030
Chain 1: 400 -8235.888 1.086 1.030
Chain 1: 500 -8422.039 0.873 1.000
Chain 1: 600 -8348.637 0.729 1.000
Chain 1: 700 -9471.536 0.642 0.119
Chain 1: 800 -8460.911 0.576 0.119
Chain 1: 900 -7767.646 0.522 0.119
Chain 1: 1000 -8335.766 0.477 0.119
Chain 1: 1100 -8039.663 0.381 0.100
Chain 1: 1200 -7823.504 0.162 0.089
Chain 1: 1300 -7887.134 0.060 0.068
Chain 1: 1400 -7978.699 0.051 0.037
Chain 1: 1500 -7696.074 0.052 0.037
Chain 1: 1600 -8042.683 0.056 0.043
Chain 1: 1700 -7678.447 0.049 0.043
Chain 1: 1800 -7808.227 0.039 0.037
Chain 1: 1900 -7745.168 0.030 0.037
Chain 1: 2000 -7907.054 0.026 0.028
Chain 1: 2100 -7726.342 0.024 0.023
Chain 1: 2200 -7965.225 0.025 0.023
Chain 1: 2300 -7697.285 0.027 0.030
Chain 1: 2400 -7793.724 0.027 0.030
Chain 1: 2500 -7723.665 0.025 0.023
Chain 1: 2600 -7673.716 0.021 0.020
Chain 1: 2700 -7650.298 0.016 0.017
Chain 1: 2800 -7667.139 0.015 0.012
Chain 1: 2900 -7517.549 0.016 0.020
Chain 1: 3000 -7672.557 0.016 0.020
Chain 1: 3100 -7670.065 0.014 0.012
Chain 1: 3200 -7893.207 0.014 0.012
Chain 1: 3300 -7598.824 0.014 0.012
Chain 1: 3400 -7849.921 0.016 0.020
Chain 1: 3500 -7587.319 0.019 0.020
Chain 1: 3600 -7647.232 0.019 0.020
Chain 1: 3700 -7599.897 0.019 0.020
Chain 1: 3800 -7606.827 0.019 0.020
Chain 1: 3900 -7577.700 0.017 0.020
Chain 1: 4000 -7546.207 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003525 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.25 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86850.843 1.000 1.000
Chain 1: 200 -14153.942 3.068 5.136
Chain 1: 300 -10420.365 2.165 1.000
Chain 1: 400 -11696.772 1.651 1.000
Chain 1: 500 -9052.963 1.379 0.358
Chain 1: 600 -8871.892 1.153 0.358
Chain 1: 700 -8826.624 0.989 0.292
Chain 1: 800 -9145.714 0.870 0.292
Chain 1: 900 -9125.137 0.773 0.109
Chain 1: 1000 -9207.952 0.697 0.109
Chain 1: 1100 -8991.719 0.599 0.035 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8714.773 0.089 0.032
Chain 1: 1300 -9032.615 0.056 0.032
Chain 1: 1400 -9066.130 0.046 0.024
Chain 1: 1500 -8894.570 0.019 0.020
Chain 1: 1600 -9005.138 0.018 0.019
Chain 1: 1700 -9058.195 0.018 0.019
Chain 1: 1800 -8612.720 0.020 0.019
Chain 1: 1900 -8719.827 0.021 0.019
Chain 1: 2000 -8701.788 0.020 0.019
Chain 1: 2100 -8839.748 0.019 0.016
Chain 1: 2200 -8614.366 0.018 0.016
Chain 1: 2300 -8711.834 0.016 0.012
Chain 1: 2400 -8786.810 0.016 0.012
Chain 1: 2500 -8726.918 0.015 0.012
Chain 1: 2600 -8743.641 0.014 0.011
Chain 1: 2700 -8649.763 0.015 0.011
Chain 1: 2800 -8595.259 0.010 0.011
Chain 1: 2900 -8700.029 0.010 0.011
Chain 1: 3000 -8539.544 0.012 0.011
Chain 1: 3100 -8679.701 0.012 0.011
Chain 1: 3200 -8548.954 0.011 0.011
Chain 1: 3300 -8779.004 0.012 0.012
Chain 1: 3400 -8785.718 0.012 0.012
Chain 1: 3500 -8654.381 0.012 0.015
Chain 1: 3600 -8506.274 0.014 0.015
Chain 1: 3700 -8653.216 0.015 0.016
Chain 1: 3800 -8508.946 0.016 0.017
Chain 1: 3900 -8440.749 0.015 0.017
Chain 1: 4000 -8551.392 0.015 0.016
Chain 1: 4100 -8515.976 0.013 0.015
Chain 1: 4200 -8501.779 0.012 0.015
Chain 1: 4300 -8535.253 0.010 0.013 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003177 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.77 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8411214.962 1.000 1.000
Chain 1: 200 -1587318.179 2.650 4.299
Chain 1: 300 -891666.459 2.026 1.000
Chain 1: 400 -458854.217 1.756 1.000
Chain 1: 500 -359097.187 1.460 0.943
Chain 1: 600 -233758.192 1.306 0.943
Chain 1: 700 -119909.278 1.255 0.943
Chain 1: 800 -87132.999 1.145 0.943
Chain 1: 900 -67470.208 1.050 0.780
Chain 1: 1000 -52277.027 0.974 0.780
Chain 1: 1100 -39759.083 0.906 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38940.194 0.478 0.376
Chain 1: 1300 -26888.682 0.445 0.376
Chain 1: 1400 -26609.629 0.352 0.315
Chain 1: 1500 -23195.573 0.339 0.315
Chain 1: 1600 -22412.850 0.288 0.291
Chain 1: 1700 -21284.694 0.199 0.291
Chain 1: 1800 -21228.988 0.161 0.147
Chain 1: 1900 -21555.748 0.134 0.053
Chain 1: 2000 -20064.946 0.112 0.053
Chain 1: 2100 -20303.380 0.082 0.035
Chain 1: 2200 -20530.563 0.081 0.035
Chain 1: 2300 -20146.950 0.038 0.019
Chain 1: 2400 -19918.796 0.038 0.019
Chain 1: 2500 -19720.970 0.024 0.015
Chain 1: 2600 -19350.436 0.023 0.015
Chain 1: 2700 -19307.115 0.018 0.012
Chain 1: 2800 -19023.794 0.019 0.015
Chain 1: 2900 -19305.326 0.019 0.015
Chain 1: 3000 -19291.372 0.011 0.012
Chain 1: 3100 -19376.538 0.011 0.011
Chain 1: 3200 -19066.769 0.011 0.015
Chain 1: 3300 -19271.827 0.010 0.011
Chain 1: 3400 -18746.035 0.012 0.015
Chain 1: 3500 -19359.073 0.014 0.015
Chain 1: 3600 -18664.158 0.016 0.015
Chain 1: 3700 -19052.182 0.018 0.016
Chain 1: 3800 -18009.541 0.022 0.020
Chain 1: 3900 -18005.635 0.021 0.020
Chain 1: 4000 -18122.912 0.021 0.020
Chain 1: 4100 -18036.626 0.021 0.020
Chain 1: 4200 -17852.296 0.021 0.020
Chain 1: 4300 -17991.069 0.020 0.020
Chain 1: 4400 -17947.456 0.018 0.010
Chain 1: 4500 -17849.923 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001242 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.42 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49364.995 1.000 1.000
Chain 1: 200 -16523.719 1.494 1.988
Chain 1: 300 -13936.577 1.058 1.000
Chain 1: 400 -18222.176 0.852 1.000
Chain 1: 500 -15694.346 0.714 0.235
Chain 1: 600 -11729.678 0.651 0.338
Chain 1: 700 -14331.165 0.584 0.235
Chain 1: 800 -13753.961 0.516 0.235
Chain 1: 900 -13806.735 0.459 0.186
Chain 1: 1000 -14661.751 0.419 0.186
Chain 1: 1100 -10802.471 0.355 0.186
Chain 1: 1200 -15945.288 0.189 0.186
Chain 1: 1300 -12602.179 0.196 0.235
Chain 1: 1400 -11512.600 0.182 0.182
Chain 1: 1500 -13045.486 0.178 0.182
Chain 1: 1600 -10392.907 0.170 0.182
Chain 1: 1700 -13018.028 0.172 0.202
Chain 1: 1800 -12235.781 0.174 0.202
Chain 1: 1900 -10192.650 0.194 0.202
Chain 1: 2000 -14506.376 0.218 0.255
Chain 1: 2100 -13188.756 0.192 0.202
Chain 1: 2200 -11166.346 0.178 0.200
Chain 1: 2300 -14115.805 0.172 0.200
Chain 1: 2400 -9930.102 0.205 0.202
Chain 1: 2500 -18233.374 0.239 0.209
Chain 1: 2600 -12166.689 0.263 0.209
Chain 1: 2700 -16021.102 0.267 0.241
Chain 1: 2800 -9766.639 0.324 0.297
Chain 1: 2900 -10653.445 0.313 0.297
Chain 1: 3000 -9459.307 0.296 0.241
Chain 1: 3100 -10926.485 0.299 0.241
Chain 1: 3200 -9521.019 0.296 0.241
Chain 1: 3300 -9609.074 0.276 0.241
Chain 1: 3400 -9537.895 0.234 0.148
Chain 1: 3500 -10033.526 0.194 0.134
Chain 1: 3600 -9824.239 0.146 0.126
Chain 1: 3700 -9835.705 0.122 0.083
Chain 1: 3800 -10529.627 0.065 0.066
Chain 1: 3900 -10610.244 0.057 0.049
Chain 1: 4000 -14604.577 0.072 0.049
Chain 1: 4100 -15377.338 0.063 0.049
Chain 1: 4200 -11023.576 0.088 0.049
Chain 1: 4300 -14479.071 0.111 0.050
Chain 1: 4400 -9350.410 0.165 0.066
Chain 1: 4500 -10049.826 0.167 0.070
Chain 1: 4600 -9344.879 0.173 0.075
Chain 1: 4700 -9485.736 0.174 0.075
Chain 1: 4800 -8975.216 0.173 0.075
Chain 1: 4900 -9201.318 0.175 0.075
Chain 1: 5000 -15883.903 0.189 0.075
Chain 1: 5100 -9475.944 0.252 0.239
Chain 1: 5200 -9312.795 0.214 0.075
Chain 1: 5300 -10137.979 0.199 0.075
Chain 1: 5400 -9409.684 0.151 0.075
Chain 1: 5500 -15649.741 0.184 0.077
Chain 1: 5600 -16048.430 0.179 0.077
Chain 1: 5700 -11446.358 0.218 0.081
Chain 1: 5800 -9211.586 0.237 0.243
Chain 1: 5900 -14777.564 0.272 0.377
Chain 1: 6000 -9131.979 0.292 0.377
Chain 1: 6100 -9620.963 0.229 0.243
Chain 1: 6200 -9661.185 0.228 0.243
Chain 1: 6300 -14792.649 0.254 0.347
Chain 1: 6400 -14914.334 0.247 0.347
Chain 1: 6500 -11822.826 0.234 0.261
Chain 1: 6600 -9120.570 0.261 0.296
Chain 1: 6700 -9862.303 0.228 0.261
Chain 1: 6800 -9132.991 0.212 0.261
Chain 1: 6900 -9103.816 0.174 0.080
Chain 1: 7000 -10985.966 0.130 0.080
Chain 1: 7100 -10041.436 0.134 0.094
Chain 1: 7200 -12227.521 0.152 0.171
Chain 1: 7300 -10575.598 0.132 0.156
Chain 1: 7400 -8658.716 0.154 0.171
Chain 1: 7500 -9452.910 0.136 0.156
Chain 1: 7600 -9435.785 0.107 0.094
Chain 1: 7700 -8802.875 0.106 0.094
Chain 1: 7800 -9164.323 0.102 0.094
Chain 1: 7900 -8807.510 0.106 0.094
Chain 1: 8000 -11561.622 0.113 0.094
Chain 1: 8100 -8722.282 0.136 0.156
Chain 1: 8200 -10984.821 0.138 0.156
Chain 1: 8300 -11519.892 0.128 0.084
Chain 1: 8400 -8832.243 0.136 0.084
Chain 1: 8500 -9945.641 0.139 0.112
Chain 1: 8600 -9344.017 0.145 0.112
Chain 1: 8700 -9263.182 0.139 0.112
Chain 1: 8800 -9297.682 0.135 0.112
Chain 1: 8900 -12689.702 0.158 0.206
Chain 1: 9000 -10369.852 0.156 0.206
Chain 1: 9100 -8828.442 0.141 0.175
Chain 1: 9200 -10578.735 0.137 0.165
Chain 1: 9300 -10052.718 0.138 0.165
Chain 1: 9400 -9682.092 0.111 0.112
Chain 1: 9500 -11436.534 0.115 0.153
Chain 1: 9600 -9210.665 0.133 0.165
Chain 1: 9700 -12225.045 0.157 0.175
Chain 1: 9800 -8920.856 0.193 0.224
Chain 1: 9900 -8947.246 0.167 0.175
Chain 1: 10000 -8673.228 0.148 0.165
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001373 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.73 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58869.123 1.000 1.000
Chain 1: 200 -18290.640 1.609 2.219
Chain 1: 300 -8941.529 1.421 1.046
Chain 1: 400 -8156.009 1.090 1.046
Chain 1: 500 -9247.626 0.896 1.000
Chain 1: 600 -8210.267 0.767 1.000
Chain 1: 700 -7818.451 0.665 0.126
Chain 1: 800 -8360.434 0.590 0.126
Chain 1: 900 -8151.787 0.527 0.118
Chain 1: 1000 -8004.713 0.476 0.118
Chain 1: 1100 -7783.492 0.379 0.096
Chain 1: 1200 -7579.317 0.160 0.065
Chain 1: 1300 -7889.244 0.059 0.050
Chain 1: 1400 -7658.497 0.053 0.039
Chain 1: 1500 -7572.256 0.042 0.030
Chain 1: 1600 -7794.757 0.032 0.029
Chain 1: 1700 -7599.778 0.030 0.028
Chain 1: 1800 -7644.890 0.024 0.027
Chain 1: 1900 -7567.699 0.022 0.027
Chain 1: 2000 -7691.989 0.022 0.027
Chain 1: 2100 -7645.228 0.020 0.026
Chain 1: 2200 -7837.515 0.020 0.025
Chain 1: 2300 -7600.894 0.019 0.025
Chain 1: 2400 -7624.112 0.016 0.016
Chain 1: 2500 -7627.258 0.015 0.016
Chain 1: 2600 -7558.357 0.013 0.010
Chain 1: 2700 -7471.914 0.012 0.010
Chain 1: 2800 -7655.369 0.014 0.012
Chain 1: 2900 -7406.611 0.016 0.016
Chain 1: 3000 -7568.595 0.016 0.021
Chain 1: 3100 -7549.726 0.016 0.021
Chain 1: 3200 -7770.074 0.017 0.021
Chain 1: 3300 -7473.905 0.017 0.021
Chain 1: 3400 -7724.679 0.020 0.024
Chain 1: 3500 -7466.136 0.024 0.028
Chain 1: 3600 -7528.995 0.024 0.028
Chain 1: 3700 -7482.347 0.023 0.028
Chain 1: 3800 -7481.369 0.021 0.028
Chain 1: 3900 -7437.611 0.018 0.021
Chain 1: 4000 -7431.064 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002927 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.27 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86661.830 1.000 1.000
Chain 1: 200 -14003.965 3.094 5.188
Chain 1: 300 -10334.333 2.181 1.000
Chain 1: 400 -11369.090 1.659 1.000
Chain 1: 500 -9322.375 1.371 0.355
Chain 1: 600 -8977.574 1.149 0.355
Chain 1: 700 -8839.308 0.987 0.220
Chain 1: 800 -9391.131 0.871 0.220
Chain 1: 900 -9159.096 0.777 0.091
Chain 1: 1000 -9089.133 0.700 0.091
Chain 1: 1100 -9031.718 0.601 0.059 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8790.347 0.085 0.038
Chain 1: 1300 -9021.921 0.052 0.027
Chain 1: 1400 -9047.106 0.043 0.026
Chain 1: 1500 -8892.947 0.023 0.025
Chain 1: 1600 -9006.859 0.020 0.017
Chain 1: 1700 -9084.106 0.019 0.017
Chain 1: 1800 -8661.912 0.018 0.017
Chain 1: 1900 -8762.279 0.017 0.013
Chain 1: 2000 -8736.757 0.016 0.013
Chain 1: 2100 -8861.928 0.017 0.014
Chain 1: 2200 -8666.410 0.017 0.014
Chain 1: 2300 -8757.065 0.015 0.013
Chain 1: 2400 -8826.040 0.016 0.013
Chain 1: 2500 -8772.291 0.015 0.011
Chain 1: 2600 -8773.437 0.013 0.010
Chain 1: 2700 -8690.226 0.013 0.010
Chain 1: 2800 -8650.440 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.005762 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 57.62 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8391889.677 1.000 1.000
Chain 1: 200 -1582144.781 2.652 4.304
Chain 1: 300 -891750.703 2.026 1.000
Chain 1: 400 -458643.574 1.756 1.000
Chain 1: 500 -359056.523 1.460 0.944
Chain 1: 600 -233875.582 1.306 0.944
Chain 1: 700 -119941.592 1.255 0.944
Chain 1: 800 -87067.775 1.145 0.944
Chain 1: 900 -67380.912 1.051 0.774
Chain 1: 1000 -52153.899 0.975 0.774
Chain 1: 1100 -39605.578 0.906 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38780.007 0.478 0.378
Chain 1: 1300 -26716.542 0.446 0.378
Chain 1: 1400 -26433.156 0.352 0.317
Chain 1: 1500 -23014.899 0.340 0.317
Chain 1: 1600 -22229.506 0.290 0.292
Chain 1: 1700 -21101.249 0.200 0.292
Chain 1: 1800 -21044.965 0.162 0.149
Chain 1: 1900 -21371.151 0.135 0.053
Chain 1: 2000 -19880.938 0.113 0.053
Chain 1: 2100 -20119.410 0.083 0.035
Chain 1: 2200 -20346.052 0.082 0.035
Chain 1: 2300 -19963.101 0.038 0.019
Chain 1: 2400 -19735.158 0.038 0.019
Chain 1: 2500 -19537.109 0.025 0.015
Chain 1: 2600 -19167.221 0.023 0.015
Chain 1: 2700 -19124.155 0.018 0.012
Chain 1: 2800 -18840.931 0.019 0.015
Chain 1: 2900 -19122.290 0.019 0.015
Chain 1: 3000 -19108.476 0.012 0.012
Chain 1: 3100 -19193.462 0.011 0.012
Chain 1: 3200 -18884.081 0.011 0.015
Chain 1: 3300 -19088.855 0.011 0.012
Chain 1: 3400 -18563.615 0.012 0.015
Chain 1: 3500 -19175.706 0.014 0.015
Chain 1: 3600 -18482.173 0.016 0.015
Chain 1: 3700 -18869.137 0.018 0.016
Chain 1: 3800 -17828.444 0.022 0.021
Chain 1: 3900 -17824.590 0.021 0.021
Chain 1: 4000 -17941.904 0.021 0.021
Chain 1: 4100 -17855.619 0.022 0.021
Chain 1: 4200 -17671.791 0.021 0.021
Chain 1: 4300 -17810.257 0.021 0.021
Chain 1: 4400 -17767.021 0.018 0.010
Chain 1: 4500 -17669.543 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001243 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.43 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48805.962 1.000 1.000
Chain 1: 200 -15101.698 1.616 2.232
Chain 1: 300 -17183.854 1.118 1.000
Chain 1: 400 -22334.530 0.896 1.000
Chain 1: 500 -13719.271 0.842 0.628
Chain 1: 600 -30694.144 0.794 0.628
Chain 1: 700 -15840.019 0.815 0.628
Chain 1: 800 -14210.393 0.727 0.628
Chain 1: 900 -11106.776 0.677 0.553
Chain 1: 1000 -29741.830 0.672 0.627
Chain 1: 1100 -14080.632 0.684 0.627 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -11867.134 0.479 0.553 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1300 -9905.381 0.487 0.553 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1400 -13701.832 0.491 0.553 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1500 -12631.274 0.437 0.279
Chain 1: 1600 -21841.622 0.424 0.279
Chain 1: 1700 -18965.045 0.345 0.277
Chain 1: 1800 -9684.907 0.430 0.279
Chain 1: 1900 -10035.610 0.405 0.277
Chain 1: 2000 -9144.447 0.352 0.198
Chain 1: 2100 -17431.624 0.289 0.198
Chain 1: 2200 -9388.941 0.356 0.277
Chain 1: 2300 -9705.151 0.339 0.277
Chain 1: 2400 -18170.780 0.358 0.422
Chain 1: 2500 -8782.433 0.456 0.466
Chain 1: 2600 -8823.037 0.415 0.466
Chain 1: 2700 -8857.528 0.400 0.466
Chain 1: 2800 -10666.042 0.321 0.170
Chain 1: 2900 -8885.055 0.338 0.200
Chain 1: 3000 -18017.555 0.378 0.466
Chain 1: 3100 -9217.536 0.426 0.466
Chain 1: 3200 -8592.569 0.348 0.200
Chain 1: 3300 -9514.373 0.354 0.200
Chain 1: 3400 -9558.938 0.308 0.170
Chain 1: 3500 -8748.709 0.211 0.097
Chain 1: 3600 -15061.672 0.252 0.170
Chain 1: 3700 -9085.478 0.318 0.200
Chain 1: 3800 -10336.034 0.313 0.200
Chain 1: 3900 -8503.544 0.314 0.215
Chain 1: 4000 -13763.737 0.302 0.215
Chain 1: 4100 -8678.525 0.265 0.215
Chain 1: 4200 -13362.529 0.293 0.351
Chain 1: 4300 -11589.336 0.298 0.351
Chain 1: 4400 -8565.826 0.333 0.353
Chain 1: 4500 -9062.806 0.329 0.353
Chain 1: 4600 -8920.432 0.289 0.351
Chain 1: 4700 -11847.576 0.248 0.247
Chain 1: 4800 -8340.058 0.278 0.351
Chain 1: 4900 -9108.187 0.265 0.351
Chain 1: 5000 -9269.838 0.228 0.247
Chain 1: 5100 -8890.381 0.174 0.153
Chain 1: 5200 -8452.206 0.144 0.084
Chain 1: 5300 -15201.744 0.173 0.084
Chain 1: 5400 -8449.540 0.218 0.084
Chain 1: 5500 -12439.622 0.244 0.247
Chain 1: 5600 -9113.532 0.279 0.321
Chain 1: 5700 -8567.508 0.261 0.321
Chain 1: 5800 -8288.152 0.222 0.084
Chain 1: 5900 -8890.095 0.221 0.068
Chain 1: 6000 -8500.311 0.223 0.068
Chain 1: 6100 -9414.397 0.229 0.097
Chain 1: 6200 -8844.535 0.230 0.097
Chain 1: 6300 -13055.957 0.218 0.097
Chain 1: 6400 -11440.644 0.152 0.097
Chain 1: 6500 -11428.516 0.120 0.068
Chain 1: 6600 -10493.627 0.093 0.068
Chain 1: 6700 -8688.289 0.107 0.089
Chain 1: 6800 -8070.182 0.111 0.089
Chain 1: 6900 -10291.580 0.126 0.097
Chain 1: 7000 -8225.920 0.147 0.141
Chain 1: 7100 -7917.486 0.141 0.141
Chain 1: 7200 -10443.573 0.159 0.208
Chain 1: 7300 -8864.365 0.144 0.178
Chain 1: 7400 -10258.122 0.144 0.178
Chain 1: 7500 -8914.219 0.159 0.178
Chain 1: 7600 -8055.301 0.160 0.178
Chain 1: 7700 -8147.830 0.141 0.151
Chain 1: 7800 -9020.646 0.143 0.151
Chain 1: 7900 -8013.509 0.134 0.136
Chain 1: 8000 -8712.297 0.117 0.126
Chain 1: 8100 -10651.011 0.131 0.136
Chain 1: 8200 -8584.733 0.131 0.136
Chain 1: 8300 -12349.788 0.143 0.136
Chain 1: 8400 -10394.319 0.149 0.151
Chain 1: 8500 -8091.322 0.162 0.182
Chain 1: 8600 -11358.962 0.180 0.188
Chain 1: 8700 -9689.858 0.196 0.188
Chain 1: 8800 -8208.698 0.205 0.188
Chain 1: 8900 -10194.738 0.212 0.195
Chain 1: 9000 -8039.294 0.230 0.241
Chain 1: 9100 -8077.489 0.213 0.241
Chain 1: 9200 -9365.062 0.202 0.195
Chain 1: 9300 -10175.467 0.180 0.188
Chain 1: 9400 -11642.770 0.174 0.180
Chain 1: 9500 -11121.376 0.150 0.172
Chain 1: 9600 -7993.006 0.160 0.172
Chain 1: 9700 -10082.215 0.164 0.180
Chain 1: 9800 -8475.552 0.165 0.190
Chain 1: 9900 -10065.340 0.161 0.158
Chain 1: 10000 -8441.054 0.153 0.158
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001429 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.29 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56660.572 1.000 1.000
Chain 1: 200 -17016.380 1.665 2.330
Chain 1: 300 -8436.901 1.449 1.017
Chain 1: 400 -7794.339 1.107 1.017
Chain 1: 500 -8275.042 0.897 1.000
Chain 1: 600 -7921.495 0.755 1.000
Chain 1: 700 -7660.047 0.652 0.082
Chain 1: 800 -7995.089 0.576 0.082
Chain 1: 900 -7831.980 0.514 0.058
Chain 1: 1000 -7534.223 0.467 0.058
Chain 1: 1100 -7568.452 0.367 0.045
Chain 1: 1200 -7648.353 0.135 0.042
Chain 1: 1300 -7584.862 0.034 0.040
Chain 1: 1400 -7723.494 0.028 0.034
Chain 1: 1500 -7564.092 0.024 0.021
Chain 1: 1600 -7484.147 0.021 0.021
Chain 1: 1700 -7437.977 0.018 0.018
Chain 1: 1800 -7512.339 0.015 0.011
Chain 1: 1900 -7536.674 0.013 0.010
Chain 1: 2000 -7518.824 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003067 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.67 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86092.376 1.000 1.000
Chain 1: 200 -13099.458 3.286 5.572
Chain 1: 300 -9541.622 2.315 1.000
Chain 1: 400 -10385.777 1.757 1.000
Chain 1: 500 -8450.564 1.451 0.373
Chain 1: 600 -8092.444 1.217 0.373
Chain 1: 700 -8146.393 1.044 0.229
Chain 1: 800 -8526.206 0.919 0.229
Chain 1: 900 -8441.393 0.818 0.081
Chain 1: 1000 -8142.769 0.740 0.081
Chain 1: 1100 -8392.678 0.643 0.045 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8046.194 0.090 0.044
Chain 1: 1300 -8284.226 0.055 0.043
Chain 1: 1400 -8270.584 0.047 0.037
Chain 1: 1500 -8163.988 0.026 0.030
Chain 1: 1600 -8264.722 0.023 0.029
Chain 1: 1700 -8353.990 0.023 0.029
Chain 1: 1800 -7964.226 0.023 0.029
Chain 1: 1900 -8066.486 0.024 0.029
Chain 1: 2000 -8036.603 0.020 0.013
Chain 1: 2100 -8166.645 0.019 0.013
Chain 1: 2200 -7952.998 0.017 0.013
Chain 1: 2300 -8095.654 0.016 0.013
Chain 1: 2400 -8108.952 0.016 0.013
Chain 1: 2500 -8076.539 0.015 0.013
Chain 1: 2600 -8077.208 0.014 0.013
Chain 1: 2700 -7984.914 0.014 0.013
Chain 1: 2800 -7959.964 0.010 0.012 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.010065 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 100.65 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8433208.743 1.000 1.000
Chain 1: 200 -1587313.775 2.656 4.313
Chain 1: 300 -889653.158 2.032 1.000
Chain 1: 400 -456670.595 1.761 1.000
Chain 1: 500 -356700.092 1.465 0.948
Chain 1: 600 -231830.659 1.311 0.948
Chain 1: 700 -118430.653 1.260 0.948
Chain 1: 800 -85723.236 1.150 0.948
Chain 1: 900 -66137.269 1.055 0.784
Chain 1: 1000 -50985.758 0.980 0.784
Chain 1: 1100 -38519.804 0.912 0.539 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37698.878 0.483 0.382
Chain 1: 1300 -25723.283 0.451 0.382
Chain 1: 1400 -25446.793 0.357 0.324
Chain 1: 1500 -22051.835 0.345 0.324
Chain 1: 1600 -21273.054 0.294 0.297
Chain 1: 1700 -20155.550 0.204 0.296
Chain 1: 1800 -20101.507 0.166 0.154
Chain 1: 1900 -20427.183 0.138 0.055
Chain 1: 2000 -18943.844 0.116 0.055
Chain 1: 2100 -19181.901 0.085 0.037
Chain 1: 2200 -19407.247 0.084 0.037
Chain 1: 2300 -19025.561 0.040 0.020
Chain 1: 2400 -18797.911 0.040 0.020
Chain 1: 2500 -18599.626 0.026 0.016
Chain 1: 2600 -18230.594 0.024 0.016
Chain 1: 2700 -18187.868 0.019 0.012
Chain 1: 2800 -17904.768 0.020 0.016
Chain 1: 2900 -18185.732 0.020 0.015
Chain 1: 3000 -18172.025 0.012 0.012
Chain 1: 3100 -18256.887 0.011 0.012
Chain 1: 3200 -17948.002 0.012 0.015
Chain 1: 3300 -18152.415 0.011 0.012
Chain 1: 3400 -17627.933 0.013 0.015
Chain 1: 3500 -18238.768 0.015 0.016
Chain 1: 3600 -17546.818 0.017 0.016
Chain 1: 3700 -17932.526 0.019 0.017
Chain 1: 3800 -16894.238 0.023 0.022
Chain 1: 3900 -16890.391 0.022 0.022
Chain 1: 4000 -17007.744 0.023 0.022
Chain 1: 4100 -16921.541 0.023 0.022
Chain 1: 4200 -16738.262 0.022 0.022
Chain 1: 4300 -16876.360 0.022 0.022
Chain 1: 4400 -16833.555 0.019 0.011
Chain 1: 4500 -16736.110 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001279 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.79 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12185.413 1.000 1.000
Chain 1: 200 -9097.688 0.670 1.000
Chain 1: 300 -8015.756 0.491 0.339
Chain 1: 400 -8132.492 0.372 0.339
Chain 1: 500 -7821.237 0.306 0.135
Chain 1: 600 -7843.286 0.255 0.135
Chain 1: 700 -7801.028 0.220 0.040
Chain 1: 800 -7813.595 0.192 0.040
Chain 1: 900 -7719.932 0.172 0.014
Chain 1: 1000 -7822.074 0.156 0.014
Chain 1: 1100 -7840.412 0.057 0.013
Chain 1: 1200 -7814.485 0.023 0.012
Chain 1: 1300 -7768.203 0.010 0.006 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001427 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.27 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57314.037 1.000 1.000
Chain 1: 200 -17312.591 1.655 2.311
Chain 1: 300 -8607.383 1.441 1.011
Chain 1: 400 -7996.596 1.100 1.011
Chain 1: 500 -8233.991 0.885 1.000
Chain 1: 600 -7992.344 0.743 1.000
Chain 1: 700 -7821.876 0.640 0.076
Chain 1: 800 -8108.138 0.564 0.076
Chain 1: 900 -7969.709 0.504 0.035
Chain 1: 1000 -7869.060 0.454 0.035
Chain 1: 1100 -7803.152 0.355 0.030
Chain 1: 1200 -7960.752 0.126 0.029
Chain 1: 1300 -7697.660 0.029 0.029
Chain 1: 1400 -8015.080 0.025 0.029
Chain 1: 1500 -7704.858 0.026 0.030
Chain 1: 1600 -7619.218 0.024 0.022
Chain 1: 1700 -7617.174 0.022 0.020
Chain 1: 1800 -7633.258 0.019 0.017
Chain 1: 1900 -7714.017 0.018 0.013
Chain 1: 2000 -7700.362 0.017 0.011
Chain 1: 2100 -7724.237 0.016 0.011
Chain 1: 2200 -7770.342 0.015 0.010
Chain 1: 2300 -7679.306 0.013 0.010
Chain 1: 2400 -7731.770 0.009 0.007 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003192 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.92 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86864.551 1.000 1.000
Chain 1: 200 -13212.201 3.287 5.575
Chain 1: 300 -9659.278 2.314 1.000
Chain 1: 400 -10470.226 1.755 1.000
Chain 1: 500 -8562.405 1.449 0.368
Chain 1: 600 -8262.491 1.213 0.368
Chain 1: 700 -8225.569 1.040 0.223
Chain 1: 800 -8461.266 0.914 0.223
Chain 1: 900 -8533.858 0.813 0.077
Chain 1: 1000 -8265.000 0.735 0.077
Chain 1: 1100 -8516.369 0.638 0.036 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8199.723 0.085 0.036
Chain 1: 1300 -8386.838 0.050 0.033
Chain 1: 1400 -8330.341 0.043 0.030
Chain 1: 1500 -8286.872 0.021 0.028
Chain 1: 1600 -8283.711 0.018 0.022
Chain 1: 1700 -8217.549 0.018 0.022
Chain 1: 1800 -8099.210 0.017 0.015
Chain 1: 1900 -8214.543 0.017 0.015
Chain 1: 2000 -8175.375 0.014 0.014
Chain 1: 2100 -8310.854 0.013 0.014
Chain 1: 2200 -8098.951 0.012 0.014
Chain 1: 2300 -8239.742 0.011 0.014
Chain 1: 2400 -8250.220 0.011 0.014
Chain 1: 2500 -8218.825 0.011 0.014
Chain 1: 2600 -8214.608 0.011 0.014
Chain 1: 2700 -8124.607 0.011 0.014
Chain 1: 2800 -8104.015 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003354 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.54 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8406881.767 1.000 1.000
Chain 1: 200 -1588073.038 2.647 4.294
Chain 1: 300 -891439.987 2.025 1.000
Chain 1: 400 -457711.137 1.756 1.000
Chain 1: 500 -357531.181 1.461 0.948
Chain 1: 600 -232445.727 1.307 0.948
Chain 1: 700 -118749.071 1.257 0.948
Chain 1: 800 -85967.821 1.147 0.948
Chain 1: 900 -66340.400 1.053 0.781
Chain 1: 1000 -51155.150 0.977 0.781
Chain 1: 1100 -38656.967 0.910 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37833.006 0.482 0.381
Chain 1: 1300 -25830.147 0.451 0.381
Chain 1: 1400 -25550.879 0.357 0.323
Chain 1: 1500 -22148.381 0.344 0.323
Chain 1: 1600 -21366.882 0.294 0.297
Chain 1: 1700 -20246.323 0.204 0.296
Chain 1: 1800 -20191.450 0.166 0.154
Chain 1: 1900 -20516.932 0.138 0.055
Chain 1: 2000 -19032.052 0.116 0.055
Chain 1: 2100 -19270.394 0.085 0.037
Chain 1: 2200 -19495.782 0.084 0.037
Chain 1: 2300 -19114.040 0.040 0.020
Chain 1: 2400 -18886.376 0.040 0.020
Chain 1: 2500 -18688.155 0.026 0.016
Chain 1: 2600 -18319.286 0.024 0.016
Chain 1: 2700 -18276.526 0.019 0.012
Chain 1: 2800 -17993.520 0.020 0.016
Chain 1: 2900 -18274.454 0.020 0.015
Chain 1: 3000 -18260.773 0.012 0.012
Chain 1: 3100 -18345.629 0.011 0.012
Chain 1: 3200 -18036.813 0.012 0.015
Chain 1: 3300 -18241.140 0.011 0.012
Chain 1: 3400 -17716.856 0.013 0.015
Chain 1: 3500 -18327.462 0.015 0.016
Chain 1: 3600 -17635.781 0.017 0.016
Chain 1: 3700 -18021.312 0.019 0.017
Chain 1: 3800 -16983.507 0.023 0.021
Chain 1: 3900 -16979.655 0.022 0.021
Chain 1: 4000 -17097.008 0.022 0.021
Chain 1: 4100 -17010.844 0.023 0.021
Chain 1: 4200 -16827.638 0.022 0.021
Chain 1: 4300 -16965.680 0.022 0.021
Chain 1: 4400 -16922.959 0.019 0.011
Chain 1: 4500 -16825.516 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001423 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.23 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13291.750 1.000 1.000
Chain 1: 200 -10241.059 0.649 1.000
Chain 1: 300 -8553.485 0.498 0.298
Chain 1: 400 -8779.796 0.380 0.298
Chain 1: 500 -8686.861 0.306 0.197
Chain 1: 600 -8491.821 0.259 0.197
Chain 1: 700 -8379.219 0.224 0.026
Chain 1: 800 -8425.197 0.197 0.026
Chain 1: 900 -8511.846 0.176 0.023
Chain 1: 1000 -8445.643 0.159 0.023
Chain 1: 1100 -8418.390 0.059 0.013
Chain 1: 1200 -8410.446 0.030 0.011
Chain 1: 1300 -8349.311 0.011 0.010
Chain 1: 1400 -8381.594 0.009 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002805 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 28.05 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58132.904 1.000 1.000
Chain 1: 200 -18462.331 1.574 2.149
Chain 1: 300 -9255.234 1.381 1.000
Chain 1: 400 -8308.791 1.064 1.000
Chain 1: 500 -8874.743 0.864 0.995
Chain 1: 600 -9017.113 0.723 0.995
Chain 1: 700 -8532.115 0.628 0.114
Chain 1: 800 -8374.524 0.552 0.114
Chain 1: 900 -8324.046 0.491 0.064
Chain 1: 1000 -7893.597 0.447 0.064
Chain 1: 1100 -7576.897 0.352 0.057
Chain 1: 1200 -7664.327 0.138 0.055
Chain 1: 1300 -8203.431 0.045 0.055
Chain 1: 1400 -7696.183 0.040 0.055
Chain 1: 1500 -7621.746 0.035 0.042
Chain 1: 1600 -7867.847 0.036 0.042
Chain 1: 1700 -7743.682 0.032 0.031
Chain 1: 1800 -7718.281 0.031 0.031
Chain 1: 1900 -7672.206 0.031 0.031
Chain 1: 2000 -7774.645 0.026 0.016
Chain 1: 2100 -7579.770 0.025 0.016
Chain 1: 2200 -7966.344 0.029 0.026
Chain 1: 2300 -7632.509 0.026 0.026
Chain 1: 2400 -7615.255 0.020 0.016
Chain 1: 2500 -7443.456 0.021 0.023
Chain 1: 2600 -7627.388 0.021 0.023
Chain 1: 2700 -7586.148 0.020 0.023
Chain 1: 2800 -7544.400 0.020 0.023
Chain 1: 2900 -7573.335 0.020 0.023
Chain 1: 3000 -7578.698 0.018 0.023
Chain 1: 3100 -7580.138 0.016 0.006 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003029 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.29 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87583.579 1.000 1.000
Chain 1: 200 -14456.938 3.029 5.058
Chain 1: 300 -10642.012 2.139 1.000
Chain 1: 400 -12651.329 1.644 1.000
Chain 1: 500 -9052.086 1.395 0.398
Chain 1: 600 -9103.573 1.163 0.398
Chain 1: 700 -8845.977 1.001 0.358
Chain 1: 800 -9553.682 0.885 0.358
Chain 1: 900 -9354.408 0.789 0.159
Chain 1: 1000 -9632.849 0.713 0.159
Chain 1: 1100 -9341.114 0.616 0.074 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8816.018 0.116 0.060
Chain 1: 1300 -9238.177 0.085 0.046
Chain 1: 1400 -9161.413 0.070 0.031
Chain 1: 1500 -9078.169 0.031 0.029
Chain 1: 1600 -9120.113 0.031 0.029
Chain 1: 1700 -9205.390 0.029 0.029
Chain 1: 1800 -8738.909 0.027 0.029
Chain 1: 1900 -8853.495 0.026 0.029
Chain 1: 2000 -8870.581 0.024 0.013
Chain 1: 2100 -8965.244 0.022 0.011
Chain 1: 2200 -8731.275 0.018 0.011
Chain 1: 2300 -8902.509 0.016 0.011
Chain 1: 2400 -8760.458 0.016 0.013
Chain 1: 2500 -8821.825 0.016 0.013
Chain 1: 2600 -8728.562 0.017 0.013
Chain 1: 2700 -8764.012 0.016 0.013
Chain 1: 2800 -8722.098 0.011 0.011
Chain 1: 2900 -8830.825 0.011 0.011
Chain 1: 3000 -8738.834 0.012 0.011
Chain 1: 3100 -8706.144 0.012 0.011
Chain 1: 3200 -8675.226 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003315 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.15 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8403134.862 1.000 1.000
Chain 1: 200 -1583704.122 2.653 4.306
Chain 1: 300 -890001.256 2.028 1.000
Chain 1: 400 -457329.280 1.758 1.000
Chain 1: 500 -357658.547 1.462 0.946
Chain 1: 600 -233041.403 1.307 0.946
Chain 1: 700 -119795.077 1.256 0.945
Chain 1: 800 -87131.833 1.146 0.945
Chain 1: 900 -67577.779 1.051 0.779
Chain 1: 1000 -52458.766 0.974 0.779
Chain 1: 1100 -39996.898 0.905 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39193.062 0.477 0.375
Chain 1: 1300 -27188.330 0.443 0.375
Chain 1: 1400 -26917.131 0.349 0.312
Chain 1: 1500 -23512.806 0.336 0.312
Chain 1: 1600 -22733.844 0.286 0.289
Chain 1: 1700 -21610.604 0.197 0.288
Chain 1: 1800 -21556.367 0.159 0.145
Chain 1: 1900 -21883.604 0.132 0.052
Chain 1: 2000 -20394.366 0.111 0.052
Chain 1: 2100 -20632.980 0.081 0.034
Chain 1: 2200 -20859.805 0.080 0.034
Chain 1: 2300 -20476.390 0.037 0.019
Chain 1: 2400 -20248.116 0.037 0.019
Chain 1: 2500 -20049.923 0.024 0.015
Chain 1: 2600 -19679.313 0.022 0.015
Chain 1: 2700 -19636.066 0.017 0.012
Chain 1: 2800 -19352.345 0.019 0.015
Chain 1: 2900 -19634.028 0.019 0.014
Chain 1: 3000 -19620.242 0.011 0.012
Chain 1: 3100 -19705.353 0.011 0.011
Chain 1: 3200 -19395.410 0.011 0.014
Chain 1: 3300 -19600.655 0.010 0.011
Chain 1: 3400 -19074.350 0.012 0.014
Chain 1: 3500 -19687.977 0.014 0.015
Chain 1: 3600 -18992.357 0.016 0.015
Chain 1: 3700 -19380.822 0.018 0.016
Chain 1: 3800 -18336.911 0.022 0.020
Chain 1: 3900 -18332.917 0.020 0.020
Chain 1: 4000 -18450.282 0.021 0.020
Chain 1: 4100 -18363.817 0.021 0.020
Chain 1: 4200 -18179.292 0.020 0.020
Chain 1: 4300 -18318.285 0.020 0.020
Chain 1: 4400 -18274.465 0.018 0.010
Chain 1: 4500 -18176.820 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001365 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.65 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48587.435 1.000 1.000
Chain 1: 200 -17264.644 1.407 1.814
Chain 1: 300 -20931.313 0.996 1.000
Chain 1: 400 -14294.341 0.863 1.000
Chain 1: 500 -29343.939 0.793 0.513
Chain 1: 600 -15933.365 0.801 0.842
Chain 1: 700 -14595.327 0.700 0.513
Chain 1: 800 -11144.730 0.651 0.513
Chain 1: 900 -14148.824 0.602 0.464
Chain 1: 1000 -11511.547 0.565 0.464
Chain 1: 1100 -11843.719 0.468 0.310
Chain 1: 1200 -10653.556 0.298 0.229
Chain 1: 1300 -9630.838 0.291 0.229
Chain 1: 1400 -10122.911 0.249 0.212
Chain 1: 1500 -10390.390 0.200 0.112
Chain 1: 1600 -11135.381 0.123 0.106
Chain 1: 1700 -12584.863 0.125 0.112
Chain 1: 1800 -9767.969 0.123 0.112
Chain 1: 1900 -9956.498 0.104 0.106
Chain 1: 2000 -10827.783 0.089 0.080
Chain 1: 2100 -10319.144 0.091 0.080
Chain 1: 2200 -10475.650 0.081 0.067
Chain 1: 2300 -10121.787 0.074 0.049
Chain 1: 2400 -9133.071 0.080 0.067
Chain 1: 2500 -13320.938 0.109 0.080
Chain 1: 2600 -9186.522 0.147 0.108
Chain 1: 2700 -10022.834 0.144 0.083
Chain 1: 2800 -9576.910 0.120 0.080
Chain 1: 2900 -9153.468 0.123 0.080
Chain 1: 3000 -9550.703 0.119 0.049
Chain 1: 3100 -9536.084 0.114 0.047
Chain 1: 3200 -9338.288 0.115 0.047
Chain 1: 3300 -8898.427 0.116 0.049
Chain 1: 3400 -13681.458 0.140 0.049
Chain 1: 3500 -9641.576 0.151 0.049
Chain 1: 3600 -10162.833 0.111 0.049
Chain 1: 3700 -9379.825 0.111 0.049
Chain 1: 3800 -8664.695 0.115 0.051
Chain 1: 3900 -13406.041 0.145 0.083
Chain 1: 4000 -9874.288 0.177 0.083
Chain 1: 4100 -8701.686 0.190 0.135
Chain 1: 4200 -9659.436 0.198 0.135
Chain 1: 4300 -12107.369 0.213 0.202
Chain 1: 4400 -8596.728 0.219 0.202
Chain 1: 4500 -8919.033 0.181 0.135
Chain 1: 4600 -13241.510 0.208 0.202
Chain 1: 4700 -13008.792 0.202 0.202
Chain 1: 4800 -8763.149 0.242 0.326
Chain 1: 4900 -14607.429 0.247 0.326
Chain 1: 5000 -9104.143 0.271 0.326
Chain 1: 5100 -8971.634 0.259 0.326
Chain 1: 5200 -10776.956 0.266 0.326
Chain 1: 5300 -10463.342 0.249 0.326
Chain 1: 5400 -8463.758 0.232 0.236
Chain 1: 5500 -11985.749 0.258 0.294
Chain 1: 5600 -11636.423 0.228 0.236
Chain 1: 5700 -9298.168 0.251 0.251
Chain 1: 5800 -8339.259 0.214 0.236
Chain 1: 5900 -13436.745 0.212 0.236
Chain 1: 6000 -9225.084 0.197 0.236
Chain 1: 6100 -8562.554 0.204 0.236
Chain 1: 6200 -8257.408 0.191 0.236
Chain 1: 6300 -12425.374 0.221 0.251
Chain 1: 6400 -8749.280 0.240 0.294
Chain 1: 6500 -10950.661 0.230 0.251
Chain 1: 6600 -9030.739 0.249 0.251
Chain 1: 6700 -13026.586 0.254 0.307
Chain 1: 6800 -8633.371 0.294 0.335
Chain 1: 6900 -12448.058 0.286 0.307
Chain 1: 7000 -8851.159 0.281 0.307
Chain 1: 7100 -8220.573 0.281 0.307
Chain 1: 7200 -9219.968 0.288 0.307
Chain 1: 7300 -13797.844 0.288 0.307
Chain 1: 7400 -8688.273 0.305 0.307
Chain 1: 7500 -12460.228 0.315 0.307
Chain 1: 7600 -8817.669 0.335 0.332
Chain 1: 7700 -8412.338 0.309 0.332
Chain 1: 7800 -11629.284 0.286 0.306
Chain 1: 7900 -8987.623 0.285 0.303
Chain 1: 8000 -8574.834 0.249 0.294
Chain 1: 8100 -8321.887 0.244 0.294
Chain 1: 8200 -9112.943 0.242 0.294
Chain 1: 8300 -9161.514 0.209 0.277
Chain 1: 8400 -8469.920 0.159 0.087
Chain 1: 8500 -8362.150 0.130 0.082
Chain 1: 8600 -8505.024 0.090 0.048
Chain 1: 8700 -9808.710 0.099 0.082
Chain 1: 8800 -8230.465 0.090 0.082
Chain 1: 8900 -9931.364 0.078 0.082
Chain 1: 9000 -8070.507 0.096 0.087
Chain 1: 9100 -10401.050 0.115 0.133
Chain 1: 9200 -8636.226 0.127 0.171
Chain 1: 9300 -8397.321 0.129 0.171
Chain 1: 9400 -9616.983 0.134 0.171
Chain 1: 9500 -8089.173 0.152 0.189
Chain 1: 9600 -10367.839 0.172 0.192
Chain 1: 9700 -8163.576 0.186 0.204
Chain 1: 9800 -8215.419 0.167 0.204
Chain 1: 9900 -8823.837 0.157 0.204
Chain 1: 10000 -8202.842 0.141 0.189
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001428 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.28 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56991.513 1.000 1.000
Chain 1: 200 -17243.610 1.653 2.305
Chain 1: 300 -8662.646 1.432 1.000
Chain 1: 400 -8190.024 1.088 1.000
Chain 1: 500 -8321.830 0.874 0.991
Chain 1: 600 -8216.679 0.730 0.991
Chain 1: 700 -8095.025 0.628 0.058
Chain 1: 800 -8029.530 0.551 0.058
Chain 1: 900 -7724.079 0.494 0.040
Chain 1: 1000 -7841.520 0.446 0.040
Chain 1: 1100 -7743.798 0.347 0.016
Chain 1: 1200 -7586.956 0.119 0.016
Chain 1: 1300 -7724.656 0.022 0.016
Chain 1: 1400 -7864.698 0.018 0.016
Chain 1: 1500 -7612.855 0.019 0.018
Chain 1: 1600 -7670.902 0.019 0.018
Chain 1: 1700 -7519.693 0.019 0.018
Chain 1: 1800 -7590.255 0.019 0.018
Chain 1: 1900 -7566.983 0.016 0.018
Chain 1: 2000 -7620.288 0.015 0.018
Chain 1: 2100 -7597.988 0.014 0.018
Chain 1: 2200 -7696.394 0.013 0.013
Chain 1: 2300 -7589.718 0.013 0.013
Chain 1: 2400 -7638.062 0.012 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002592 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.92 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85854.368 1.000 1.000
Chain 1: 200 -13338.824 3.218 5.436
Chain 1: 300 -9798.195 2.266 1.000
Chain 1: 400 -10713.388 1.721 1.000
Chain 1: 500 -8697.882 1.423 0.361
Chain 1: 600 -8424.476 1.191 0.361
Chain 1: 700 -8476.051 1.022 0.232
Chain 1: 800 -8831.316 0.899 0.232
Chain 1: 900 -8685.617 0.801 0.085
Chain 1: 1000 -8357.220 0.725 0.085
Chain 1: 1100 -8708.030 0.629 0.040 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8379.730 0.089 0.040
Chain 1: 1300 -8373.233 0.053 0.039
Chain 1: 1400 -8374.716 0.045 0.039
Chain 1: 1500 -8407.232 0.022 0.032
Chain 1: 1600 -8412.911 0.019 0.017
Chain 1: 1700 -8344.426 0.019 0.017
Chain 1: 1800 -8227.475 0.016 0.014
Chain 1: 1900 -8344.381 0.016 0.014
Chain 1: 2000 -8304.569 0.013 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002503 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.03 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8389132.857 1.000 1.000
Chain 1: 200 -1581617.672 2.652 4.304
Chain 1: 300 -891344.061 2.026 1.000
Chain 1: 400 -458117.480 1.756 1.000
Chain 1: 500 -358768.302 1.460 0.946
Chain 1: 600 -233485.214 1.306 0.946
Chain 1: 700 -119373.725 1.256 0.946
Chain 1: 800 -86505.230 1.147 0.946
Chain 1: 900 -66776.335 1.052 0.774
Chain 1: 1000 -51514.699 0.977 0.774
Chain 1: 1100 -38943.866 0.909 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38108.050 0.481 0.380
Chain 1: 1300 -26025.196 0.450 0.380
Chain 1: 1400 -25737.163 0.356 0.323
Chain 1: 1500 -22315.532 0.344 0.323
Chain 1: 1600 -21528.611 0.294 0.296
Chain 1: 1700 -20398.351 0.204 0.295
Chain 1: 1800 -20341.123 0.166 0.153
Chain 1: 1900 -20666.723 0.138 0.055
Chain 1: 2000 -19176.957 0.116 0.055
Chain 1: 2100 -19415.157 0.085 0.037
Chain 1: 2200 -19641.732 0.084 0.037
Chain 1: 2300 -19258.996 0.040 0.020
Chain 1: 2400 -19031.298 0.040 0.020
Chain 1: 2500 -18833.558 0.025 0.016
Chain 1: 2600 -18464.090 0.024 0.016
Chain 1: 2700 -18421.112 0.018 0.012
Chain 1: 2800 -18138.396 0.020 0.016
Chain 1: 2900 -18419.404 0.020 0.015
Chain 1: 3000 -18405.543 0.012 0.012
Chain 1: 3100 -18490.497 0.011 0.012
Chain 1: 3200 -18181.477 0.012 0.015
Chain 1: 3300 -18385.950 0.011 0.012
Chain 1: 3400 -17861.563 0.013 0.015
Chain 1: 3500 -18472.491 0.015 0.016
Chain 1: 3600 -17780.416 0.017 0.016
Chain 1: 3700 -18166.379 0.019 0.017
Chain 1: 3800 -17128.074 0.023 0.021
Chain 1: 3900 -17124.322 0.022 0.021
Chain 1: 4000 -17241.565 0.022 0.021
Chain 1: 4100 -17155.497 0.022 0.021
Chain 1: 4200 -16972.150 0.022 0.021
Chain 1: 4300 -17110.212 0.021 0.021
Chain 1: 4400 -17067.398 0.019 0.011
Chain 1: 4500 -16970.051 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001454 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.54 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48832.371 1.000 1.000
Chain 1: 200 -16728.004 1.460 1.919
Chain 1: 300 -16078.536 0.987 1.000
Chain 1: 400 -14910.004 0.759 1.000
Chain 1: 500 -13432.432 0.630 0.110
Chain 1: 600 -14780.211 0.540 0.110
Chain 1: 700 -17628.050 0.486 0.110
Chain 1: 800 -15844.076 0.439 0.113
Chain 1: 900 -13597.086 0.409 0.113
Chain 1: 1000 -10397.735 0.399 0.162
Chain 1: 1100 -10656.624 0.301 0.113
Chain 1: 1200 -10753.438 0.110 0.110
Chain 1: 1300 -12611.490 0.121 0.113
Chain 1: 1400 -10808.974 0.130 0.147
Chain 1: 1500 -10633.824 0.120 0.147
Chain 1: 1600 -9682.620 0.121 0.147
Chain 1: 1700 -13756.586 0.134 0.147
Chain 1: 1800 -11099.633 0.147 0.165
Chain 1: 1900 -10236.659 0.139 0.147
Chain 1: 2000 -10399.358 0.110 0.098
Chain 1: 2100 -10239.640 0.109 0.098
Chain 1: 2200 -9961.388 0.111 0.098
Chain 1: 2300 -9216.305 0.104 0.084
Chain 1: 2400 -10386.250 0.099 0.084
Chain 1: 2500 -10160.733 0.099 0.084
Chain 1: 2600 -10495.265 0.093 0.081
Chain 1: 2700 -13119.726 0.083 0.081
Chain 1: 2800 -9752.489 0.094 0.081
Chain 1: 2900 -18590.176 0.133 0.081
Chain 1: 3000 -15624.819 0.150 0.113
Chain 1: 3100 -8979.771 0.223 0.190
Chain 1: 3200 -8749.854 0.222 0.190
Chain 1: 3300 -12462.140 0.244 0.200
Chain 1: 3400 -9081.580 0.270 0.298
Chain 1: 3500 -12578.827 0.296 0.298
Chain 1: 3600 -9569.744 0.324 0.314
Chain 1: 3700 -14502.639 0.338 0.340
Chain 1: 3800 -8436.773 0.375 0.340
Chain 1: 3900 -9442.009 0.338 0.314
Chain 1: 4000 -16645.140 0.363 0.340
Chain 1: 4100 -13486.430 0.312 0.314
Chain 1: 4200 -9816.315 0.347 0.340
Chain 1: 4300 -9737.078 0.318 0.340
Chain 1: 4400 -8865.907 0.291 0.314
Chain 1: 4500 -9244.151 0.267 0.314
Chain 1: 4600 -13339.554 0.266 0.307
Chain 1: 4700 -8540.976 0.288 0.307
Chain 1: 4800 -8800.510 0.219 0.234
Chain 1: 4900 -12396.900 0.238 0.290
Chain 1: 5000 -9600.593 0.224 0.290
Chain 1: 5100 -8536.628 0.213 0.290
Chain 1: 5200 -8630.811 0.176 0.125
Chain 1: 5300 -12579.601 0.207 0.290
Chain 1: 5400 -8603.771 0.243 0.291
Chain 1: 5500 -8321.448 0.243 0.291
Chain 1: 5600 -10877.374 0.235 0.290
Chain 1: 5700 -8419.895 0.208 0.290
Chain 1: 5800 -8844.634 0.210 0.290
Chain 1: 5900 -13567.742 0.216 0.291
Chain 1: 6000 -9876.937 0.224 0.292
Chain 1: 6100 -8344.566 0.230 0.292
Chain 1: 6200 -8291.041 0.230 0.292
Chain 1: 6300 -8437.158 0.200 0.235
Chain 1: 6400 -8487.112 0.154 0.184
Chain 1: 6500 -9223.706 0.159 0.184
Chain 1: 6600 -9863.187 0.142 0.080
Chain 1: 6700 -8334.106 0.131 0.080
Chain 1: 6800 -10780.936 0.149 0.183
Chain 1: 6900 -8946.959 0.135 0.183
Chain 1: 7000 -9321.587 0.101 0.080
Chain 1: 7100 -8278.427 0.096 0.080
Chain 1: 7200 -9078.158 0.104 0.088
Chain 1: 7300 -11749.875 0.125 0.126
Chain 1: 7400 -8097.000 0.169 0.183
Chain 1: 7500 -9744.426 0.178 0.183
Chain 1: 7600 -8585.154 0.185 0.183
Chain 1: 7700 -8148.597 0.172 0.169
Chain 1: 7800 -10610.598 0.173 0.169
Chain 1: 7900 -9943.728 0.159 0.135
Chain 1: 8000 -8294.915 0.175 0.169
Chain 1: 8100 -8250.244 0.163 0.169
Chain 1: 8200 -9080.387 0.163 0.169
Chain 1: 8300 -8129.810 0.152 0.135
Chain 1: 8400 -10507.617 0.130 0.135
Chain 1: 8500 -8376.846 0.138 0.135
Chain 1: 8600 -7942.592 0.130 0.117
Chain 1: 8700 -8240.422 0.128 0.117
Chain 1: 8800 -9504.883 0.118 0.117
Chain 1: 8900 -12324.344 0.135 0.133
Chain 1: 9000 -10974.346 0.127 0.123
Chain 1: 9100 -8165.032 0.161 0.133
Chain 1: 9200 -8238.316 0.153 0.133
Chain 1: 9300 -11798.380 0.171 0.226
Chain 1: 9400 -8100.009 0.194 0.229
Chain 1: 9500 -11565.785 0.199 0.229
Chain 1: 9600 -9664.006 0.213 0.229
Chain 1: 9700 -9349.024 0.213 0.229
Chain 1: 9800 -10903.941 0.214 0.229
Chain 1: 9900 -8413.276 0.220 0.296
Chain 1: 10000 -7936.063 0.214 0.296
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00138 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.8 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58186.026 1.000 1.000
Chain 1: 200 -17739.864 1.640 2.280
Chain 1: 300 -8688.551 1.441 1.042
Chain 1: 400 -8230.003 1.094 1.042
Chain 1: 500 -8454.047 0.881 1.000
Chain 1: 600 -7872.754 0.746 1.000
Chain 1: 700 -7836.018 0.640 0.074
Chain 1: 800 -8242.919 0.566 0.074
Chain 1: 900 -7953.292 0.508 0.056
Chain 1: 1000 -7819.997 0.459 0.056
Chain 1: 1100 -7652.949 0.361 0.049
Chain 1: 1200 -7583.242 0.134 0.036
Chain 1: 1300 -7746.301 0.032 0.027
Chain 1: 1400 -7815.311 0.027 0.022
Chain 1: 1500 -7609.532 0.027 0.022
Chain 1: 1600 -7863.363 0.023 0.022
Chain 1: 1700 -7514.555 0.027 0.027
Chain 1: 1800 -7620.698 0.023 0.022
Chain 1: 1900 -7485.744 0.022 0.021
Chain 1: 2000 -7595.409 0.021 0.021
Chain 1: 2100 -7595.120 0.019 0.018
Chain 1: 2200 -7720.623 0.020 0.018
Chain 1: 2300 -7575.046 0.020 0.018
Chain 1: 2400 -7641.467 0.020 0.018
Chain 1: 2500 -7574.611 0.018 0.016
Chain 1: 2600 -7515.775 0.015 0.014
Chain 1: 2700 -7514.562 0.011 0.014
Chain 1: 2800 -7577.003 0.010 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003111 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.11 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86627.323 1.000 1.000
Chain 1: 200 -13500.214 3.208 5.417
Chain 1: 300 -9801.662 2.265 1.000
Chain 1: 400 -10925.607 1.724 1.000
Chain 1: 500 -8791.964 1.428 0.377
Chain 1: 600 -8225.828 1.201 0.377
Chain 1: 700 -8386.446 1.033 0.243
Chain 1: 800 -8884.571 0.910 0.243
Chain 1: 900 -8560.228 0.814 0.103
Chain 1: 1000 -8680.149 0.734 0.103
Chain 1: 1100 -8386.411 0.637 0.069 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8182.502 0.098 0.056
Chain 1: 1300 -8482.131 0.064 0.038
Chain 1: 1400 -8432.792 0.054 0.035
Chain 1: 1500 -8324.146 0.031 0.035
Chain 1: 1600 -8430.442 0.025 0.025
Chain 1: 1700 -8505.099 0.024 0.025
Chain 1: 1800 -8072.711 0.024 0.025
Chain 1: 1900 -8176.918 0.022 0.014
Chain 1: 2000 -8152.206 0.020 0.013
Chain 1: 2100 -8129.189 0.017 0.013
Chain 1: 2200 -8094.798 0.015 0.013
Chain 1: 2300 -8224.508 0.013 0.013
Chain 1: 2400 -8079.500 0.014 0.013
Chain 1: 2500 -8148.445 0.014 0.013
Chain 1: 2600 -8067.601 0.014 0.010
Chain 1: 2700 -8096.885 0.013 0.010
Chain 1: 2800 -8057.922 0.008 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003336 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.36 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8395505.843 1.000 1.000
Chain 1: 200 -1582606.521 2.652 4.305
Chain 1: 300 -890140.466 2.028 1.000
Chain 1: 400 -457714.681 1.757 1.000
Chain 1: 500 -357991.996 1.461 0.945
Chain 1: 600 -233072.787 1.307 0.945
Chain 1: 700 -119256.395 1.257 0.945
Chain 1: 800 -86467.824 1.147 0.945
Chain 1: 900 -66810.062 1.052 0.778
Chain 1: 1000 -51610.389 0.976 0.778
Chain 1: 1100 -39084.987 0.908 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38265.609 0.480 0.379
Chain 1: 1300 -26212.399 0.448 0.379
Chain 1: 1400 -25932.655 0.355 0.320
Chain 1: 1500 -22517.482 0.342 0.320
Chain 1: 1600 -21733.672 0.292 0.295
Chain 1: 1700 -20605.581 0.202 0.294
Chain 1: 1800 -20549.645 0.165 0.152
Chain 1: 1900 -20876.157 0.137 0.055
Chain 1: 2000 -19385.770 0.115 0.055
Chain 1: 2100 -19624.262 0.084 0.036
Chain 1: 2200 -19851.134 0.083 0.036
Chain 1: 2300 -19467.882 0.039 0.020
Chain 1: 2400 -19239.828 0.039 0.020
Chain 1: 2500 -19041.934 0.025 0.016
Chain 1: 2600 -18671.739 0.024 0.016
Chain 1: 2700 -18628.562 0.018 0.012
Chain 1: 2800 -18345.324 0.020 0.015
Chain 1: 2900 -18626.772 0.020 0.015
Chain 1: 3000 -18612.883 0.012 0.012
Chain 1: 3100 -18697.947 0.011 0.012
Chain 1: 3200 -18388.394 0.012 0.015
Chain 1: 3300 -18593.302 0.011 0.012
Chain 1: 3400 -18067.868 0.013 0.015
Chain 1: 3500 -18680.300 0.015 0.015
Chain 1: 3600 -17986.238 0.017 0.015
Chain 1: 3700 -18373.646 0.019 0.017
Chain 1: 3800 -17332.191 0.023 0.021
Chain 1: 3900 -17328.305 0.022 0.021
Chain 1: 4000 -17445.615 0.022 0.021
Chain 1: 4100 -17359.335 0.022 0.021
Chain 1: 4200 -17175.293 0.022 0.021
Chain 1: 4300 -17313.879 0.021 0.021
Chain 1: 4400 -17270.503 0.019 0.011
Chain 1: 4500 -17172.994 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001372 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.72 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12601.475 1.000 1.000
Chain 1: 200 -9399.279 0.670 1.000
Chain 1: 300 -7977.298 0.506 0.341
Chain 1: 400 -8040.578 0.382 0.341
Chain 1: 500 -7948.680 0.308 0.178
Chain 1: 600 -7896.736 0.257 0.178
Chain 1: 700 -7778.082 0.223 0.015
Chain 1: 800 -7780.071 0.195 0.015
Chain 1: 900 -7732.270 0.174 0.012
Chain 1: 1000 -7871.483 0.158 0.015
Chain 1: 1100 -7890.614 0.059 0.012
Chain 1: 1200 -7821.011 0.025 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001384 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.84 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57019.998 1.000 1.000
Chain 1: 200 -17628.640 1.617 2.235
Chain 1: 300 -8849.169 1.409 1.000
Chain 1: 400 -8388.574 1.070 1.000
Chain 1: 500 -8501.071 0.859 0.992
Chain 1: 600 -8685.833 0.719 0.992
Chain 1: 700 -7814.328 0.633 0.112
Chain 1: 800 -8099.479 0.558 0.112
Chain 1: 900 -7722.087 0.501 0.055
Chain 1: 1000 -7710.006 0.451 0.055
Chain 1: 1100 -7805.717 0.353 0.049
Chain 1: 1200 -7819.316 0.129 0.035
Chain 1: 1300 -7669.799 0.032 0.021
Chain 1: 1400 -7809.555 0.028 0.019
Chain 1: 1500 -7614.969 0.030 0.021
Chain 1: 1600 -7591.961 0.028 0.019
Chain 1: 1700 -7612.013 0.017 0.018
Chain 1: 1800 -7639.311 0.014 0.012
Chain 1: 1900 -7638.686 0.009 0.004 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003151 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.51 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87167.118 1.000 1.000
Chain 1: 200 -13641.720 3.195 5.390
Chain 1: 300 -9895.971 2.256 1.000
Chain 1: 400 -11415.909 1.725 1.000
Chain 1: 500 -8751.273 1.441 0.379
Chain 1: 600 -9222.448 1.209 0.379
Chain 1: 700 -8687.788 1.046 0.304
Chain 1: 800 -8173.355 0.923 0.304
Chain 1: 900 -8291.745 0.822 0.133
Chain 1: 1000 -8249.610 0.740 0.133
Chain 1: 1100 -8671.245 0.645 0.063 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8237.753 0.111 0.062
Chain 1: 1300 -8441.303 0.076 0.053
Chain 1: 1400 -8551.241 0.064 0.051
Chain 1: 1500 -8394.606 0.035 0.049
Chain 1: 1600 -8498.622 0.031 0.024
Chain 1: 1700 -8562.724 0.026 0.019
Chain 1: 1800 -8120.847 0.025 0.019
Chain 1: 1900 -8227.465 0.025 0.019
Chain 1: 2000 -8211.809 0.025 0.019
Chain 1: 2100 -8337.542 0.021 0.015
Chain 1: 2200 -8126.230 0.019 0.015
Chain 1: 2300 -8221.060 0.017 0.013
Chain 1: 2400 -8288.262 0.017 0.013
Chain 1: 2500 -8236.176 0.016 0.012
Chain 1: 2600 -8248.610 0.015 0.012
Chain 1: 2700 -8156.960 0.015 0.012
Chain 1: 2800 -8105.170 0.010 0.011
Chain 1: 2900 -8199.201 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003252 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.52 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8420824.065 1.000 1.000
Chain 1: 200 -1592949.336 2.643 4.286
Chain 1: 300 -892899.847 2.023 1.000
Chain 1: 400 -458433.873 1.755 1.000
Chain 1: 500 -357801.502 1.460 0.948
Chain 1: 600 -232624.922 1.306 0.948
Chain 1: 700 -119076.481 1.256 0.948
Chain 1: 800 -86328.672 1.146 0.948
Chain 1: 900 -66739.721 1.052 0.784
Chain 1: 1000 -51598.284 0.976 0.784
Chain 1: 1100 -39126.257 0.908 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38316.229 0.481 0.379
Chain 1: 1300 -26320.351 0.448 0.379
Chain 1: 1400 -26046.884 0.355 0.319
Chain 1: 1500 -22644.791 0.341 0.319
Chain 1: 1600 -21865.068 0.291 0.294
Chain 1: 1700 -20744.208 0.201 0.293
Chain 1: 1800 -20689.990 0.164 0.150
Chain 1: 1900 -21016.600 0.136 0.054
Chain 1: 2000 -19529.369 0.114 0.054
Chain 1: 2100 -19767.993 0.083 0.036
Chain 1: 2200 -19994.101 0.082 0.036
Chain 1: 2300 -19611.435 0.039 0.020
Chain 1: 2400 -19383.383 0.039 0.020
Chain 1: 2500 -19185.014 0.025 0.016
Chain 1: 2600 -18815.099 0.023 0.016
Chain 1: 2700 -18772.059 0.018 0.012
Chain 1: 2800 -18488.451 0.019 0.015
Chain 1: 2900 -18769.903 0.019 0.015
Chain 1: 3000 -18756.222 0.012 0.012
Chain 1: 3100 -18841.224 0.011 0.012
Chain 1: 3200 -18531.669 0.012 0.015
Chain 1: 3300 -18736.594 0.011 0.012
Chain 1: 3400 -18210.895 0.012 0.015
Chain 1: 3500 -18823.549 0.015 0.015
Chain 1: 3600 -18129.237 0.017 0.015
Chain 1: 3700 -18516.668 0.018 0.017
Chain 1: 3800 -17474.726 0.023 0.021
Chain 1: 3900 -17470.754 0.021 0.021
Chain 1: 4000 -17588.147 0.022 0.021
Chain 1: 4100 -17501.734 0.022 0.021
Chain 1: 4200 -17317.654 0.021 0.021
Chain 1: 4300 -17456.347 0.021 0.021
Chain 1: 4400 -17412.883 0.018 0.011
Chain 1: 4500 -17315.292 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001242 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.42 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12485.678 1.000 1.000
Chain 1: 200 -9403.574 0.664 1.000
Chain 1: 300 -8287.306 0.487 0.328
Chain 1: 400 -8270.093 0.366 0.328
Chain 1: 500 -8177.584 0.295 0.135
Chain 1: 600 -8083.999 0.248 0.135
Chain 1: 700 -7993.530 0.214 0.012
Chain 1: 800 -8035.968 0.188 0.012
Chain 1: 900 -8159.269 0.169 0.012
Chain 1: 1000 -8068.808 0.153 0.012
Chain 1: 1100 -8099.895 0.053 0.011
Chain 1: 1200 -8004.522 0.022 0.011
Chain 1: 1300 -7963.333 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.0014 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58058.655 1.000 1.000
Chain 1: 200 -17711.065 1.639 2.278
Chain 1: 300 -8701.010 1.438 1.036
Chain 1: 400 -8123.871 1.096 1.036
Chain 1: 500 -8242.318 0.880 1.000
Chain 1: 600 -8112.612 0.736 1.000
Chain 1: 700 -7936.045 0.634 0.071
Chain 1: 800 -8300.159 0.560 0.071
Chain 1: 900 -7937.353 0.503 0.046
Chain 1: 1000 -7994.974 0.453 0.046
Chain 1: 1100 -7701.818 0.357 0.044
Chain 1: 1200 -7757.665 0.130 0.038
Chain 1: 1300 -7772.480 0.027 0.022
Chain 1: 1400 -7785.120 0.020 0.016
Chain 1: 1500 -7547.076 0.022 0.022
Chain 1: 1600 -7726.393 0.022 0.023
Chain 1: 1700 -7588.652 0.022 0.023
Chain 1: 1800 -7577.560 0.018 0.018
Chain 1: 1900 -7597.098 0.013 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00256 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.6 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86356.244 1.000 1.000
Chain 1: 200 -13549.041 3.187 5.374
Chain 1: 300 -9934.222 2.246 1.000
Chain 1: 400 -10753.237 1.703 1.000
Chain 1: 500 -8896.750 1.404 0.364
Chain 1: 600 -8580.690 1.177 0.364
Chain 1: 700 -8599.506 1.009 0.209
Chain 1: 800 -9081.943 0.889 0.209
Chain 1: 900 -8765.557 0.795 0.076
Chain 1: 1000 -8528.453 0.718 0.076
Chain 1: 1100 -8777.712 0.621 0.053 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8423.244 0.088 0.042
Chain 1: 1300 -8629.071 0.054 0.037
Chain 1: 1400 -8637.435 0.046 0.036
Chain 1: 1500 -8527.224 0.026 0.028
Chain 1: 1600 -8630.773 0.024 0.028
Chain 1: 1700 -8719.554 0.025 0.028
Chain 1: 1800 -8312.813 0.024 0.028
Chain 1: 1900 -8410.038 0.022 0.024
Chain 1: 2000 -8381.991 0.019 0.013
Chain 1: 2100 -8502.267 0.018 0.013
Chain 1: 2200 -8304.722 0.016 0.013
Chain 1: 2300 -8448.361 0.015 0.013
Chain 1: 2400 -8455.256 0.015 0.013
Chain 1: 2500 -8423.448 0.015 0.012
Chain 1: 2600 -8421.833 0.013 0.012
Chain 1: 2700 -8335.018 0.013 0.012
Chain 1: 2800 -8300.638 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004313 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 43.13 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8407431.957 1.000 1.000
Chain 1: 200 -1586882.053 2.649 4.298
Chain 1: 300 -891879.171 2.026 1.000
Chain 1: 400 -457890.422 1.756 1.000
Chain 1: 500 -357848.672 1.461 0.948
Chain 1: 600 -232829.799 1.307 0.948
Chain 1: 700 -119171.184 1.256 0.948
Chain 1: 800 -86381.612 1.147 0.948
Chain 1: 900 -66754.153 1.052 0.779
Chain 1: 1000 -51570.187 0.976 0.779
Chain 1: 1100 -39061.661 0.908 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38242.557 0.481 0.380
Chain 1: 1300 -26218.851 0.449 0.380
Chain 1: 1400 -25939.741 0.355 0.320
Chain 1: 1500 -22531.264 0.342 0.320
Chain 1: 1600 -21748.906 0.292 0.294
Chain 1: 1700 -20625.263 0.202 0.294
Chain 1: 1800 -20570.063 0.164 0.151
Chain 1: 1900 -20896.000 0.137 0.054
Chain 1: 2000 -19408.804 0.115 0.054
Chain 1: 2100 -19647.198 0.084 0.036
Chain 1: 2200 -19873.157 0.083 0.036
Chain 1: 2300 -19490.857 0.039 0.020
Chain 1: 2400 -19263.055 0.039 0.020
Chain 1: 2500 -19064.885 0.025 0.016
Chain 1: 2600 -18695.425 0.023 0.016
Chain 1: 2700 -18652.554 0.018 0.012
Chain 1: 2800 -18369.336 0.020 0.015
Chain 1: 2900 -18650.535 0.019 0.015
Chain 1: 3000 -18636.836 0.012 0.012
Chain 1: 3100 -18721.735 0.011 0.012
Chain 1: 3200 -18412.575 0.012 0.015
Chain 1: 3300 -18617.213 0.011 0.012
Chain 1: 3400 -18092.319 0.013 0.015
Chain 1: 3500 -18703.791 0.015 0.015
Chain 1: 3600 -18011.090 0.017 0.015
Chain 1: 3700 -18397.370 0.018 0.017
Chain 1: 3800 -17357.909 0.023 0.021
Chain 1: 3900 -17354.066 0.021 0.021
Chain 1: 4000 -17471.412 0.022 0.021
Chain 1: 4100 -17385.119 0.022 0.021
Chain 1: 4200 -17201.634 0.021 0.021
Chain 1: 4300 -17339.879 0.021 0.021
Chain 1: 4400 -17296.863 0.019 0.011
Chain 1: 4500 -17199.421 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001278 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.78 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12137.487 1.000 1.000
Chain 1: 200 -8933.798 0.679 1.000
Chain 1: 300 -7901.168 0.496 0.359
Chain 1: 400 -8073.515 0.378 0.359
Chain 1: 500 -7929.909 0.306 0.131
Chain 1: 600 -7800.655 0.258 0.131
Chain 1: 700 -7728.783 0.222 0.021
Chain 1: 800 -7735.958 0.194 0.021
Chain 1: 900 -7650.391 0.174 0.018
Chain 1: 1000 -7828.810 0.159 0.021
Chain 1: 1100 -7840.953 0.059 0.018
Chain 1: 1200 -7744.763 0.024 0.017
Chain 1: 1300 -7707.388 0.012 0.012
Chain 1: 1400 -7725.509 0.010 0.011
Chain 1: 1500 -7817.216 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001411 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.11 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61496.763 1.000 1.000
Chain 1: 200 -17597.203 1.747 2.495
Chain 1: 300 -8743.454 1.502 1.013
Chain 1: 400 -9006.460 1.134 1.013
Chain 1: 500 -8330.921 0.924 1.000
Chain 1: 600 -8586.876 0.775 1.000
Chain 1: 700 -8010.840 0.674 0.081
Chain 1: 800 -8078.632 0.591 0.081
Chain 1: 900 -7891.486 0.528 0.072
Chain 1: 1000 -7766.851 0.477 0.072
Chain 1: 1100 -7680.551 0.378 0.030
Chain 1: 1200 -7609.815 0.129 0.029
Chain 1: 1300 -7679.370 0.029 0.024
Chain 1: 1400 -7687.346 0.026 0.016
Chain 1: 1500 -7636.677 0.019 0.011
Chain 1: 1600 -7604.505 0.016 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002529 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.29 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85528.494 1.000 1.000
Chain 1: 200 -13234.154 3.231 5.463
Chain 1: 300 -9658.865 2.278 1.000
Chain 1: 400 -10608.735 1.731 1.000
Chain 1: 500 -8580.758 1.432 0.370
Chain 1: 600 -8331.136 1.198 0.370
Chain 1: 700 -8285.836 1.028 0.236
Chain 1: 800 -8762.356 0.906 0.236
Chain 1: 900 -8534.722 0.808 0.090
Chain 1: 1000 -8246.078 0.731 0.090
Chain 1: 1100 -8522.964 0.634 0.054 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8292.389 0.091 0.035
Chain 1: 1300 -8352.796 0.054 0.032
Chain 1: 1400 -8375.292 0.046 0.030
Chain 1: 1500 -8246.571 0.024 0.028
Chain 1: 1600 -8356.630 0.022 0.027
Chain 1: 1700 -8444.618 0.023 0.027
Chain 1: 1800 -8042.719 0.022 0.027
Chain 1: 1900 -8142.732 0.021 0.016
Chain 1: 2000 -8113.939 0.018 0.013
Chain 1: 2100 -8233.905 0.016 0.013
Chain 1: 2200 -8023.434 0.016 0.013
Chain 1: 2300 -8174.140 0.017 0.015
Chain 1: 2400 -8055.504 0.018 0.015
Chain 1: 2500 -8118.312 0.017 0.015
Chain 1: 2600 -8139.635 0.016 0.015
Chain 1: 2700 -8058.731 0.016 0.015
Chain 1: 2800 -8032.748 0.011 0.012
Chain 1: 2900 -8088.132 0.011 0.010
Chain 1: 3000 -7972.356 0.012 0.015
Chain 1: 3100 -8110.110 0.012 0.015
Chain 1: 3200 -7990.084 0.011 0.015
Chain 1: 3300 -8011.560 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003211 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.11 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8401794.730 1.000 1.000
Chain 1: 200 -1583453.675 2.653 4.306
Chain 1: 300 -889520.569 2.029 1.000
Chain 1: 400 -456705.642 1.758 1.000
Chain 1: 500 -357012.000 1.463 0.948
Chain 1: 600 -232265.687 1.308 0.948
Chain 1: 700 -118733.925 1.258 0.948
Chain 1: 800 -86024.081 1.148 0.948
Chain 1: 900 -66410.445 1.054 0.780
Chain 1: 1000 -51238.433 0.978 0.780
Chain 1: 1100 -38744.929 0.910 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37924.274 0.482 0.380
Chain 1: 1300 -25908.174 0.450 0.380
Chain 1: 1400 -25630.194 0.356 0.322
Chain 1: 1500 -22224.562 0.344 0.322
Chain 1: 1600 -21443.284 0.294 0.296
Chain 1: 1700 -20320.171 0.204 0.295
Chain 1: 1800 -20265.200 0.166 0.153
Chain 1: 1900 -20591.060 0.138 0.055
Chain 1: 2000 -19104.626 0.116 0.055
Chain 1: 2100 -19342.812 0.085 0.036
Chain 1: 2200 -19568.790 0.084 0.036
Chain 1: 2300 -19186.521 0.040 0.020
Chain 1: 2400 -18958.713 0.040 0.020
Chain 1: 2500 -18760.744 0.025 0.016
Chain 1: 2600 -18391.201 0.024 0.016
Chain 1: 2700 -18348.379 0.019 0.012
Chain 1: 2800 -18065.278 0.020 0.016
Chain 1: 2900 -18346.421 0.020 0.015
Chain 1: 3000 -18332.641 0.012 0.012
Chain 1: 3100 -18417.527 0.011 0.012
Chain 1: 3200 -18108.456 0.012 0.015
Chain 1: 3300 -18313.052 0.011 0.012
Chain 1: 3400 -17788.318 0.013 0.015
Chain 1: 3500 -18399.622 0.015 0.016
Chain 1: 3600 -17707.109 0.017 0.016
Chain 1: 3700 -18093.221 0.019 0.017
Chain 1: 3800 -17054.155 0.023 0.021
Chain 1: 3900 -17050.348 0.022 0.021
Chain 1: 4000 -17167.644 0.022 0.021
Chain 1: 4100 -17081.384 0.022 0.021
Chain 1: 4200 -16897.992 0.022 0.021
Chain 1: 4300 -17036.145 0.022 0.021
Chain 1: 4400 -16993.164 0.019 0.011
Chain 1: 4500 -16895.757 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001122 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 11.22 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49122.178 1.000 1.000
Chain 1: 200 -21180.989 1.160 1.319
Chain 1: 300 -15988.985 0.881 1.000
Chain 1: 400 -21441.759 0.725 1.000
Chain 1: 500 -12953.274 0.711 0.655
Chain 1: 600 -15145.841 0.616 0.655
Chain 1: 700 -11837.706 0.568 0.325
Chain 1: 800 -14356.658 0.519 0.325
Chain 1: 900 -11314.103 0.491 0.279
Chain 1: 1000 -13637.256 0.459 0.279
Chain 1: 1100 -11664.897 0.376 0.269
Chain 1: 1200 -16933.553 0.275 0.269
Chain 1: 1300 -12391.188 0.280 0.269
Chain 1: 1400 -11580.307 0.261 0.269
Chain 1: 1500 -10508.485 0.206 0.175
Chain 1: 1600 -9984.276 0.197 0.175
Chain 1: 1700 -12736.885 0.190 0.175
Chain 1: 1800 -10351.847 0.196 0.216
Chain 1: 1900 -11823.132 0.181 0.170
Chain 1: 2000 -11467.454 0.167 0.169
Chain 1: 2100 -10921.250 0.155 0.124
Chain 1: 2200 -10127.637 0.132 0.102
Chain 1: 2300 -9980.483 0.097 0.078
Chain 1: 2400 -9852.593 0.091 0.078
Chain 1: 2500 -14389.017 0.113 0.078
Chain 1: 2600 -18572.925 0.130 0.124
Chain 1: 2700 -9607.339 0.202 0.124
Chain 1: 2800 -9285.702 0.182 0.078
Chain 1: 2900 -9921.314 0.176 0.064
Chain 1: 3000 -16302.050 0.212 0.078
Chain 1: 3100 -9104.393 0.286 0.225
Chain 1: 3200 -16573.066 0.323 0.315
Chain 1: 3300 -15646.338 0.328 0.315
Chain 1: 3400 -10731.886 0.372 0.391
Chain 1: 3500 -10589.931 0.342 0.391
Chain 1: 3600 -10339.128 0.322 0.391
Chain 1: 3700 -20333.139 0.278 0.391
Chain 1: 3800 -10342.931 0.371 0.451
Chain 1: 3900 -11997.954 0.378 0.451
Chain 1: 4000 -9824.060 0.361 0.451
Chain 1: 4100 -9325.542 0.288 0.221
Chain 1: 4200 -13033.520 0.271 0.221
Chain 1: 4300 -10604.327 0.288 0.229
Chain 1: 4400 -8999.497 0.260 0.221
Chain 1: 4500 -9670.832 0.266 0.221
Chain 1: 4600 -13928.274 0.294 0.229
Chain 1: 4700 -10139.996 0.282 0.229
Chain 1: 4800 -9339.222 0.194 0.221
Chain 1: 4900 -9416.711 0.181 0.221
Chain 1: 5000 -14286.566 0.193 0.229
Chain 1: 5100 -9318.024 0.241 0.284
Chain 1: 5200 -16335.621 0.255 0.306
Chain 1: 5300 -12966.267 0.258 0.306
Chain 1: 5400 -9004.740 0.285 0.341
Chain 1: 5500 -11550.913 0.300 0.341
Chain 1: 5600 -9727.585 0.288 0.341
Chain 1: 5700 -8959.974 0.259 0.260
Chain 1: 5800 -9075.455 0.252 0.260
Chain 1: 5900 -10596.980 0.265 0.260
Chain 1: 6000 -12400.272 0.246 0.220
Chain 1: 6100 -8642.065 0.236 0.220
Chain 1: 6200 -11671.990 0.219 0.220
Chain 1: 6300 -11247.845 0.197 0.187
Chain 1: 6400 -12644.767 0.164 0.145
Chain 1: 6500 -10376.563 0.164 0.145
Chain 1: 6600 -9112.444 0.159 0.144
Chain 1: 6700 -9222.325 0.151 0.144
Chain 1: 6800 -9332.540 0.151 0.144
Chain 1: 6900 -13234.849 0.166 0.145
Chain 1: 7000 -8569.511 0.206 0.219
Chain 1: 7100 -8698.504 0.164 0.139
Chain 1: 7200 -8993.305 0.142 0.110
Chain 1: 7300 -8386.128 0.145 0.110
Chain 1: 7400 -14873.159 0.178 0.139
Chain 1: 7500 -8513.380 0.230 0.139
Chain 1: 7600 -11403.459 0.242 0.253
Chain 1: 7700 -8620.846 0.273 0.295
Chain 1: 7800 -9244.005 0.279 0.295
Chain 1: 7900 -8746.939 0.255 0.253
Chain 1: 8000 -9954.105 0.212 0.121
Chain 1: 8100 -8836.226 0.224 0.127
Chain 1: 8200 -10549.463 0.237 0.162
Chain 1: 8300 -8986.728 0.247 0.174
Chain 1: 8400 -12255.666 0.230 0.174
Chain 1: 8500 -8742.118 0.195 0.174
Chain 1: 8600 -9448.551 0.177 0.162
Chain 1: 8700 -9218.396 0.148 0.127
Chain 1: 8800 -9856.339 0.147 0.127
Chain 1: 8900 -9259.708 0.148 0.127
Chain 1: 9000 -11591.705 0.156 0.162
Chain 1: 9100 -12002.381 0.147 0.162
Chain 1: 9200 -9369.343 0.159 0.174
Chain 1: 9300 -8496.423 0.152 0.103
Chain 1: 9400 -8731.974 0.128 0.075
Chain 1: 9500 -9417.928 0.095 0.073
Chain 1: 9600 -8513.395 0.098 0.073
Chain 1: 9700 -8269.335 0.098 0.073
Chain 1: 9800 -12111.580 0.124 0.103
Chain 1: 9900 -10342.352 0.134 0.106
Chain 1: 10000 -8389.755 0.137 0.106
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001378 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.78 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -62095.939 1.000 1.000
Chain 1: 200 -18256.626 1.701 2.401
Chain 1: 300 -9034.299 1.474 1.021
Chain 1: 400 -9467.281 1.117 1.021
Chain 1: 500 -7855.344 0.935 1.000
Chain 1: 600 -8800.634 0.797 1.000
Chain 1: 700 -8354.876 0.691 0.205
Chain 1: 800 -8227.165 0.606 0.205
Chain 1: 900 -7942.176 0.543 0.107
Chain 1: 1000 -7723.219 0.491 0.107
Chain 1: 1100 -7899.724 0.394 0.053
Chain 1: 1200 -7565.104 0.158 0.046
Chain 1: 1300 -7796.766 0.059 0.044
Chain 1: 1400 -7895.190 0.055 0.036
Chain 1: 1500 -7524.401 0.040 0.036
Chain 1: 1600 -7615.967 0.030 0.030
Chain 1: 1700 -7601.293 0.025 0.028
Chain 1: 1800 -7642.363 0.024 0.028
Chain 1: 1900 -7535.439 0.022 0.022
Chain 1: 2000 -7647.154 0.021 0.015
Chain 1: 2100 -7649.363 0.018 0.014
Chain 1: 2200 -7746.059 0.015 0.012
Chain 1: 2300 -7542.162 0.015 0.012
Chain 1: 2400 -7574.636 0.014 0.012
Chain 1: 2500 -7415.056 0.011 0.012
Chain 1: 2600 -7530.091 0.012 0.014
Chain 1: 2700 -7510.710 0.012 0.014
Chain 1: 2800 -7507.509 0.011 0.014
Chain 1: 2900 -7372.802 0.012 0.015
Chain 1: 3000 -7522.790 0.012 0.015
Chain 1: 3100 -7520.920 0.012 0.015
Chain 1: 3200 -7740.906 0.014 0.018
Chain 1: 3300 -7445.953 0.015 0.018
Chain 1: 3400 -7688.234 0.018 0.020
Chain 1: 3500 -7432.266 0.019 0.020
Chain 1: 3600 -7499.616 0.018 0.020
Chain 1: 3700 -7447.553 0.019 0.020
Chain 1: 3800 -7448.488 0.019 0.020
Chain 1: 3900 -7407.961 0.018 0.020
Chain 1: 4000 -7399.976 0.016 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003069 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.69 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86233.335 1.000 1.000
Chain 1: 200 -13892.578 3.104 5.207
Chain 1: 300 -10170.836 2.191 1.000
Chain 1: 400 -11567.115 1.673 1.000
Chain 1: 500 -9087.939 1.393 0.366
Chain 1: 600 -8938.252 1.164 0.366
Chain 1: 700 -8917.982 0.998 0.273
Chain 1: 800 -8679.998 0.877 0.273
Chain 1: 900 -8586.497 0.780 0.121
Chain 1: 1000 -8813.013 0.705 0.121
Chain 1: 1100 -8825.319 0.605 0.027 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8508.372 0.088 0.027
Chain 1: 1300 -8840.700 0.055 0.027
Chain 1: 1400 -8796.096 0.044 0.026
Chain 1: 1500 -8685.765 0.018 0.017
Chain 1: 1600 -8788.104 0.017 0.013
Chain 1: 1700 -8850.873 0.018 0.013
Chain 1: 1800 -8412.792 0.020 0.013
Chain 1: 1900 -8517.592 0.020 0.013
Chain 1: 2000 -8496.130 0.018 0.012
Chain 1: 2100 -8471.586 0.018 0.012
Chain 1: 2200 -8436.228 0.015 0.012
Chain 1: 2300 -8571.289 0.013 0.012
Chain 1: 2400 -8416.033 0.014 0.012
Chain 1: 2500 -8487.257 0.014 0.012
Chain 1: 2600 -8400.120 0.013 0.010
Chain 1: 2700 -8437.373 0.013 0.010
Chain 1: 2800 -8395.400 0.008 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003746 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 37.46 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8369948.180 1.000 1.000
Chain 1: 200 -1576207.096 2.655 4.310
Chain 1: 300 -889208.858 2.028 1.000
Chain 1: 400 -456947.327 1.757 1.000
Chain 1: 500 -358011.498 1.461 0.946
Chain 1: 600 -233435.058 1.306 0.946
Chain 1: 700 -119742.416 1.255 0.946
Chain 1: 800 -86953.667 1.146 0.946
Chain 1: 900 -67293.943 1.051 0.773
Chain 1: 1000 -52081.770 0.975 0.773
Chain 1: 1100 -39538.624 0.907 0.534 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38720.687 0.478 0.377
Chain 1: 1300 -26639.507 0.446 0.377
Chain 1: 1400 -26358.952 0.352 0.317
Chain 1: 1500 -22935.036 0.340 0.317
Chain 1: 1600 -22149.609 0.290 0.292
Chain 1: 1700 -21017.672 0.200 0.292
Chain 1: 1800 -20961.218 0.163 0.149
Chain 1: 1900 -21287.831 0.135 0.054
Chain 1: 2000 -19795.236 0.113 0.054
Chain 1: 2100 -20033.916 0.083 0.035
Chain 1: 2200 -20261.097 0.082 0.035
Chain 1: 2300 -19877.526 0.039 0.019
Chain 1: 2400 -19649.369 0.039 0.019
Chain 1: 2500 -19451.578 0.025 0.015
Chain 1: 2600 -19081.109 0.023 0.015
Chain 1: 2700 -19037.962 0.018 0.012
Chain 1: 2800 -18754.654 0.019 0.015
Chain 1: 2900 -19036.180 0.019 0.015
Chain 1: 3000 -19022.303 0.012 0.012
Chain 1: 3100 -19107.368 0.011 0.012
Chain 1: 3200 -18797.705 0.011 0.015
Chain 1: 3300 -19002.737 0.011 0.012
Chain 1: 3400 -18477.109 0.012 0.015
Chain 1: 3500 -19089.887 0.014 0.015
Chain 1: 3600 -18395.405 0.016 0.015
Chain 1: 3700 -18783.089 0.018 0.016
Chain 1: 3800 -17741.059 0.022 0.021
Chain 1: 3900 -17737.198 0.021 0.021
Chain 1: 4000 -17854.455 0.022 0.021
Chain 1: 4100 -17768.102 0.022 0.021
Chain 1: 4200 -17584.042 0.021 0.021
Chain 1: 4300 -17722.664 0.021 0.021
Chain 1: 4400 -17679.177 0.018 0.010
Chain 1: 4500 -17581.668 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001352 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.52 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48898.386 1.000 1.000
Chain 1: 200 -16205.276 1.509 2.017
Chain 1: 300 -23782.375 1.112 1.000
Chain 1: 400 -18672.312 0.902 1.000
Chain 1: 500 -15509.945 0.763 0.319
Chain 1: 600 -13013.885 0.668 0.319
Chain 1: 700 -12724.159 0.575 0.274
Chain 1: 800 -16244.158 0.531 0.274
Chain 1: 900 -14254.388 0.487 0.217
Chain 1: 1000 -12312.819 0.454 0.217
Chain 1: 1100 -11384.227 0.362 0.204
Chain 1: 1200 -11707.974 0.163 0.192
Chain 1: 1300 -11448.498 0.134 0.158
Chain 1: 1400 -12095.898 0.112 0.140
Chain 1: 1500 -9981.077 0.113 0.140
Chain 1: 1600 -10586.389 0.099 0.082
Chain 1: 1700 -9505.137 0.108 0.114
Chain 1: 1800 -10376.091 0.095 0.084
Chain 1: 1900 -9833.274 0.087 0.082
Chain 1: 2000 -11215.781 0.083 0.082
Chain 1: 2100 -9517.208 0.093 0.084
Chain 1: 2200 -9808.391 0.093 0.084
Chain 1: 2300 -11765.214 0.107 0.114
Chain 1: 2400 -9956.501 0.120 0.123
Chain 1: 2500 -10914.048 0.108 0.114
Chain 1: 2600 -10096.932 0.110 0.114
Chain 1: 2700 -12458.240 0.118 0.123
Chain 1: 2800 -10909.327 0.123 0.142
Chain 1: 2900 -9495.326 0.133 0.149
Chain 1: 3000 -12184.375 0.143 0.166
Chain 1: 3100 -11673.021 0.129 0.149
Chain 1: 3200 -17223.019 0.158 0.166
Chain 1: 3300 -16353.397 0.147 0.149
Chain 1: 3400 -8908.865 0.212 0.149
Chain 1: 3500 -10001.355 0.215 0.149
Chain 1: 3600 -10828.287 0.214 0.149
Chain 1: 3700 -14117.084 0.219 0.149
Chain 1: 3800 -14587.959 0.208 0.149
Chain 1: 3900 -14052.613 0.196 0.109
Chain 1: 4000 -12039.800 0.191 0.109
Chain 1: 4100 -9530.867 0.213 0.167
Chain 1: 4200 -13137.088 0.208 0.167
Chain 1: 4300 -13801.799 0.208 0.167
Chain 1: 4400 -10348.869 0.158 0.167
Chain 1: 4500 -8591.590 0.167 0.205
Chain 1: 4600 -9228.115 0.166 0.205
Chain 1: 4700 -9741.133 0.148 0.167
Chain 1: 4800 -8936.320 0.154 0.167
Chain 1: 4900 -8988.365 0.151 0.167
Chain 1: 5000 -9324.892 0.138 0.090
Chain 1: 5100 -8679.102 0.119 0.074
Chain 1: 5200 -8782.169 0.093 0.069
Chain 1: 5300 -14495.277 0.127 0.074
Chain 1: 5400 -12066.147 0.114 0.074
Chain 1: 5500 -8469.511 0.136 0.074
Chain 1: 5600 -10006.531 0.144 0.090
Chain 1: 5700 -12083.759 0.156 0.154
Chain 1: 5800 -10420.638 0.163 0.160
Chain 1: 5900 -9130.203 0.177 0.160
Chain 1: 6000 -9941.445 0.181 0.160
Chain 1: 6100 -8853.307 0.186 0.160
Chain 1: 6200 -8524.886 0.189 0.160
Chain 1: 6300 -9027.227 0.155 0.154
Chain 1: 6400 -9217.304 0.137 0.141
Chain 1: 6500 -10380.009 0.106 0.123
Chain 1: 6600 -10397.251 0.091 0.112
Chain 1: 6700 -10929.443 0.078 0.082
Chain 1: 6800 -8478.688 0.091 0.082
Chain 1: 6900 -8515.327 0.078 0.056
Chain 1: 7000 -9061.435 0.075 0.056
Chain 1: 7100 -8621.711 0.068 0.051
Chain 1: 7200 -8712.804 0.065 0.051
Chain 1: 7300 -9095.321 0.064 0.049
Chain 1: 7400 -11271.479 0.081 0.051
Chain 1: 7500 -9024.811 0.095 0.051
Chain 1: 7600 -8465.917 0.101 0.060
Chain 1: 7700 -8335.490 0.098 0.060
Chain 1: 7800 -10726.176 0.091 0.060
Chain 1: 7900 -8879.659 0.112 0.066
Chain 1: 8000 -10594.894 0.122 0.162
Chain 1: 8100 -8826.081 0.137 0.193
Chain 1: 8200 -8500.054 0.140 0.193
Chain 1: 8300 -10858.058 0.157 0.200
Chain 1: 8400 -8952.657 0.159 0.208
Chain 1: 8500 -9246.624 0.137 0.200
Chain 1: 8600 -8972.069 0.134 0.200
Chain 1: 8700 -8525.042 0.138 0.200
Chain 1: 8800 -8574.084 0.116 0.162
Chain 1: 8900 -11736.019 0.122 0.162
Chain 1: 9000 -8811.522 0.139 0.200
Chain 1: 9100 -9358.859 0.125 0.058
Chain 1: 9200 -8577.453 0.130 0.091
Chain 1: 9300 -10330.290 0.125 0.091
Chain 1: 9400 -9387.206 0.114 0.091
Chain 1: 9500 -8558.723 0.121 0.097
Chain 1: 9600 -10185.899 0.134 0.100
Chain 1: 9700 -11762.449 0.142 0.134
Chain 1: 9800 -8402.952 0.181 0.160
Chain 1: 9900 -8466.683 0.155 0.134
Chain 1: 10000 -8748.917 0.125 0.100
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001518 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.18 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57092.688 1.000 1.000
Chain 1: 200 -17483.510 1.633 2.266
Chain 1: 300 -8759.861 1.420 1.000
Chain 1: 400 -8387.079 1.076 1.000
Chain 1: 500 -7770.150 0.877 0.996
Chain 1: 600 -8876.222 0.752 0.996
Chain 1: 700 -8043.885 0.659 0.125
Chain 1: 800 -8164.993 0.579 0.125
Chain 1: 900 -8101.302 0.515 0.103
Chain 1: 1000 -7723.254 0.468 0.103
Chain 1: 1100 -7690.297 0.369 0.079
Chain 1: 1200 -7785.640 0.144 0.049
Chain 1: 1300 -7771.732 0.044 0.044
Chain 1: 1400 -7672.888 0.041 0.015
Chain 1: 1500 -7599.167 0.034 0.013
Chain 1: 1600 -7595.202 0.022 0.012
Chain 1: 1700 -7549.948 0.012 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002671 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 26.71 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87201.083 1.000 1.000
Chain 1: 200 -13526.469 3.223 5.447
Chain 1: 300 -9900.752 2.271 1.000
Chain 1: 400 -10803.043 1.724 1.000
Chain 1: 500 -8816.482 1.424 0.366
Chain 1: 600 -8452.338 1.194 0.366
Chain 1: 700 -8636.247 1.027 0.225
Chain 1: 800 -9392.752 0.908 0.225
Chain 1: 900 -8739.153 0.816 0.084
Chain 1: 1000 -8483.921 0.737 0.084
Chain 1: 1100 -8683.507 0.639 0.081 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8255.906 0.100 0.075
Chain 1: 1300 -8584.156 0.067 0.052
Chain 1: 1400 -8586.125 0.059 0.043
Chain 1: 1500 -8461.632 0.038 0.038
Chain 1: 1600 -8579.387 0.035 0.030
Chain 1: 1700 -8658.294 0.034 0.030
Chain 1: 1800 -8247.035 0.031 0.030
Chain 1: 1900 -8342.601 0.024 0.023
Chain 1: 2000 -8315.908 0.022 0.015
Chain 1: 2100 -8438.297 0.021 0.015
Chain 1: 2200 -8257.758 0.018 0.015
Chain 1: 2300 -8337.982 0.015 0.014
Chain 1: 2400 -8407.520 0.016 0.014
Chain 1: 2500 -8352.823 0.015 0.011
Chain 1: 2600 -8352.053 0.013 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002543 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.43 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8413494.787 1.000 1.000
Chain 1: 200 -1586425.086 2.652 4.303
Chain 1: 300 -890178.648 2.029 1.000
Chain 1: 400 -456656.295 1.759 1.000
Chain 1: 500 -356959.110 1.463 0.949
Chain 1: 600 -232127.312 1.309 0.949
Chain 1: 700 -118801.276 1.258 0.949
Chain 1: 800 -86139.030 1.148 0.949
Chain 1: 900 -66570.648 1.053 0.782
Chain 1: 1000 -51437.045 0.977 0.782
Chain 1: 1100 -38977.076 0.909 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38161.174 0.481 0.379
Chain 1: 1300 -26178.470 0.449 0.379
Chain 1: 1400 -25903.786 0.355 0.320
Chain 1: 1500 -22506.392 0.342 0.320
Chain 1: 1600 -21727.659 0.292 0.294
Chain 1: 1700 -20608.373 0.202 0.294
Chain 1: 1800 -20554.211 0.164 0.151
Chain 1: 1900 -20880.360 0.136 0.054
Chain 1: 2000 -19395.109 0.115 0.054
Chain 1: 2100 -19633.460 0.084 0.036
Chain 1: 2200 -19859.262 0.083 0.036
Chain 1: 2300 -19476.992 0.039 0.020
Chain 1: 2400 -19249.112 0.039 0.020
Chain 1: 2500 -19050.898 0.025 0.016
Chain 1: 2600 -18681.485 0.023 0.016
Chain 1: 2700 -18638.530 0.018 0.012
Chain 1: 2800 -18355.317 0.020 0.015
Chain 1: 2900 -18636.425 0.019 0.015
Chain 1: 3000 -18622.705 0.012 0.012
Chain 1: 3100 -18707.695 0.011 0.012
Chain 1: 3200 -18398.493 0.012 0.015
Chain 1: 3300 -18603.107 0.011 0.012
Chain 1: 3400 -18078.161 0.013 0.015
Chain 1: 3500 -18689.810 0.015 0.015
Chain 1: 3600 -17996.678 0.017 0.015
Chain 1: 3700 -18383.297 0.018 0.017
Chain 1: 3800 -17343.352 0.023 0.021
Chain 1: 3900 -17339.431 0.021 0.021
Chain 1: 4000 -17456.782 0.022 0.021
Chain 1: 4100 -17370.569 0.022 0.021
Chain 1: 4200 -17186.849 0.021 0.021
Chain 1: 4300 -17325.264 0.021 0.021
Chain 1: 4400 -17282.141 0.019 0.011
Chain 1: 4500 -17184.606 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001279 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.79 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48348.065 1.000 1.000
Chain 1: 200 -20534.004 1.177 1.355
Chain 1: 300 -13681.088 0.952 1.000
Chain 1: 400 -19563.371 0.789 1.000
Chain 1: 500 -21262.295 0.647 0.501
Chain 1: 600 -11673.723 0.676 0.821
Chain 1: 700 -11087.129 0.587 0.501
Chain 1: 800 -14044.135 0.540 0.501
Chain 1: 900 -15756.470 0.492 0.301
Chain 1: 1000 -12725.574 0.467 0.301
Chain 1: 1100 -19546.422 0.402 0.301
Chain 1: 1200 -10592.044 0.351 0.301
Chain 1: 1300 -11536.521 0.309 0.238
Chain 1: 1400 -11704.546 0.280 0.211
Chain 1: 1500 -10760.593 0.281 0.211
Chain 1: 1600 -11999.855 0.209 0.109
Chain 1: 1700 -9558.227 0.229 0.211
Chain 1: 1800 -11626.500 0.226 0.178
Chain 1: 1900 -9609.265 0.236 0.210
Chain 1: 2000 -9436.067 0.214 0.178
Chain 1: 2100 -9188.827 0.182 0.103
Chain 1: 2200 -12649.902 0.125 0.103
Chain 1: 2300 -11445.988 0.127 0.105
Chain 1: 2400 -17444.154 0.160 0.178
Chain 1: 2500 -9363.896 0.238 0.210
Chain 1: 2600 -8948.062 0.232 0.210
Chain 1: 2700 -9183.194 0.209 0.178
Chain 1: 2800 -9720.480 0.197 0.105
Chain 1: 2900 -9344.453 0.180 0.055
Chain 1: 3000 -15078.456 0.216 0.105
Chain 1: 3100 -9378.492 0.274 0.274
Chain 1: 3200 -8533.936 0.257 0.105
Chain 1: 3300 -13627.669 0.284 0.344
Chain 1: 3400 -8656.462 0.307 0.374
Chain 1: 3500 -9911.904 0.233 0.127
Chain 1: 3600 -12708.502 0.250 0.220
Chain 1: 3700 -8604.272 0.295 0.374
Chain 1: 3800 -11075.545 0.312 0.374
Chain 1: 3900 -9535.781 0.324 0.374
Chain 1: 4000 -8513.865 0.298 0.223
Chain 1: 4100 -8550.555 0.238 0.220
Chain 1: 4200 -12583.247 0.260 0.223
Chain 1: 4300 -11301.113 0.234 0.220
Chain 1: 4400 -9102.701 0.201 0.220
Chain 1: 4500 -10588.754 0.202 0.220
Chain 1: 4600 -8437.477 0.206 0.223
Chain 1: 4700 -11172.679 0.182 0.223
Chain 1: 4800 -14399.770 0.183 0.224
Chain 1: 4900 -9402.231 0.220 0.242
Chain 1: 5000 -9061.833 0.211 0.242
Chain 1: 5100 -8789.493 0.214 0.242
Chain 1: 5200 -8539.014 0.185 0.224
Chain 1: 5300 -10570.520 0.193 0.224
Chain 1: 5400 -8369.025 0.195 0.224
Chain 1: 5500 -11090.738 0.205 0.245
Chain 1: 5600 -8465.862 0.211 0.245
Chain 1: 5700 -8448.359 0.187 0.224
Chain 1: 5800 -8423.255 0.165 0.192
Chain 1: 5900 -8223.953 0.114 0.038
Chain 1: 6000 -8332.910 0.111 0.031
Chain 1: 6100 -8938.926 0.115 0.068
Chain 1: 6200 -8365.890 0.119 0.068
Chain 1: 6300 -12139.326 0.131 0.068
Chain 1: 6400 -8888.505 0.141 0.068
Chain 1: 6500 -9195.342 0.120 0.068
Chain 1: 6600 -9112.814 0.090 0.033
Chain 1: 6700 -9814.853 0.097 0.068
Chain 1: 6800 -8553.445 0.111 0.068
Chain 1: 6900 -8287.918 0.112 0.068
Chain 1: 7000 -8523.735 0.113 0.068
Chain 1: 7100 -8158.348 0.111 0.068
Chain 1: 7200 -9674.328 0.120 0.072
Chain 1: 7300 -8386.515 0.104 0.072
Chain 1: 7400 -9440.494 0.079 0.072
Chain 1: 7500 -11179.170 0.091 0.112
Chain 1: 7600 -8304.805 0.125 0.147
Chain 1: 7700 -8297.758 0.118 0.147
Chain 1: 7800 -8382.387 0.104 0.112
Chain 1: 7900 -8198.704 0.103 0.112
Chain 1: 8000 -8291.885 0.101 0.112
Chain 1: 8100 -8204.749 0.098 0.112
Chain 1: 8200 -8124.526 0.083 0.022
Chain 1: 8300 -8133.577 0.068 0.011
Chain 1: 8400 -10648.232 0.080 0.011
Chain 1: 8500 -11427.288 0.072 0.011
Chain 1: 8600 -8406.666 0.073 0.011
Chain 1: 8700 -7936.235 0.079 0.022
Chain 1: 8800 -8017.380 0.079 0.022
Chain 1: 8900 -8335.709 0.080 0.038
Chain 1: 9000 -8319.966 0.079 0.038
Chain 1: 9100 -9820.088 0.094 0.059
Chain 1: 9200 -8660.695 0.106 0.068
Chain 1: 9300 -8301.655 0.110 0.068
Chain 1: 9400 -9492.322 0.099 0.068
Chain 1: 9500 -8815.131 0.100 0.077
Chain 1: 9600 -10634.765 0.081 0.077
Chain 1: 9700 -11328.696 0.081 0.077
Chain 1: 9800 -10549.991 0.088 0.077
Chain 1: 9900 -7890.882 0.118 0.125
Chain 1: 10000 -8366.977 0.123 0.125
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001553 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.53 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61372.086 1.000 1.000
Chain 1: 200 -17538.433 1.750 2.499
Chain 1: 300 -8670.862 1.507 1.023
Chain 1: 400 -8907.997 1.137 1.023
Chain 1: 500 -7880.467 0.936 1.000
Chain 1: 600 -8822.924 0.798 1.000
Chain 1: 700 -7828.879 0.702 0.130
Chain 1: 800 -8076.851 0.618 0.130
Chain 1: 900 -7605.552 0.556 0.127
Chain 1: 1000 -7702.085 0.502 0.127
Chain 1: 1100 -7687.643 0.402 0.107
Chain 1: 1200 -7593.028 0.153 0.062
Chain 1: 1300 -7576.104 0.051 0.031
Chain 1: 1400 -7679.844 0.050 0.031
Chain 1: 1500 -7573.053 0.038 0.014
Chain 1: 1600 -7518.444 0.028 0.014
Chain 1: 1700 -7461.782 0.016 0.013
Chain 1: 1800 -7522.642 0.014 0.012
Chain 1: 1900 -7522.414 0.008 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00251 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.1 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85732.830 1.000 1.000
Chain 1: 200 -13178.989 3.253 5.505
Chain 1: 300 -9646.929 2.290 1.000
Chain 1: 400 -10521.918 1.739 1.000
Chain 1: 500 -8527.670 1.438 0.366
Chain 1: 600 -8393.492 1.201 0.366
Chain 1: 700 -8550.349 1.032 0.234
Chain 1: 800 -8751.509 0.906 0.234
Chain 1: 900 -8543.603 0.808 0.083
Chain 1: 1000 -8215.572 0.731 0.083
Chain 1: 1100 -8511.678 0.634 0.040 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8189.838 0.088 0.039
Chain 1: 1300 -8390.882 0.054 0.035
Chain 1: 1400 -8407.616 0.046 0.024
Chain 1: 1500 -8296.652 0.023 0.024
Chain 1: 1600 -8384.021 0.023 0.024
Chain 1: 1700 -8486.153 0.022 0.024
Chain 1: 1800 -8100.188 0.025 0.024
Chain 1: 1900 -8201.623 0.024 0.024
Chain 1: 2000 -8171.208 0.020 0.013
Chain 1: 2100 -8310.682 0.018 0.013
Chain 1: 2200 -8091.802 0.017 0.013
Chain 1: 2300 -8233.978 0.016 0.013
Chain 1: 2400 -8118.544 0.017 0.014
Chain 1: 2500 -8177.628 0.017 0.014
Chain 1: 2600 -8192.062 0.016 0.014
Chain 1: 2700 -8114.429 0.016 0.014
Chain 1: 2800 -8095.893 0.011 0.012
Chain 1: 2900 -8108.086 0.010 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003275 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.75 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8374486.348 1.000 1.000
Chain 1: 200 -1579462.353 2.651 4.302
Chain 1: 300 -890390.631 2.025 1.000
Chain 1: 400 -457443.505 1.756 1.000
Chain 1: 500 -358012.152 1.460 0.946
Chain 1: 600 -233194.733 1.306 0.946
Chain 1: 700 -119175.505 1.256 0.946
Chain 1: 800 -86294.335 1.147 0.946
Chain 1: 900 -66591.034 1.052 0.774
Chain 1: 1000 -51333.435 0.977 0.774
Chain 1: 1100 -38764.500 0.909 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37932.133 0.481 0.381
Chain 1: 1300 -25854.238 0.450 0.381
Chain 1: 1400 -25567.907 0.357 0.324
Chain 1: 1500 -22145.622 0.345 0.324
Chain 1: 1600 -21358.364 0.295 0.297
Chain 1: 1700 -20228.609 0.205 0.296
Chain 1: 1800 -20171.582 0.167 0.155
Chain 1: 1900 -20497.046 0.139 0.056
Chain 1: 2000 -19007.446 0.117 0.056
Chain 1: 2100 -19245.985 0.086 0.037
Chain 1: 2200 -19472.165 0.085 0.037
Chain 1: 2300 -19089.749 0.040 0.020
Chain 1: 2400 -18862.008 0.040 0.020
Chain 1: 2500 -18664.150 0.026 0.016
Chain 1: 2600 -18294.935 0.024 0.016
Chain 1: 2700 -18252.075 0.019 0.012
Chain 1: 2800 -17969.218 0.020 0.016
Chain 1: 2900 -18250.242 0.020 0.015
Chain 1: 3000 -18236.524 0.012 0.012
Chain 1: 3100 -18321.379 0.011 0.012
Chain 1: 3200 -18012.488 0.012 0.015
Chain 1: 3300 -18216.884 0.011 0.012
Chain 1: 3400 -17692.592 0.013 0.015
Chain 1: 3500 -18303.336 0.015 0.016
Chain 1: 3600 -17611.537 0.017 0.016
Chain 1: 3700 -17997.204 0.019 0.017
Chain 1: 3800 -16959.302 0.023 0.021
Chain 1: 3900 -16955.521 0.022 0.021
Chain 1: 4000 -17072.800 0.023 0.021
Chain 1: 4100 -16986.645 0.023 0.021
Chain 1: 4200 -16803.445 0.022 0.021
Chain 1: 4300 -16941.458 0.022 0.021
Chain 1: 4400 -16898.716 0.019 0.011
Chain 1: 4500 -16801.328 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001303 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.03 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12857.533 1.000 1.000
Chain 1: 200 -9683.400 0.664 1.000
Chain 1: 300 -8567.771 0.486 0.328
Chain 1: 400 -8666.691 0.367 0.328
Chain 1: 500 -8625.409 0.295 0.130
Chain 1: 600 -8473.531 0.249 0.130
Chain 1: 700 -8391.707 0.215 0.018
Chain 1: 800 -8356.118 0.188 0.018
Chain 1: 900 -8403.859 0.168 0.011
Chain 1: 1000 -8472.660 0.152 0.011
Chain 1: 1100 -8533.853 0.053 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00171 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 17.1 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58399.693 1.000 1.000
Chain 1: 200 -18033.853 1.619 2.238
Chain 1: 300 -8870.086 1.424 1.033
Chain 1: 400 -8054.439 1.093 1.033
Chain 1: 500 -8849.776 0.893 1.000
Chain 1: 600 -9492.505 0.755 1.000
Chain 1: 700 -8297.004 0.668 0.144
Chain 1: 800 -8273.830 0.585 0.144
Chain 1: 900 -7948.104 0.524 0.101
Chain 1: 1000 -7759.046 0.474 0.101
Chain 1: 1100 -7743.685 0.374 0.090
Chain 1: 1200 -7718.624 0.151 0.068
Chain 1: 1300 -7616.277 0.049 0.041
Chain 1: 1400 -7939.643 0.043 0.041
Chain 1: 1500 -7610.744 0.038 0.041
Chain 1: 1600 -7760.514 0.033 0.024
Chain 1: 1700 -7594.154 0.021 0.022
Chain 1: 1800 -7605.060 0.021 0.022
Chain 1: 1900 -7624.691 0.017 0.019
Chain 1: 2000 -7685.235 0.016 0.013
Chain 1: 2100 -7597.180 0.017 0.013
Chain 1: 2200 -7836.267 0.019 0.019
Chain 1: 2300 -7561.189 0.022 0.022
Chain 1: 2400 -7562.904 0.018 0.019
Chain 1: 2500 -7627.172 0.014 0.012
Chain 1: 2600 -7549.859 0.013 0.010
Chain 1: 2700 -7476.630 0.012 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003196 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.96 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87131.396 1.000 1.000
Chain 1: 200 -13928.667 3.128 5.256
Chain 1: 300 -10346.883 2.201 1.000
Chain 1: 400 -11182.653 1.669 1.000
Chain 1: 500 -9298.823 1.376 0.346
Chain 1: 600 -8871.502 1.155 0.346
Chain 1: 700 -8771.321 0.991 0.203
Chain 1: 800 -9284.214 0.874 0.203
Chain 1: 900 -9162.644 0.779 0.075
Chain 1: 1000 -8941.759 0.703 0.075
Chain 1: 1100 -9161.829 0.606 0.055 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8840.257 0.084 0.048
Chain 1: 1300 -9057.474 0.051 0.036
Chain 1: 1400 -9041.654 0.044 0.025
Chain 1: 1500 -8938.870 0.025 0.024
Chain 1: 1600 -9042.600 0.021 0.024
Chain 1: 1700 -9131.312 0.021 0.024
Chain 1: 1800 -8730.206 0.020 0.024
Chain 1: 1900 -8829.967 0.020 0.024
Chain 1: 2000 -8801.278 0.018 0.011
Chain 1: 2100 -8921.174 0.017 0.011
Chain 1: 2200 -8706.650 0.016 0.011
Chain 1: 2300 -8860.734 0.015 0.011
Chain 1: 2400 -8742.320 0.016 0.013
Chain 1: 2500 -8805.546 0.016 0.013
Chain 1: 2600 -8826.873 0.015 0.013
Chain 1: 2700 -8745.957 0.015 0.013
Chain 1: 2800 -8720.230 0.011 0.011
Chain 1: 2900 -8775.576 0.010 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003202 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.02 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8423598.240 1.000 1.000
Chain 1: 200 -1587510.175 2.653 4.306
Chain 1: 300 -891701.932 2.029 1.000
Chain 1: 400 -458905.695 1.757 1.000
Chain 1: 500 -358785.725 1.462 0.943
Chain 1: 600 -233412.038 1.308 0.943
Chain 1: 700 -119555.955 1.257 0.943
Chain 1: 800 -86809.302 1.147 0.943
Chain 1: 900 -67144.270 1.052 0.780
Chain 1: 1000 -51948.018 0.976 0.780
Chain 1: 1100 -39445.714 0.908 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38617.284 0.479 0.377
Chain 1: 1300 -26595.597 0.446 0.377
Chain 1: 1400 -26315.794 0.353 0.317
Chain 1: 1500 -22909.891 0.340 0.317
Chain 1: 1600 -22128.400 0.290 0.293
Chain 1: 1700 -21004.693 0.200 0.293
Chain 1: 1800 -20949.330 0.163 0.149
Chain 1: 1900 -21275.189 0.135 0.053
Chain 1: 2000 -19788.427 0.113 0.053
Chain 1: 2100 -20026.566 0.083 0.035
Chain 1: 2200 -20252.774 0.082 0.035
Chain 1: 2300 -19870.225 0.038 0.019
Chain 1: 2400 -19642.401 0.038 0.019
Chain 1: 2500 -19444.519 0.025 0.015
Chain 1: 2600 -19074.949 0.023 0.015
Chain 1: 2700 -19031.945 0.018 0.012
Chain 1: 2800 -18749.032 0.019 0.015
Chain 1: 2900 -19030.025 0.019 0.015
Chain 1: 3000 -19016.228 0.012 0.012
Chain 1: 3100 -19101.236 0.011 0.012
Chain 1: 3200 -18792.088 0.011 0.015
Chain 1: 3300 -18996.627 0.011 0.012
Chain 1: 3400 -18471.957 0.012 0.015
Chain 1: 3500 -19083.312 0.014 0.015
Chain 1: 3600 -18390.531 0.016 0.015
Chain 1: 3700 -18776.933 0.018 0.016
Chain 1: 3800 -17737.643 0.022 0.021
Chain 1: 3900 -17733.787 0.021 0.021
Chain 1: 4000 -17851.077 0.022 0.021
Chain 1: 4100 -17764.976 0.022 0.021
Chain 1: 4200 -17581.351 0.021 0.021
Chain 1: 4300 -17719.628 0.021 0.021
Chain 1: 4400 -17676.605 0.018 0.010
Chain 1: 4500 -17579.154 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001396 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.96 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12528.725 1.000 1.000
Chain 1: 200 -9446.929 0.663 1.000
Chain 1: 300 -8089.372 0.498 0.326
Chain 1: 400 -8314.600 0.380 0.326
Chain 1: 500 -8212.058 0.307 0.168
Chain 1: 600 -8048.023 0.259 0.168
Chain 1: 700 -7941.235 0.224 0.027
Chain 1: 800 -7929.892 0.196 0.027
Chain 1: 900 -7897.526 0.175 0.020
Chain 1: 1000 -8074.259 0.159 0.022
Chain 1: 1100 -8082.829 0.060 0.020
Chain 1: 1200 -7964.442 0.028 0.015
Chain 1: 1300 -7935.743 0.012 0.013
Chain 1: 1400 -7939.595 0.009 0.012 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001394 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.94 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58854.270 1.000 1.000
Chain 1: 200 -17942.692 1.640 2.280
Chain 1: 300 -8784.358 1.441 1.043
Chain 1: 400 -8150.012 1.100 1.043
Chain 1: 500 -8644.522 0.892 1.000
Chain 1: 600 -8039.690 0.755 1.000
Chain 1: 700 -7902.766 0.650 0.078
Chain 1: 800 -8311.852 0.575 0.078
Chain 1: 900 -8067.357 0.514 0.075
Chain 1: 1000 -7909.723 0.465 0.075
Chain 1: 1100 -7612.505 0.369 0.057
Chain 1: 1200 -7768.687 0.143 0.049
Chain 1: 1300 -7775.578 0.039 0.039
Chain 1: 1400 -7950.359 0.033 0.030
Chain 1: 1500 -7551.083 0.033 0.030
Chain 1: 1600 -7747.461 0.028 0.025
Chain 1: 1700 -7542.713 0.029 0.027
Chain 1: 1800 -7565.476 0.024 0.025
Chain 1: 1900 -7596.518 0.021 0.022
Chain 1: 2000 -7629.265 0.020 0.022
Chain 1: 2100 -7589.812 0.016 0.020
Chain 1: 2200 -7718.961 0.016 0.017
Chain 1: 2300 -7572.268 0.018 0.019
Chain 1: 2400 -7636.678 0.017 0.017
Chain 1: 2500 -7565.624 0.012 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003331 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.31 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86144.426 1.000 1.000
Chain 1: 200 -13691.845 3.146 5.292
Chain 1: 300 -10005.745 2.220 1.000
Chain 1: 400 -11318.207 1.694 1.000
Chain 1: 500 -8992.275 1.407 0.368
Chain 1: 600 -8496.207 1.182 0.368
Chain 1: 700 -8630.822 1.016 0.259
Chain 1: 800 -8882.646 0.892 0.259
Chain 1: 900 -8812.186 0.794 0.116
Chain 1: 1000 -8773.232 0.715 0.116
Chain 1: 1100 -8602.613 0.617 0.058 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8387.445 0.090 0.028
Chain 1: 1300 -8685.738 0.057 0.028
Chain 1: 1400 -8638.361 0.046 0.026
Chain 1: 1500 -8527.907 0.021 0.020
Chain 1: 1600 -8635.187 0.017 0.016
Chain 1: 1700 -8708.581 0.016 0.013
Chain 1: 1800 -8276.371 0.018 0.013
Chain 1: 1900 -8380.534 0.019 0.013
Chain 1: 2000 -8355.862 0.019 0.013
Chain 1: 2100 -8328.107 0.017 0.012
Chain 1: 2200 -8297.860 0.015 0.012
Chain 1: 2300 -8428.101 0.013 0.012
Chain 1: 2400 -8283.316 0.014 0.012
Chain 1: 2500 -8352.151 0.014 0.012
Chain 1: 2600 -8271.446 0.013 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003274 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.74 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8412366.111 1.000 1.000
Chain 1: 200 -1586631.415 2.651 4.302
Chain 1: 300 -891088.696 2.028 1.000
Chain 1: 400 -458363.170 1.757 1.000
Chain 1: 500 -358483.425 1.461 0.944
Chain 1: 600 -233307.933 1.307 0.944
Chain 1: 700 -119465.688 1.256 0.944
Chain 1: 800 -86664.688 1.147 0.944
Chain 1: 900 -67000.093 1.052 0.781
Chain 1: 1000 -51799.601 0.976 0.781
Chain 1: 1100 -39277.765 0.908 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38455.804 0.480 0.378
Chain 1: 1300 -26410.031 0.447 0.378
Chain 1: 1400 -26129.910 0.354 0.319
Chain 1: 1500 -22716.756 0.341 0.319
Chain 1: 1600 -21933.719 0.291 0.294
Chain 1: 1700 -20806.799 0.201 0.293
Chain 1: 1800 -20751.090 0.164 0.150
Chain 1: 1900 -21077.500 0.136 0.054
Chain 1: 2000 -19587.836 0.114 0.054
Chain 1: 2100 -19826.234 0.083 0.036
Chain 1: 2200 -20053.012 0.082 0.036
Chain 1: 2300 -19669.855 0.039 0.019
Chain 1: 2400 -19441.836 0.039 0.019
Chain 1: 2500 -19243.906 0.025 0.015
Chain 1: 2600 -18873.790 0.023 0.015
Chain 1: 2700 -18830.607 0.018 0.012
Chain 1: 2800 -18547.355 0.019 0.015
Chain 1: 2900 -18828.738 0.019 0.015
Chain 1: 3000 -18814.887 0.012 0.012
Chain 1: 3100 -18899.968 0.011 0.012
Chain 1: 3200 -18590.436 0.012 0.015
Chain 1: 3300 -18795.307 0.011 0.012
Chain 1: 3400 -18269.886 0.012 0.015
Chain 1: 3500 -18882.316 0.015 0.015
Chain 1: 3600 -18188.224 0.016 0.015
Chain 1: 3700 -18575.613 0.018 0.017
Chain 1: 3800 -17534.200 0.023 0.021
Chain 1: 3900 -17530.319 0.021 0.021
Chain 1: 4000 -17647.613 0.022 0.021
Chain 1: 4100 -17561.350 0.022 0.021
Chain 1: 4200 -17377.322 0.021 0.021
Chain 1: 4300 -17515.908 0.021 0.021
Chain 1: 4400 -17472.522 0.018 0.011
Chain 1: 4500 -17375.028 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001327 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.27 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49532.656 1.000 1.000
Chain 1: 200 -23046.233 1.075 1.149
Chain 1: 300 -16785.902 0.841 1.000
Chain 1: 400 -17527.600 0.641 1.000
Chain 1: 500 -12319.643 0.597 0.423
Chain 1: 600 -16102.625 0.537 0.423
Chain 1: 700 -20345.538 0.490 0.373
Chain 1: 800 -11824.533 0.519 0.423
Chain 1: 900 -16560.760 0.493 0.373
Chain 1: 1000 -12419.903 0.477 0.373
Chain 1: 1100 -20647.898 0.417 0.373
Chain 1: 1200 -15841.571 0.332 0.333
Chain 1: 1300 -11801.498 0.329 0.333
Chain 1: 1400 -11298.171 0.330 0.333
Chain 1: 1500 -13463.353 0.303 0.303
Chain 1: 1600 -12258.084 0.290 0.303
Chain 1: 1700 -15018.823 0.287 0.303
Chain 1: 1800 -10252.492 0.262 0.303
Chain 1: 1900 -10800.409 0.238 0.303
Chain 1: 2000 -10737.055 0.205 0.184
Chain 1: 2100 -10392.413 0.169 0.161
Chain 1: 2200 -18779.089 0.183 0.161
Chain 1: 2300 -17909.885 0.154 0.098
Chain 1: 2400 -9942.775 0.229 0.161
Chain 1: 2500 -14850.632 0.246 0.184
Chain 1: 2600 -9552.235 0.292 0.330
Chain 1: 2700 -9351.393 0.276 0.330
Chain 1: 2800 -9639.875 0.232 0.051
Chain 1: 2900 -9172.773 0.232 0.051
Chain 1: 3000 -9474.946 0.235 0.051
Chain 1: 3100 -9970.417 0.237 0.051
Chain 1: 3200 -11349.164 0.204 0.051
Chain 1: 3300 -10842.138 0.204 0.051
Chain 1: 3400 -13968.089 0.146 0.051
Chain 1: 3500 -9169.012 0.165 0.051
Chain 1: 3600 -9456.642 0.113 0.050
Chain 1: 3700 -9794.935 0.114 0.050
Chain 1: 3800 -8759.613 0.123 0.051
Chain 1: 3900 -9522.373 0.126 0.080
Chain 1: 4000 -9674.199 0.124 0.080
Chain 1: 4100 -9731.840 0.120 0.080
Chain 1: 4200 -16104.374 0.147 0.080
Chain 1: 4300 -10154.819 0.201 0.118
Chain 1: 4400 -16166.672 0.216 0.118
Chain 1: 4500 -17709.187 0.173 0.087
Chain 1: 4600 -12700.275 0.209 0.118
Chain 1: 4700 -10444.740 0.227 0.216
Chain 1: 4800 -9976.045 0.220 0.216
Chain 1: 4900 -9289.744 0.219 0.216
Chain 1: 5000 -13934.597 0.251 0.333
Chain 1: 5100 -14011.536 0.251 0.333
Chain 1: 5200 -9584.144 0.258 0.333
Chain 1: 5300 -11229.535 0.214 0.216
Chain 1: 5400 -16481.255 0.208 0.216
Chain 1: 5500 -11141.283 0.248 0.319
Chain 1: 5600 -9992.961 0.220 0.216
Chain 1: 5700 -10015.300 0.198 0.147
Chain 1: 5800 -8680.377 0.209 0.154
Chain 1: 5900 -9008.052 0.205 0.154
Chain 1: 6000 -11410.066 0.193 0.154
Chain 1: 6100 -10599.520 0.200 0.154
Chain 1: 6200 -10079.925 0.159 0.147
Chain 1: 6300 -13656.268 0.171 0.154
Chain 1: 6400 -10144.865 0.173 0.154
Chain 1: 6500 -9359.181 0.134 0.115
Chain 1: 6600 -8681.268 0.130 0.084
Chain 1: 6700 -8720.396 0.130 0.084
Chain 1: 6800 -13544.149 0.151 0.084
Chain 1: 6900 -8708.671 0.202 0.211
Chain 1: 7000 -9341.838 0.188 0.084
Chain 1: 7100 -9194.434 0.182 0.084
Chain 1: 7200 -8540.889 0.185 0.084
Chain 1: 7300 -8444.826 0.160 0.078
Chain 1: 7400 -8383.539 0.126 0.077
Chain 1: 7500 -9547.143 0.129 0.077
Chain 1: 7600 -8941.678 0.128 0.068
Chain 1: 7700 -8929.086 0.128 0.068
Chain 1: 7800 -11829.829 0.117 0.068
Chain 1: 7900 -8990.779 0.093 0.068
Chain 1: 8000 -8410.129 0.093 0.069
Chain 1: 8100 -9565.513 0.104 0.077
Chain 1: 8200 -8522.111 0.108 0.121
Chain 1: 8300 -13813.065 0.145 0.122
Chain 1: 8400 -10144.451 0.181 0.122
Chain 1: 8500 -8483.567 0.188 0.196
Chain 1: 8600 -11429.321 0.207 0.245
Chain 1: 8700 -9318.081 0.230 0.245
Chain 1: 8800 -9459.270 0.207 0.227
Chain 1: 8900 -9651.034 0.177 0.196
Chain 1: 9000 -8687.954 0.181 0.196
Chain 1: 9100 -8648.669 0.170 0.196
Chain 1: 9200 -8124.062 0.164 0.196
Chain 1: 9300 -10503.519 0.148 0.196
Chain 1: 9400 -8897.444 0.130 0.181
Chain 1: 9500 -8695.918 0.113 0.111
Chain 1: 9600 -10537.624 0.105 0.111
Chain 1: 9700 -8004.790 0.114 0.111
Chain 1: 9800 -13514.553 0.153 0.175
Chain 1: 9900 -11932.340 0.164 0.175
Chain 1: 10000 -9302.225 0.181 0.181
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001412 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.12 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -46580.706 1.000 1.000
Chain 1: 200 -15818.371 1.472 1.945
Chain 1: 300 -8814.088 1.246 1.000
Chain 1: 400 -8150.168 0.955 1.000
Chain 1: 500 -7903.177 0.770 0.795
Chain 1: 600 -8208.239 0.648 0.795
Chain 1: 700 -7784.314 0.563 0.081
Chain 1: 800 -8207.777 0.499 0.081
Chain 1: 900 -7874.975 0.449 0.054
Chain 1: 1000 -7691.925 0.406 0.054
Chain 1: 1100 -7644.739 0.307 0.052
Chain 1: 1200 -7561.410 0.113 0.042
Chain 1: 1300 -7547.610 0.034 0.037
Chain 1: 1400 -7806.020 0.029 0.033
Chain 1: 1500 -7542.329 0.030 0.035
Chain 1: 1600 -7681.644 0.028 0.033
Chain 1: 1700 -7517.805 0.024 0.024
Chain 1: 1800 -7648.076 0.021 0.022
Chain 1: 1900 -7545.125 0.018 0.018
Chain 1: 2000 -7514.406 0.016 0.017
Chain 1: 2100 -7529.229 0.016 0.017
Chain 1: 2200 -7683.805 0.017 0.018
Chain 1: 2300 -7507.356 0.019 0.020
Chain 1: 2400 -7485.876 0.016 0.018
Chain 1: 2500 -7532.934 0.013 0.017
Chain 1: 2600 -7468.082 0.012 0.014
Chain 1: 2700 -7395.439 0.011 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002584 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.84 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87197.370 1.000 1.000
Chain 1: 200 -13792.639 3.161 5.322
Chain 1: 300 -10052.921 2.231 1.000
Chain 1: 400 -11248.050 1.700 1.000
Chain 1: 500 -8658.387 1.420 0.372
Chain 1: 600 -8396.103 1.188 0.372
Chain 1: 700 -8470.926 1.020 0.299
Chain 1: 800 -8691.571 0.896 0.299
Chain 1: 900 -8763.486 0.797 0.106
Chain 1: 1000 -8868.439 0.718 0.106
Chain 1: 1100 -8740.500 0.620 0.031 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8376.235 0.092 0.031
Chain 1: 1300 -8677.145 0.058 0.031
Chain 1: 1400 -8480.119 0.050 0.025
Chain 1: 1500 -8539.985 0.021 0.023
Chain 1: 1600 -8647.195 0.019 0.015
Chain 1: 1700 -8704.722 0.019 0.015
Chain 1: 1800 -8260.960 0.022 0.015
Chain 1: 1900 -8368.849 0.022 0.015
Chain 1: 2000 -8355.526 0.021 0.015
Chain 1: 2100 -8472.110 0.021 0.014
Chain 1: 2200 -8266.473 0.019 0.014
Chain 1: 2300 -8361.931 0.017 0.013
Chain 1: 2400 -8428.859 0.015 0.012
Chain 1: 2500 -8377.273 0.015 0.012
Chain 1: 2600 -8391.295 0.014 0.011
Chain 1: 2700 -8298.755 0.015 0.011
Chain 1: 2800 -8246.188 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002953 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.53 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8427715.595 1.000 1.000
Chain 1: 200 -1587636.329 2.654 4.308
Chain 1: 300 -890413.587 2.030 1.000
Chain 1: 400 -457723.929 1.759 1.000
Chain 1: 500 -357716.632 1.463 0.945
Chain 1: 600 -232795.685 1.309 0.945
Chain 1: 700 -119224.025 1.258 0.945
Chain 1: 800 -86536.702 1.148 0.945
Chain 1: 900 -66928.455 1.053 0.783
Chain 1: 1000 -51779.325 0.977 0.783
Chain 1: 1100 -39303.819 0.909 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38488.372 0.480 0.378
Chain 1: 1300 -26479.896 0.447 0.378
Chain 1: 1400 -26205.366 0.353 0.317
Chain 1: 1500 -22801.702 0.340 0.317
Chain 1: 1600 -22021.982 0.290 0.293
Chain 1: 1700 -20899.044 0.200 0.293
Chain 1: 1800 -20844.395 0.163 0.149
Chain 1: 1900 -21171.084 0.135 0.054
Chain 1: 2000 -19683.069 0.113 0.054
Chain 1: 2100 -19921.482 0.083 0.035
Chain 1: 2200 -20148.036 0.082 0.035
Chain 1: 2300 -19764.967 0.039 0.019
Chain 1: 2400 -19536.860 0.039 0.019
Chain 1: 2500 -19338.782 0.025 0.015
Chain 1: 2600 -18968.562 0.023 0.015
Chain 1: 2700 -18925.392 0.018 0.012
Chain 1: 2800 -18642.002 0.019 0.015
Chain 1: 2900 -18923.394 0.019 0.015
Chain 1: 3000 -18909.586 0.012 0.012
Chain 1: 3100 -18994.669 0.011 0.012
Chain 1: 3200 -18685.012 0.011 0.015
Chain 1: 3300 -18889.976 0.011 0.012
Chain 1: 3400 -18364.303 0.012 0.015
Chain 1: 3500 -18977.080 0.015 0.015
Chain 1: 3600 -18282.479 0.016 0.015
Chain 1: 3700 -18670.198 0.018 0.017
Chain 1: 3800 -17627.979 0.023 0.021
Chain 1: 3900 -17624.015 0.021 0.021
Chain 1: 4000 -17741.350 0.022 0.021
Chain 1: 4100 -17655.046 0.022 0.021
Chain 1: 4200 -17470.813 0.021 0.021
Chain 1: 4300 -17609.557 0.021 0.021
Chain 1: 4400 -17566.016 0.018 0.011
Chain 1: 4500 -17468.441 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002536 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.36 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48428.758 1.000 1.000
Chain 1: 200 -13853.399 1.748 2.496
Chain 1: 300 -12273.146 1.208 1.000
Chain 1: 400 -17225.661 0.978 1.000
Chain 1: 500 -17556.264 0.786 0.288
Chain 1: 600 -17902.966 0.658 0.288
Chain 1: 700 -10726.931 0.660 0.288
Chain 1: 800 -13701.962 0.605 0.288
Chain 1: 900 -17791.085 0.563 0.230
Chain 1: 1000 -13454.740 0.539 0.288
Chain 1: 1100 -17256.184 0.461 0.230
Chain 1: 1200 -13072.821 0.243 0.230
Chain 1: 1300 -9749.954 0.265 0.288
Chain 1: 1400 -11595.125 0.252 0.230
Chain 1: 1500 -10279.292 0.263 0.230
Chain 1: 1600 -9778.211 0.266 0.230
Chain 1: 1700 -9804.068 0.199 0.220
Chain 1: 1800 -9289.783 0.183 0.220
Chain 1: 1900 -14720.601 0.197 0.220
Chain 1: 2000 -11403.154 0.194 0.220
Chain 1: 2100 -9467.463 0.192 0.204
Chain 1: 2200 -14254.100 0.194 0.204
Chain 1: 2300 -9399.204 0.211 0.204
Chain 1: 2400 -8373.043 0.208 0.204
Chain 1: 2500 -14619.712 0.238 0.291
Chain 1: 2600 -8572.765 0.303 0.336
Chain 1: 2700 -10030.526 0.317 0.336
Chain 1: 2800 -8826.585 0.325 0.336
Chain 1: 2900 -12444.250 0.318 0.291
Chain 1: 3000 -11272.256 0.299 0.291
Chain 1: 3100 -9291.530 0.300 0.291
Chain 1: 3200 -13685.953 0.298 0.291
Chain 1: 3300 -13847.603 0.248 0.213
Chain 1: 3400 -9764.352 0.277 0.291
Chain 1: 3500 -9678.320 0.235 0.213
Chain 1: 3600 -10500.743 0.173 0.145
Chain 1: 3700 -8497.346 0.182 0.213
Chain 1: 3800 -10785.445 0.189 0.213
Chain 1: 3900 -13155.594 0.178 0.212
Chain 1: 4000 -8955.287 0.215 0.213
Chain 1: 4100 -8480.441 0.199 0.212
Chain 1: 4200 -12400.472 0.199 0.212
Chain 1: 4300 -8503.580 0.243 0.236
Chain 1: 4400 -8657.748 0.203 0.212
Chain 1: 4500 -9636.085 0.213 0.212
Chain 1: 4600 -9688.206 0.205 0.212
Chain 1: 4700 -8275.536 0.199 0.180
Chain 1: 4800 -8183.059 0.179 0.171
Chain 1: 4900 -8998.890 0.170 0.102
Chain 1: 5000 -14777.784 0.162 0.102
Chain 1: 5100 -13646.109 0.165 0.102
Chain 1: 5200 -12561.595 0.142 0.091
Chain 1: 5300 -13239.770 0.101 0.086
Chain 1: 5400 -8377.770 0.157 0.091
Chain 1: 5500 -12722.183 0.181 0.091
Chain 1: 5600 -11024.022 0.196 0.154
Chain 1: 5700 -8256.038 0.212 0.154
Chain 1: 5800 -8512.312 0.214 0.154
Chain 1: 5900 -11111.394 0.229 0.234
Chain 1: 6000 -10313.615 0.197 0.154
Chain 1: 6100 -8979.448 0.204 0.154
Chain 1: 6200 -8519.903 0.201 0.154
Chain 1: 6300 -13844.472 0.234 0.234
Chain 1: 6400 -12375.304 0.188 0.154
Chain 1: 6500 -9608.997 0.182 0.154
Chain 1: 6600 -9575.567 0.167 0.149
Chain 1: 6700 -9153.864 0.138 0.119
Chain 1: 6800 -11748.766 0.158 0.149
Chain 1: 6900 -8036.486 0.180 0.149
Chain 1: 7000 -8022.564 0.173 0.149
Chain 1: 7100 -7967.985 0.159 0.119
Chain 1: 7200 -8449.604 0.159 0.119
Chain 1: 7300 -11583.454 0.148 0.119
Chain 1: 7400 -11603.587 0.136 0.057
Chain 1: 7500 -10103.148 0.122 0.057
Chain 1: 7600 -7969.090 0.148 0.149
Chain 1: 7700 -9656.580 0.161 0.175
Chain 1: 7800 -8535.147 0.152 0.149
Chain 1: 7900 -7957.039 0.113 0.131
Chain 1: 8000 -9230.166 0.127 0.138
Chain 1: 8100 -8289.988 0.138 0.138
Chain 1: 8200 -8606.260 0.136 0.138
Chain 1: 8300 -7973.106 0.116 0.131
Chain 1: 8400 -11350.877 0.146 0.138
Chain 1: 8500 -7780.183 0.177 0.138
Chain 1: 8600 -11357.501 0.182 0.138
Chain 1: 8700 -8174.013 0.203 0.138
Chain 1: 8800 -8466.404 0.194 0.138
Chain 1: 8900 -10021.539 0.202 0.155
Chain 1: 9000 -9122.296 0.198 0.155
Chain 1: 9100 -8582.043 0.193 0.155
Chain 1: 9200 -7884.168 0.198 0.155
Chain 1: 9300 -8386.559 0.196 0.155
Chain 1: 9400 -8164.569 0.169 0.099
Chain 1: 9500 -10619.670 0.146 0.099
Chain 1: 9600 -9919.425 0.122 0.089
Chain 1: 9700 -7913.972 0.108 0.089
Chain 1: 9800 -8843.921 0.115 0.099
Chain 1: 9900 -10179.790 0.113 0.099
Chain 1: 10000 -8202.097 0.127 0.105
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001526 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.26 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61478.555 1.000 1.000
Chain 1: 200 -17413.019 1.765 2.531
Chain 1: 300 -8607.597 1.518 1.023
Chain 1: 400 -8154.981 1.152 1.023
Chain 1: 500 -8478.152 0.929 1.000
Chain 1: 600 -8070.988 0.783 1.000
Chain 1: 700 -7824.148 0.676 0.056
Chain 1: 800 -7966.811 0.593 0.056
Chain 1: 900 -7695.194 0.531 0.050
Chain 1: 1000 -7600.868 0.479 0.050
Chain 1: 1100 -7541.544 0.380 0.038
Chain 1: 1200 -7486.613 0.128 0.035
Chain 1: 1300 -7532.743 0.026 0.032
Chain 1: 1400 -7747.119 0.023 0.028
Chain 1: 1500 -7503.492 0.023 0.028
Chain 1: 1600 -7431.991 0.019 0.018
Chain 1: 1700 -7406.037 0.016 0.012
Chain 1: 1800 -7446.845 0.015 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003159 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.59 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85259.674 1.000 1.000
Chain 1: 200 -13048.532 3.267 5.534
Chain 1: 300 -9482.203 2.303 1.000
Chain 1: 400 -10230.286 1.746 1.000
Chain 1: 500 -8392.562 1.440 0.376
Chain 1: 600 -8019.988 1.208 0.376
Chain 1: 700 -8161.338 1.038 0.219
Chain 1: 800 -8675.637 0.916 0.219
Chain 1: 900 -8325.892 0.819 0.073
Chain 1: 1000 -8056.702 0.740 0.073
Chain 1: 1100 -8356.496 0.644 0.059 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -7895.760 0.096 0.058
Chain 1: 1300 -8210.392 0.062 0.046
Chain 1: 1400 -8238.967 0.055 0.042
Chain 1: 1500 -8087.096 0.035 0.038
Chain 1: 1600 -8193.250 0.032 0.036
Chain 1: 1700 -8272.145 0.031 0.036
Chain 1: 1800 -7874.001 0.030 0.036
Chain 1: 1900 -7977.198 0.027 0.033
Chain 1: 2000 -7947.425 0.024 0.019
Chain 1: 2100 -8069.283 0.022 0.015
Chain 1: 2200 -7849.133 0.019 0.015
Chain 1: 2300 -8005.638 0.017 0.015
Chain 1: 2400 -8018.938 0.017 0.015
Chain 1: 2500 -7989.112 0.016 0.013
Chain 1: 2600 -7991.853 0.015 0.013
Chain 1: 2700 -7898.038 0.015 0.013
Chain 1: 2800 -7868.741 0.010 0.012
Chain 1: 2900 -7926.429 0.010 0.007 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003457 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.57 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8386080.958 1.000 1.000
Chain 1: 200 -1578836.164 2.656 4.312
Chain 1: 300 -889449.596 2.029 1.000
Chain 1: 400 -457459.813 1.758 1.000
Chain 1: 500 -358282.219 1.462 0.944
Chain 1: 600 -233133.316 1.307 0.944
Chain 1: 700 -119045.432 1.258 0.944
Chain 1: 800 -86230.141 1.148 0.944
Chain 1: 900 -66497.348 1.053 0.775
Chain 1: 1000 -51244.123 0.978 0.775
Chain 1: 1100 -38679.410 0.910 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37843.747 0.481 0.381
Chain 1: 1300 -25757.688 0.451 0.381
Chain 1: 1400 -25470.581 0.357 0.325
Chain 1: 1500 -22049.084 0.345 0.325
Chain 1: 1600 -21262.658 0.295 0.298
Chain 1: 1700 -20131.322 0.205 0.297
Chain 1: 1800 -20074.139 0.167 0.155
Chain 1: 1900 -20399.955 0.139 0.056
Chain 1: 2000 -18909.485 0.117 0.056
Chain 1: 2100 -19147.545 0.086 0.037
Chain 1: 2200 -19374.551 0.085 0.037
Chain 1: 2300 -18991.432 0.040 0.020
Chain 1: 2400 -18763.607 0.040 0.020
Chain 1: 2500 -18566.056 0.026 0.016
Chain 1: 2600 -18196.154 0.024 0.016
Chain 1: 2700 -18153.050 0.019 0.012
Chain 1: 2800 -17870.331 0.020 0.016
Chain 1: 2900 -18151.443 0.020 0.015
Chain 1: 3000 -18137.487 0.012 0.012
Chain 1: 3100 -18222.512 0.011 0.012
Chain 1: 3200 -17913.294 0.012 0.015
Chain 1: 3300 -18117.914 0.011 0.012
Chain 1: 3400 -17593.271 0.013 0.015
Chain 1: 3500 -18204.647 0.015 0.016
Chain 1: 3600 -17511.949 0.017 0.016
Chain 1: 3700 -17898.391 0.019 0.017
Chain 1: 3800 -16859.189 0.024 0.022
Chain 1: 3900 -16855.436 0.022 0.022
Chain 1: 4000 -16972.660 0.023 0.022
Chain 1: 4100 -16886.576 0.023 0.022
Chain 1: 4200 -16703.006 0.022 0.022
Chain 1: 4300 -16841.199 0.022 0.022
Chain 1: 4400 -16798.194 0.019 0.011
Chain 1: 4500 -16700.840 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001345 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.45 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12540.727 1.000 1.000
Chain 1: 200 -9412.499 0.666 1.000
Chain 1: 300 -8142.569 0.496 0.332
Chain 1: 400 -8322.468 0.377 0.332
Chain 1: 500 -8292.028 0.303 0.156
Chain 1: 600 -8102.570 0.256 0.156
Chain 1: 700 -7993.266 0.222 0.023
Chain 1: 800 -8024.908 0.194 0.023
Chain 1: 900 -8119.398 0.174 0.022
Chain 1: 1000 -8075.030 0.157 0.022
Chain 1: 1100 -8119.269 0.058 0.014
Chain 1: 1200 -8011.044 0.026 0.014
Chain 1: 1300 -8104.110 0.011 0.012
Chain 1: 1400 -7993.564 0.011 0.012
Chain 1: 1500 -8090.700 0.011 0.012
Chain 1: 1600 -8025.815 0.010 0.012 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001393 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.93 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58672.752 1.000 1.000
Chain 1: 200 -17935.902 1.636 2.271
Chain 1: 300 -8984.284 1.423 1.000
Chain 1: 400 -8621.714 1.077 1.000
Chain 1: 500 -8662.436 0.863 0.996
Chain 1: 600 -8615.253 0.720 0.996
Chain 1: 700 -8420.521 0.620 0.042
Chain 1: 800 -8200.434 0.546 0.042
Chain 1: 900 -7889.535 0.490 0.039
Chain 1: 1000 -7936.294 0.442 0.039
Chain 1: 1100 -7871.421 0.342 0.027
Chain 1: 1200 -7754.410 0.117 0.023
Chain 1: 1300 -7897.833 0.019 0.018
Chain 1: 1400 -7911.140 0.015 0.015
Chain 1: 1500 -7670.465 0.018 0.018
Chain 1: 1600 -7842.782 0.019 0.022
Chain 1: 1700 -7635.053 0.020 0.022
Chain 1: 1800 -7737.484 0.018 0.018
Chain 1: 1900 -7733.969 0.014 0.015
Chain 1: 2000 -7664.915 0.015 0.015
Chain 1: 2100 -7734.937 0.015 0.015
Chain 1: 2200 -7780.924 0.014 0.013
Chain 1: 2300 -7643.576 0.014 0.013
Chain 1: 2400 -7720.871 0.015 0.013
Chain 1: 2500 -7687.376 0.012 0.010
Chain 1: 2600 -7602.621 0.011 0.010
Chain 1: 2700 -7624.173 0.008 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002882 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 28.82 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86183.444 1.000 1.000
Chain 1: 200 -13740.225 3.136 5.272
Chain 1: 300 -10057.598 2.213 1.000
Chain 1: 400 -11191.429 1.685 1.000
Chain 1: 500 -9029.183 1.396 0.366
Chain 1: 600 -8612.155 1.171 0.366
Chain 1: 700 -8466.801 1.006 0.239
Chain 1: 800 -8694.873 0.884 0.239
Chain 1: 900 -8950.580 0.789 0.101
Chain 1: 1000 -8545.162 0.715 0.101
Chain 1: 1100 -8862.957 0.618 0.048 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8455.601 0.096 0.048
Chain 1: 1300 -8765.133 0.063 0.047
Chain 1: 1400 -8718.971 0.053 0.036
Chain 1: 1500 -8573.209 0.031 0.035
Chain 1: 1600 -8688.871 0.027 0.029
Chain 1: 1700 -8761.871 0.027 0.029
Chain 1: 1800 -8328.975 0.029 0.035
Chain 1: 1900 -8433.352 0.028 0.035
Chain 1: 2000 -8409.384 0.023 0.017
Chain 1: 2100 -8544.237 0.021 0.016
Chain 1: 2200 -8339.100 0.019 0.016
Chain 1: 2300 -8482.096 0.017 0.016
Chain 1: 2400 -8336.788 0.018 0.017
Chain 1: 2500 -8408.115 0.017 0.016
Chain 1: 2600 -8322.038 0.017 0.016
Chain 1: 2700 -8353.843 0.016 0.016
Chain 1: 2800 -8316.747 0.012 0.012
Chain 1: 2900 -8407.117 0.012 0.011
Chain 1: 3000 -8231.925 0.013 0.016
Chain 1: 3100 -8397.970 0.014 0.017
Chain 1: 3200 -8271.330 0.013 0.015
Chain 1: 3300 -8283.463 0.011 0.011
Chain 1: 3400 -8421.209 0.011 0.011
Chain 1: 3500 -8405.666 0.011 0.011
Chain 1: 3600 -8229.569 0.012 0.015
Chain 1: 3700 -8370.429 0.013 0.016
Chain 1: 3800 -8236.572 0.014 0.016
Chain 1: 3900 -8172.256 0.014 0.016
Chain 1: 4000 -8247.301 0.013 0.016
Chain 1: 4100 -8236.770 0.011 0.015
Chain 1: 4200 -8225.242 0.009 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003492 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.92 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8412961.409 1.000 1.000
Chain 1: 200 -1582313.498 2.658 4.317
Chain 1: 300 -891454.167 2.031 1.000
Chain 1: 400 -458185.911 1.759 1.000
Chain 1: 500 -358789.676 1.463 0.946
Chain 1: 600 -233569.097 1.308 0.946
Chain 1: 700 -119655.499 1.258 0.946
Chain 1: 800 -86835.123 1.148 0.946
Chain 1: 900 -67137.350 1.053 0.775
Chain 1: 1000 -51908.637 0.977 0.775
Chain 1: 1100 -39365.837 0.909 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38540.813 0.479 0.378
Chain 1: 1300 -26469.861 0.447 0.378
Chain 1: 1400 -26187.272 0.354 0.319
Chain 1: 1500 -22767.645 0.341 0.319
Chain 1: 1600 -21982.850 0.291 0.293
Chain 1: 1700 -20852.982 0.201 0.293
Chain 1: 1800 -20796.559 0.164 0.150
Chain 1: 1900 -21122.951 0.136 0.054
Chain 1: 2000 -19631.976 0.114 0.054
Chain 1: 2100 -19870.323 0.083 0.036
Chain 1: 2200 -20097.340 0.082 0.036
Chain 1: 2300 -19714.033 0.039 0.019
Chain 1: 2400 -19486.019 0.039 0.019
Chain 1: 2500 -19288.190 0.025 0.015
Chain 1: 2600 -18917.880 0.023 0.015
Chain 1: 2700 -18874.762 0.018 0.012
Chain 1: 2800 -18591.578 0.019 0.015
Chain 1: 2900 -18872.962 0.019 0.015
Chain 1: 3000 -18859.064 0.012 0.012
Chain 1: 3100 -18944.124 0.011 0.012
Chain 1: 3200 -18634.562 0.012 0.015
Chain 1: 3300 -18839.501 0.011 0.012
Chain 1: 3400 -18314.062 0.012 0.015
Chain 1: 3500 -18926.490 0.015 0.015
Chain 1: 3600 -18232.466 0.016 0.015
Chain 1: 3700 -18619.814 0.018 0.017
Chain 1: 3800 -17578.457 0.023 0.021
Chain 1: 3900 -17574.621 0.021 0.021
Chain 1: 4000 -17691.881 0.022 0.021
Chain 1: 4100 -17605.588 0.022 0.021
Chain 1: 4200 -17421.648 0.021 0.021
Chain 1: 4300 -17560.158 0.021 0.021
Chain 1: 4400 -17516.786 0.018 0.011
Chain 1: 4500 -17419.325 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001328 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.28 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12096.716 1.000 1.000
Chain 1: 200 -9028.110 0.670 1.000
Chain 1: 300 -7982.815 0.490 0.340
Chain 1: 400 -8075.229 0.371 0.340
Chain 1: 500 -7939.594 0.300 0.131
Chain 1: 600 -7815.866 0.253 0.131
Chain 1: 700 -7741.693 0.218 0.017
Chain 1: 800 -7772.547 0.191 0.017
Chain 1: 900 -7882.310 0.171 0.016
Chain 1: 1000 -7814.446 0.155 0.016
Chain 1: 1100 -7859.334 0.056 0.014
Chain 1: 1200 -7773.464 0.023 0.011
Chain 1: 1300 -7730.486 0.010 0.011
Chain 1: 1400 -7733.654 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001429 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.29 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57417.732 1.000 1.000
Chain 1: 200 -17314.557 1.658 2.316
Chain 1: 300 -8512.303 1.450 1.034
Chain 1: 400 -8119.479 1.100 1.034
Chain 1: 500 -8192.832 0.882 1.000
Chain 1: 600 -8994.687 0.749 1.000
Chain 1: 700 -7788.187 0.665 0.155
Chain 1: 800 -7928.745 0.584 0.155
Chain 1: 900 -7876.008 0.520 0.089
Chain 1: 1000 -7997.916 0.469 0.089
Chain 1: 1100 -7619.876 0.374 0.050
Chain 1: 1200 -7632.612 0.143 0.048
Chain 1: 1300 -7586.352 0.040 0.018
Chain 1: 1400 -7719.845 0.037 0.017
Chain 1: 1500 -7569.918 0.038 0.018
Chain 1: 1600 -7491.143 0.030 0.017
Chain 1: 1700 -7480.452 0.015 0.015
Chain 1: 1800 -7515.083 0.013 0.011
Chain 1: 1900 -7568.282 0.013 0.011
Chain 1: 2000 -7570.060 0.012 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003288 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.88 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85994.969 1.000 1.000
Chain 1: 200 -13154.549 3.269 5.537
Chain 1: 300 -9606.217 2.302 1.000
Chain 1: 400 -10413.633 1.746 1.000
Chain 1: 500 -8506.752 1.442 0.369
Chain 1: 600 -8160.988 1.208 0.369
Chain 1: 700 -8196.143 1.036 0.224
Chain 1: 800 -8408.218 0.910 0.224
Chain 1: 900 -8474.283 0.810 0.078
Chain 1: 1000 -8231.968 0.732 0.078
Chain 1: 1100 -8522.338 0.635 0.042 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8173.564 0.086 0.042
Chain 1: 1300 -8182.151 0.049 0.034
Chain 1: 1400 -8266.599 0.042 0.029
Chain 1: 1500 -8217.748 0.020 0.025
Chain 1: 1600 -8222.307 0.016 0.010
Chain 1: 1700 -8156.954 0.016 0.010
Chain 1: 1800 -8037.192 0.015 0.010
Chain 1: 1900 -8154.150 0.016 0.014
Chain 1: 2000 -8114.023 0.014 0.010
Chain 1: 2100 -8248.060 0.012 0.010
Chain 1: 2200 -8036.908 0.010 0.010
Chain 1: 2300 -8177.539 0.012 0.014
Chain 1: 2400 -8189.604 0.011 0.014
Chain 1: 2500 -8157.662 0.011 0.014
Chain 1: 2600 -8154.309 0.011 0.014
Chain 1: 2700 -8063.824 0.011 0.014
Chain 1: 2800 -8041.822 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003323 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.23 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8364797.186 1.000 1.000
Chain 1: 200 -1580686.091 2.646 4.292
Chain 1: 300 -891091.285 2.022 1.000
Chain 1: 400 -458342.854 1.752 1.000
Chain 1: 500 -358752.149 1.458 0.944
Chain 1: 600 -233535.476 1.304 0.944
Chain 1: 700 -119299.799 1.254 0.944
Chain 1: 800 -86391.087 1.145 0.944
Chain 1: 900 -66639.737 1.051 0.774
Chain 1: 1000 -51360.471 0.976 0.774
Chain 1: 1100 -38774.510 0.908 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37937.602 0.481 0.381
Chain 1: 1300 -25842.454 0.450 0.381
Chain 1: 1400 -25554.242 0.357 0.325
Chain 1: 1500 -22128.320 0.345 0.325
Chain 1: 1600 -21339.818 0.295 0.297
Chain 1: 1700 -20208.026 0.205 0.296
Chain 1: 1800 -20150.550 0.167 0.155
Chain 1: 1900 -20476.042 0.139 0.056
Chain 1: 2000 -18985.323 0.117 0.056
Chain 1: 2100 -19223.856 0.086 0.037
Chain 1: 2200 -19450.336 0.085 0.037
Chain 1: 2300 -19067.634 0.040 0.020
Chain 1: 2400 -18839.831 0.040 0.020
Chain 1: 2500 -18642.110 0.026 0.016
Chain 1: 2600 -18272.718 0.024 0.016
Chain 1: 2700 -18229.778 0.019 0.012
Chain 1: 2800 -17946.985 0.020 0.016
Chain 1: 2900 -18228.052 0.020 0.015
Chain 1: 3000 -18214.238 0.012 0.012
Chain 1: 3100 -18299.146 0.011 0.012
Chain 1: 3200 -17990.198 0.012 0.015
Chain 1: 3300 -18194.610 0.011 0.012
Chain 1: 3400 -17670.285 0.013 0.015
Chain 1: 3500 -18281.167 0.015 0.016
Chain 1: 3600 -17589.155 0.017 0.016
Chain 1: 3700 -17975.016 0.019 0.017
Chain 1: 3800 -16936.840 0.023 0.021
Chain 1: 3900 -16933.057 0.022 0.021
Chain 1: 4000 -17050.326 0.023 0.021
Chain 1: 4100 -16964.204 0.023 0.021
Chain 1: 4200 -16780.887 0.022 0.021
Chain 1: 4300 -16918.948 0.022 0.021
Chain 1: 4400 -16876.154 0.019 0.011
Chain 1: 4500 -16778.761 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001287 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.87 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12037.739 1.000 1.000
Chain 1: 200 -8887.954 0.677 1.000
Chain 1: 300 -7995.652 0.489 0.354
Chain 1: 400 -7995.918 0.367 0.354
Chain 1: 500 -7879.373 0.296 0.112
Chain 1: 600 -7817.495 0.248 0.112
Chain 1: 700 -7750.960 0.214 0.015
Chain 1: 800 -7788.675 0.188 0.015
Chain 1: 900 -7915.416 0.169 0.015
Chain 1: 1000 -7807.983 0.153 0.015
Chain 1: 1100 -7779.539 0.054 0.014
Chain 1: 1200 -7764.672 0.018 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001382 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.82 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56395.241 1.000 1.000
Chain 1: 200 -16948.505 1.664 2.327
Chain 1: 300 -8555.237 1.436 1.000
Chain 1: 400 -8751.482 1.083 1.000
Chain 1: 500 -8396.975 0.875 0.981
Chain 1: 600 -8656.960 0.734 0.981
Chain 1: 700 -8122.458 0.638 0.066
Chain 1: 800 -8136.870 0.559 0.066
Chain 1: 900 -7836.653 0.501 0.042
Chain 1: 1000 -7977.945 0.453 0.042
Chain 1: 1100 -7654.547 0.357 0.042
Chain 1: 1200 -7740.365 0.125 0.038
Chain 1: 1300 -7656.270 0.028 0.030
Chain 1: 1400 -7837.226 0.028 0.030
Chain 1: 1500 -7637.285 0.027 0.026
Chain 1: 1600 -7556.578 0.025 0.023
Chain 1: 1700 -7524.803 0.019 0.018
Chain 1: 1800 -7555.553 0.019 0.018
Chain 1: 1900 -7634.302 0.016 0.011
Chain 1: 2000 -7621.278 0.014 0.011
Chain 1: 2100 -7634.203 0.010 0.011
Chain 1: 2200 -7695.659 0.010 0.010
Chain 1: 2300 -7600.761 0.010 0.010
Chain 1: 2400 -7644.605 0.009 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00334 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.4 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85717.098 1.000 1.000
Chain 1: 200 -13110.039 3.269 5.538
Chain 1: 300 -9591.947 2.302 1.000
Chain 1: 400 -10443.592 1.747 1.000
Chain 1: 500 -8493.242 1.443 0.367
Chain 1: 600 -8165.364 1.209 0.367
Chain 1: 700 -8155.214 1.037 0.230
Chain 1: 800 -8458.149 0.912 0.230
Chain 1: 900 -8473.354 0.811 0.082
Chain 1: 1000 -8198.957 0.733 0.082
Chain 1: 1100 -8441.554 0.636 0.040 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8157.895 0.085 0.036
Chain 1: 1300 -8348.682 0.051 0.035
Chain 1: 1400 -8279.006 0.044 0.033
Chain 1: 1500 -8225.121 0.021 0.029
Chain 1: 1600 -8221.968 0.017 0.023
Chain 1: 1700 -8159.053 0.018 0.023
Chain 1: 1800 -8038.964 0.016 0.015
Chain 1: 1900 -8152.954 0.017 0.015
Chain 1: 2000 -8114.283 0.014 0.014
Chain 1: 2100 -8253.132 0.013 0.014
Chain 1: 2200 -8039.115 0.012 0.014
Chain 1: 2300 -8180.766 0.012 0.014
Chain 1: 2400 -8187.470 0.011 0.014
Chain 1: 2500 -8156.984 0.011 0.014
Chain 1: 2600 -8151.141 0.011 0.014
Chain 1: 2700 -8062.078 0.011 0.014
Chain 1: 2800 -8043.764 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003498 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.98 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8408263.217 1.000 1.000
Chain 1: 200 -1585919.169 2.651 4.302
Chain 1: 300 -890293.316 2.028 1.000
Chain 1: 400 -456907.344 1.758 1.000
Chain 1: 500 -356888.537 1.462 0.949
Chain 1: 600 -232063.010 1.308 0.949
Chain 1: 700 -118545.643 1.258 0.949
Chain 1: 800 -85827.387 1.149 0.949
Chain 1: 900 -66215.673 1.054 0.781
Chain 1: 1000 -51046.917 0.978 0.781
Chain 1: 1100 -38559.512 0.911 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37734.395 0.483 0.381
Chain 1: 1300 -25735.614 0.451 0.381
Chain 1: 1400 -25456.394 0.357 0.324
Chain 1: 1500 -22055.894 0.345 0.324
Chain 1: 1600 -21275.100 0.295 0.297
Chain 1: 1700 -20154.840 0.204 0.296
Chain 1: 1800 -20100.029 0.167 0.154
Chain 1: 1900 -20425.492 0.139 0.056
Chain 1: 2000 -18941.130 0.117 0.056
Chain 1: 2100 -19179.186 0.085 0.037
Chain 1: 2200 -19404.685 0.084 0.037
Chain 1: 2300 -19022.947 0.040 0.020
Chain 1: 2400 -18795.351 0.040 0.020
Chain 1: 2500 -18597.240 0.026 0.016
Chain 1: 2600 -18228.278 0.024 0.016
Chain 1: 2700 -18185.567 0.019 0.012
Chain 1: 2800 -17902.673 0.020 0.016
Chain 1: 2900 -18183.535 0.020 0.015
Chain 1: 3000 -18169.832 0.012 0.012
Chain 1: 3100 -18254.666 0.011 0.012
Chain 1: 3200 -17945.887 0.012 0.015
Chain 1: 3300 -18150.209 0.011 0.012
Chain 1: 3400 -17626.011 0.013 0.015
Chain 1: 3500 -18236.491 0.015 0.016
Chain 1: 3600 -17545.028 0.017 0.016
Chain 1: 3700 -17930.398 0.019 0.017
Chain 1: 3800 -16892.918 0.023 0.021
Chain 1: 3900 -16889.126 0.022 0.021
Chain 1: 4000 -17006.446 0.023 0.021
Chain 1: 4100 -16920.306 0.023 0.021
Chain 1: 4200 -16737.207 0.022 0.021
Chain 1: 4300 -16875.155 0.022 0.021
Chain 1: 4400 -16832.474 0.019 0.011
Chain 1: 4500 -16735.098 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001387 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.87 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13061.639 1.000 1.000
Chain 1: 200 -9896.399 0.660 1.000
Chain 1: 300 -8575.442 0.491 0.320
Chain 1: 400 -8826.647 0.376 0.320
Chain 1: 500 -8666.164 0.304 0.154
Chain 1: 600 -8524.531 0.256 0.154
Chain 1: 700 -8422.194 0.221 0.028
Chain 1: 800 -8425.224 0.194 0.028
Chain 1: 900 -8361.793 0.173 0.019
Chain 1: 1000 -8549.966 0.158 0.022
Chain 1: 1100 -8566.934 0.058 0.019
Chain 1: 1200 -8440.364 0.028 0.017
Chain 1: 1300 -8407.946 0.013 0.015
Chain 1: 1400 -8416.377 0.010 0.012 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001377 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.77 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -46700.654 1.000 1.000
Chain 1: 200 -16160.945 1.445 1.890
Chain 1: 300 -9015.107 1.227 1.000
Chain 1: 400 -8089.119 0.949 1.000
Chain 1: 500 -8766.804 0.775 0.793
Chain 1: 600 -8930.516 0.649 0.793
Chain 1: 700 -8153.700 0.570 0.114
Chain 1: 800 -8380.230 0.502 0.114
Chain 1: 900 -7985.515 0.452 0.095
Chain 1: 1000 -7786.116 0.409 0.095
Chain 1: 1100 -7783.220 0.309 0.077
Chain 1: 1200 -7768.294 0.120 0.049
Chain 1: 1300 -7760.910 0.041 0.027
Chain 1: 1400 -7673.534 0.031 0.026
Chain 1: 1500 -7554.261 0.025 0.018
Chain 1: 1600 -7745.253 0.025 0.025
Chain 1: 1700 -7619.000 0.017 0.017
Chain 1: 1800 -7716.601 0.016 0.016
Chain 1: 1900 -7547.644 0.013 0.016
Chain 1: 2000 -7668.072 0.012 0.016
Chain 1: 2100 -7636.613 0.013 0.016
Chain 1: 2200 -7767.365 0.014 0.016
Chain 1: 2300 -7512.478 0.017 0.017
Chain 1: 2400 -7709.563 0.019 0.017
Chain 1: 2500 -7495.042 0.020 0.022
Chain 1: 2600 -7538.072 0.018 0.017
Chain 1: 2700 -7513.276 0.017 0.017
Chain 1: 2800 -7520.424 0.016 0.017
Chain 1: 2900 -7355.357 0.016 0.017
Chain 1: 3000 -7512.010 0.016 0.021
Chain 1: 3100 -7510.443 0.016 0.021
Chain 1: 3200 -7713.094 0.017 0.022
Chain 1: 3300 -7391.079 0.018 0.022
Chain 1: 3400 -7629.370 0.018 0.022
Chain 1: 3500 -7418.948 0.018 0.022
Chain 1: 3600 -7489.568 0.019 0.022
Chain 1: 3700 -7445.638 0.019 0.022
Chain 1: 3800 -7416.316 0.019 0.022
Chain 1: 3900 -7389.175 0.017 0.021
Chain 1: 4000 -7384.545 0.015 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003251 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.51 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87333.370 1.000 1.000
Chain 1: 200 -14199.856 3.075 5.150
Chain 1: 300 -10488.405 2.168 1.000
Chain 1: 400 -11688.019 1.652 1.000
Chain 1: 500 -9493.730 1.368 0.354
Chain 1: 600 -9166.182 1.146 0.354
Chain 1: 700 -9087.770 0.983 0.231
Chain 1: 800 -9438.994 0.865 0.231
Chain 1: 900 -9258.122 0.771 0.103
Chain 1: 1000 -9192.848 0.695 0.103
Chain 1: 1100 -9229.493 0.595 0.037 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8766.785 0.085 0.037
Chain 1: 1300 -9302.706 0.056 0.037
Chain 1: 1400 -9101.600 0.048 0.036
Chain 1: 1500 -9019.289 0.025 0.022
Chain 1: 1600 -9129.693 0.023 0.020
Chain 1: 1700 -9196.642 0.023 0.020
Chain 1: 1800 -8761.798 0.024 0.020
Chain 1: 1900 -8866.256 0.023 0.012
Chain 1: 2000 -8841.829 0.023 0.012
Chain 1: 2100 -8960.370 0.024 0.013
Chain 1: 2200 -8772.367 0.021 0.013
Chain 1: 2300 -8929.092 0.017 0.013
Chain 1: 2400 -8772.521 0.016 0.013
Chain 1: 2500 -8841.107 0.016 0.013
Chain 1: 2600 -8754.789 0.016 0.013
Chain 1: 2700 -8786.773 0.016 0.013
Chain 1: 2800 -8748.103 0.011 0.012
Chain 1: 2900 -8839.821 0.011 0.010
Chain 1: 3000 -8665.109 0.013 0.013
Chain 1: 3100 -8829.787 0.013 0.018
Chain 1: 3200 -8702.932 0.012 0.015
Chain 1: 3300 -8712.337 0.011 0.010
Chain 1: 3400 -8863.259 0.011 0.010
Chain 1: 3500 -8851.266 0.010 0.010
Chain 1: 3600 -8660.370 0.011 0.015
Chain 1: 3700 -8803.028 0.013 0.016
Chain 1: 3800 -8667.250 0.014 0.016
Chain 1: 3900 -8602.573 0.013 0.016
Chain 1: 4000 -8676.920 0.012 0.016
Chain 1: 4100 -8667.726 0.011 0.015
Chain 1: 4200 -8656.182 0.009 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002958 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.58 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8387271.673 1.000 1.000
Chain 1: 200 -1582519.669 2.650 4.300
Chain 1: 300 -891669.640 2.025 1.000
Chain 1: 400 -458599.959 1.755 1.000
Chain 1: 500 -359150.072 1.459 0.944
Chain 1: 600 -234021.095 1.305 0.944
Chain 1: 700 -120097.032 1.254 0.944
Chain 1: 800 -87261.047 1.144 0.944
Chain 1: 900 -67579.363 1.050 0.775
Chain 1: 1000 -52360.185 0.974 0.775
Chain 1: 1100 -39816.175 0.905 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38993.603 0.477 0.376
Chain 1: 1300 -26921.334 0.445 0.376
Chain 1: 1400 -26639.220 0.351 0.315
Chain 1: 1500 -23218.352 0.338 0.315
Chain 1: 1600 -22432.803 0.288 0.291
Chain 1: 1700 -21302.672 0.199 0.291
Chain 1: 1800 -21246.183 0.162 0.147
Chain 1: 1900 -21572.664 0.134 0.053
Chain 1: 2000 -20081.044 0.112 0.053
Chain 1: 2100 -20319.671 0.082 0.035
Chain 1: 2200 -20546.657 0.081 0.035
Chain 1: 2300 -20163.270 0.038 0.019
Chain 1: 2400 -19935.197 0.038 0.019
Chain 1: 2500 -19737.281 0.024 0.015
Chain 1: 2600 -19367.090 0.023 0.015
Chain 1: 2700 -19323.931 0.018 0.012
Chain 1: 2800 -19040.692 0.019 0.015
Chain 1: 2900 -19322.090 0.019 0.015
Chain 1: 3000 -19308.270 0.011 0.012
Chain 1: 3100 -19393.326 0.011 0.011
Chain 1: 3200 -19083.774 0.011 0.015
Chain 1: 3300 -19288.659 0.010 0.011
Chain 1: 3400 -18763.210 0.012 0.015
Chain 1: 3500 -19375.741 0.014 0.015
Chain 1: 3600 -18681.557 0.016 0.015
Chain 1: 3700 -19069.012 0.018 0.016
Chain 1: 3800 -18027.450 0.022 0.020
Chain 1: 3900 -18023.566 0.021 0.020
Chain 1: 4000 -18140.847 0.021 0.020
Chain 1: 4100 -18054.575 0.021 0.020
Chain 1: 4200 -17870.525 0.021 0.020
Chain 1: 4300 -18009.131 0.020 0.020
Chain 1: 4400 -17965.728 0.018 0.010
Chain 1: 4500 -17868.213 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001539 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.39 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49138.862 1.000 1.000
Chain 1: 200 -24228.724 1.014 1.028
Chain 1: 300 -14038.129 0.918 1.000
Chain 1: 400 -22063.091 0.779 1.000
Chain 1: 500 -18429.159 0.663 0.726
Chain 1: 600 -14819.462 0.593 0.726
Chain 1: 700 -13607.349 0.521 0.364
Chain 1: 800 -13763.295 0.457 0.364
Chain 1: 900 -11174.705 0.432 0.244
Chain 1: 1000 -11831.324 0.395 0.244
Chain 1: 1100 -15570.157 0.319 0.240
Chain 1: 1200 -17739.993 0.228 0.232
Chain 1: 1300 -12330.655 0.199 0.232
Chain 1: 1400 -10976.867 0.175 0.197
Chain 1: 1500 -11761.773 0.162 0.123
Chain 1: 1600 -11064.231 0.144 0.122
Chain 1: 1700 -12981.270 0.150 0.123
Chain 1: 1800 -10551.184 0.172 0.148
Chain 1: 1900 -11962.743 0.161 0.123
Chain 1: 2000 -13588.109 0.167 0.123
Chain 1: 2100 -13334.770 0.145 0.122
Chain 1: 2200 -10387.024 0.161 0.123
Chain 1: 2300 -16628.854 0.155 0.123
Chain 1: 2400 -9293.831 0.221 0.148
Chain 1: 2500 -11287.358 0.232 0.177
Chain 1: 2600 -9950.463 0.239 0.177
Chain 1: 2700 -14631.522 0.257 0.230
Chain 1: 2800 -9537.407 0.287 0.284
Chain 1: 2900 -9931.663 0.279 0.284
Chain 1: 3000 -9591.733 0.271 0.284
Chain 1: 3100 -8903.499 0.277 0.284
Chain 1: 3200 -15388.763 0.290 0.320
Chain 1: 3300 -9268.592 0.319 0.320
Chain 1: 3400 -13600.470 0.272 0.319
Chain 1: 3500 -9273.481 0.301 0.320
Chain 1: 3600 -9674.862 0.291 0.320
Chain 1: 3700 -10263.360 0.265 0.319
Chain 1: 3800 -14644.789 0.242 0.299
Chain 1: 3900 -10162.800 0.282 0.319
Chain 1: 4000 -11554.452 0.290 0.319
Chain 1: 4100 -11774.163 0.284 0.319
Chain 1: 4200 -11375.440 0.246 0.299
Chain 1: 4300 -9698.656 0.197 0.173
Chain 1: 4400 -11280.274 0.179 0.140
Chain 1: 4500 -9221.966 0.155 0.140
Chain 1: 4600 -13923.535 0.185 0.173
Chain 1: 4700 -9019.851 0.233 0.223
Chain 1: 4800 -8738.018 0.207 0.173
Chain 1: 4900 -8836.486 0.164 0.140
Chain 1: 5000 -9740.568 0.161 0.140
Chain 1: 5100 -12952.248 0.184 0.173
Chain 1: 5200 -16440.550 0.201 0.212
Chain 1: 5300 -13205.770 0.209 0.223
Chain 1: 5400 -11202.560 0.212 0.223
Chain 1: 5500 -14076.870 0.211 0.212
Chain 1: 5600 -15212.392 0.184 0.204
Chain 1: 5700 -13581.319 0.142 0.179
Chain 1: 5800 -9042.783 0.189 0.204
Chain 1: 5900 -9522.519 0.193 0.204
Chain 1: 6000 -11935.772 0.204 0.204
Chain 1: 6100 -9881.277 0.200 0.204
Chain 1: 6200 -8342.820 0.197 0.202
Chain 1: 6300 -8950.009 0.179 0.184
Chain 1: 6400 -13052.601 0.193 0.202
Chain 1: 6500 -8940.765 0.218 0.202
Chain 1: 6600 -9695.421 0.219 0.202
Chain 1: 6700 -8965.058 0.215 0.202
Chain 1: 6800 -9700.127 0.172 0.184
Chain 1: 6900 -11408.177 0.182 0.184
Chain 1: 7000 -8822.440 0.191 0.184
Chain 1: 7100 -14342.274 0.209 0.184
Chain 1: 7200 -11250.281 0.218 0.275
Chain 1: 7300 -8396.630 0.245 0.293
Chain 1: 7400 -11061.452 0.238 0.275
Chain 1: 7500 -8534.916 0.221 0.275
Chain 1: 7600 -8698.357 0.216 0.275
Chain 1: 7700 -8312.278 0.212 0.275
Chain 1: 7800 -10581.609 0.226 0.275
Chain 1: 7900 -8291.457 0.239 0.276
Chain 1: 8000 -8480.345 0.211 0.275
Chain 1: 8100 -10654.978 0.193 0.241
Chain 1: 8200 -9653.310 0.176 0.214
Chain 1: 8300 -9515.144 0.144 0.204
Chain 1: 8400 -8381.280 0.133 0.135
Chain 1: 8500 -8375.399 0.104 0.104
Chain 1: 8600 -12013.280 0.132 0.135
Chain 1: 8700 -8922.554 0.162 0.204
Chain 1: 8800 -8768.384 0.142 0.135
Chain 1: 8900 -9950.033 0.127 0.119
Chain 1: 9000 -8751.803 0.138 0.135
Chain 1: 9100 -9919.955 0.129 0.119
Chain 1: 9200 -8846.822 0.131 0.121
Chain 1: 9300 -9308.384 0.135 0.121
Chain 1: 9400 -12051.037 0.144 0.121
Chain 1: 9500 -8851.562 0.180 0.137
Chain 1: 9600 -9731.785 0.159 0.121
Chain 1: 9700 -9136.571 0.131 0.119
Chain 1: 9800 -11541.983 0.150 0.121
Chain 1: 9900 -8679.166 0.171 0.137
Chain 1: 10000 -8680.917 0.157 0.121
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001425 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.25 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -47528.667 1.000 1.000
Chain 1: 200 -15880.744 1.496 1.993
Chain 1: 300 -8607.446 1.279 1.000
Chain 1: 400 -8480.948 0.963 1.000
Chain 1: 500 -8681.489 0.775 0.845
Chain 1: 600 -8268.679 0.654 0.845
Chain 1: 700 -7744.604 0.570 0.068
Chain 1: 800 -8212.891 0.506 0.068
Chain 1: 900 -8078.741 0.452 0.057
Chain 1: 1000 -7711.876 0.411 0.057
Chain 1: 1100 -7753.888 0.312 0.050
Chain 1: 1200 -7567.834 0.115 0.048
Chain 1: 1300 -7758.203 0.033 0.025
Chain 1: 1400 -7888.912 0.033 0.025
Chain 1: 1500 -7577.446 0.035 0.041
Chain 1: 1600 -7790.173 0.033 0.027
Chain 1: 1700 -7505.815 0.030 0.027
Chain 1: 1800 -7569.725 0.025 0.025
Chain 1: 1900 -7591.064 0.024 0.025
Chain 1: 2000 -7632.030 0.019 0.025
Chain 1: 2100 -7582.990 0.020 0.025
Chain 1: 2200 -7696.160 0.019 0.017
Chain 1: 2300 -7550.478 0.018 0.017
Chain 1: 2400 -7625.646 0.017 0.015
Chain 1: 2500 -7606.363 0.013 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003296 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.96 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86764.697 1.000 1.000
Chain 1: 200 -13577.414 3.195 5.390
Chain 1: 300 -9960.895 2.251 1.000
Chain 1: 400 -10667.751 1.705 1.000
Chain 1: 500 -8944.528 1.402 0.363
Chain 1: 600 -8483.418 1.178 0.363
Chain 1: 700 -8588.177 1.011 0.193
Chain 1: 800 -9258.265 0.894 0.193
Chain 1: 900 -8842.355 0.800 0.072
Chain 1: 1000 -8513.780 0.724 0.072
Chain 1: 1100 -8823.319 0.627 0.066 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8328.131 0.094 0.059
Chain 1: 1300 -8664.701 0.062 0.054
Chain 1: 1400 -8665.275 0.055 0.047
Chain 1: 1500 -8543.879 0.037 0.039
Chain 1: 1600 -8650.554 0.033 0.039
Chain 1: 1700 -8735.057 0.033 0.039
Chain 1: 1800 -8326.762 0.030 0.039
Chain 1: 1900 -8423.137 0.027 0.035
Chain 1: 2000 -8395.499 0.023 0.014
Chain 1: 2100 -8516.542 0.021 0.014
Chain 1: 2200 -8429.255 0.016 0.012
Chain 1: 2300 -8464.468 0.013 0.011
Chain 1: 2400 -8358.667 0.014 0.012
Chain 1: 2500 -8399.905 0.013 0.011
Chain 1: 2600 -8425.103 0.012 0.010
Chain 1: 2700 -8343.260 0.012 0.010
Chain 1: 2800 -8311.715 0.008 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00298 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.8 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8411581.475 1.000 1.000
Chain 1: 200 -1586873.580 2.650 4.301
Chain 1: 300 -891073.837 2.027 1.000
Chain 1: 400 -457275.707 1.758 1.000
Chain 1: 500 -357214.998 1.462 0.949
Chain 1: 600 -232398.945 1.308 0.949
Chain 1: 700 -119001.632 1.257 0.949
Chain 1: 800 -86254.578 1.147 0.949
Chain 1: 900 -66676.510 1.053 0.781
Chain 1: 1000 -51521.492 0.977 0.781
Chain 1: 1100 -39040.451 0.909 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38224.107 0.481 0.380
Chain 1: 1300 -26231.587 0.448 0.380
Chain 1: 1400 -25955.174 0.355 0.320
Chain 1: 1500 -22554.484 0.342 0.320
Chain 1: 1600 -21774.167 0.292 0.294
Chain 1: 1700 -20654.523 0.202 0.294
Chain 1: 1800 -20600.133 0.164 0.151
Chain 1: 1900 -20926.114 0.136 0.054
Chain 1: 2000 -19440.731 0.114 0.054
Chain 1: 2100 -19679.140 0.084 0.036
Chain 1: 2200 -19904.737 0.083 0.036
Chain 1: 2300 -19522.731 0.039 0.020
Chain 1: 2400 -19294.928 0.039 0.020
Chain 1: 2500 -19096.534 0.025 0.016
Chain 1: 2600 -18727.300 0.023 0.016
Chain 1: 2700 -18684.466 0.018 0.012
Chain 1: 2800 -18401.163 0.019 0.015
Chain 1: 2900 -18682.334 0.019 0.015
Chain 1: 3000 -18668.664 0.012 0.012
Chain 1: 3100 -18753.545 0.011 0.012
Chain 1: 3200 -18444.442 0.012 0.015
Chain 1: 3300 -18649.026 0.011 0.012
Chain 1: 3400 -18124.114 0.012 0.015
Chain 1: 3500 -18735.558 0.015 0.015
Chain 1: 3600 -18042.867 0.017 0.015
Chain 1: 3700 -18429.129 0.018 0.017
Chain 1: 3800 -17389.621 0.023 0.021
Chain 1: 3900 -17385.731 0.021 0.021
Chain 1: 4000 -17503.120 0.022 0.021
Chain 1: 4100 -17416.813 0.022 0.021
Chain 1: 4200 -17233.286 0.021 0.021
Chain 1: 4300 -17371.595 0.021 0.021
Chain 1: 4400 -17328.582 0.018 0.011
Chain 1: 4500 -17231.072 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001259 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.59 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12338.769 1.000 1.000
Chain 1: 200 -9331.869 0.661 1.000
Chain 1: 300 -8041.249 0.494 0.322
Chain 1: 400 -8266.668 0.377 0.322
Chain 1: 500 -8149.703 0.305 0.161
Chain 1: 600 -8001.142 0.257 0.161
Chain 1: 700 -7911.871 0.222 0.027
Chain 1: 800 -7920.880 0.194 0.027
Chain 1: 900 -7845.395 0.174 0.019
Chain 1: 1000 -8023.088 0.159 0.022
Chain 1: 1100 -8050.776 0.059 0.019
Chain 1: 1200 -7946.557 0.028 0.014
Chain 1: 1300 -7892.180 0.013 0.013
Chain 1: 1400 -7909.411 0.010 0.011
Chain 1: 1500 -7995.969 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001457 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.57 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -55595.002 1.000 1.000
Chain 1: 200 -17217.744 1.614 2.229
Chain 1: 300 -8687.570 1.404 1.000
Chain 1: 400 -8484.448 1.059 1.000
Chain 1: 500 -8043.087 0.858 0.982
Chain 1: 600 -8430.834 0.723 0.982
Chain 1: 700 -8042.729 0.626 0.055
Chain 1: 800 -8176.253 0.550 0.055
Chain 1: 900 -8391.552 0.492 0.048
Chain 1: 1000 -8030.605 0.447 0.048
Chain 1: 1100 -7758.120 0.351 0.046
Chain 1: 1200 -7703.171 0.128 0.045
Chain 1: 1300 -7812.676 0.032 0.035
Chain 1: 1400 -7784.930 0.030 0.035
Chain 1: 1500 -7570.517 0.027 0.028
Chain 1: 1600 -7541.401 0.023 0.026
Chain 1: 1700 -7570.984 0.018 0.016
Chain 1: 1800 -7627.556 0.017 0.014
Chain 1: 1900 -7633.259 0.015 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003036 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.36 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86387.767 1.000 1.000
Chain 1: 200 -13485.963 3.203 5.406
Chain 1: 300 -9860.020 2.258 1.000
Chain 1: 400 -10687.477 1.713 1.000
Chain 1: 500 -8825.918 1.412 0.368
Chain 1: 600 -8428.027 1.185 0.368
Chain 1: 700 -8299.042 1.018 0.211
Chain 1: 800 -9198.887 0.903 0.211
Chain 1: 900 -8645.272 0.810 0.098
Chain 1: 1000 -8510.759 0.730 0.098
Chain 1: 1100 -8569.965 0.631 0.077 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8377.279 0.093 0.064
Chain 1: 1300 -8561.718 0.058 0.047
Chain 1: 1400 -8574.767 0.050 0.023
Chain 1: 1500 -8438.962 0.031 0.022
Chain 1: 1600 -8550.416 0.028 0.016
Chain 1: 1700 -8636.593 0.027 0.016
Chain 1: 1800 -8228.242 0.022 0.016
Chain 1: 1900 -8324.175 0.017 0.016
Chain 1: 2000 -8296.689 0.016 0.013
Chain 1: 2100 -8417.920 0.016 0.014
Chain 1: 2200 -8248.718 0.016 0.014
Chain 1: 2300 -8323.628 0.015 0.013
Chain 1: 2400 -8388.624 0.016 0.013
Chain 1: 2500 -8334.304 0.015 0.012
Chain 1: 2600 -8332.895 0.013 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003307 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.07 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8387595.197 1.000 1.000
Chain 1: 200 -1583318.257 2.649 4.297
Chain 1: 300 -891587.761 2.024 1.000
Chain 1: 400 -458106.642 1.755 1.000
Chain 1: 500 -358579.438 1.459 0.946
Chain 1: 600 -233413.675 1.306 0.946
Chain 1: 700 -119434.249 1.255 0.946
Chain 1: 800 -86568.492 1.146 0.946
Chain 1: 900 -66874.695 1.051 0.776
Chain 1: 1000 -51637.462 0.976 0.776
Chain 1: 1100 -39082.986 0.908 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38255.960 0.480 0.380
Chain 1: 1300 -26187.799 0.449 0.380
Chain 1: 1400 -25903.553 0.355 0.321
Chain 1: 1500 -22484.427 0.343 0.321
Chain 1: 1600 -21698.547 0.293 0.295
Chain 1: 1700 -20569.920 0.203 0.294
Chain 1: 1800 -20513.356 0.165 0.152
Chain 1: 1900 -20839.349 0.137 0.055
Chain 1: 2000 -19349.469 0.115 0.055
Chain 1: 2100 -19587.913 0.084 0.036
Chain 1: 2200 -19814.421 0.083 0.036
Chain 1: 2300 -19431.638 0.039 0.020
Chain 1: 2400 -19203.779 0.039 0.020
Chain 1: 2500 -19005.809 0.025 0.016
Chain 1: 2600 -18636.104 0.024 0.016
Chain 1: 2700 -18593.137 0.018 0.012
Chain 1: 2800 -18310.039 0.020 0.015
Chain 1: 2900 -18591.310 0.020 0.015
Chain 1: 3000 -18577.480 0.012 0.012
Chain 1: 3100 -18662.416 0.011 0.012
Chain 1: 3200 -18353.198 0.012 0.015
Chain 1: 3300 -18557.871 0.011 0.012
Chain 1: 3400 -18032.943 0.013 0.015
Chain 1: 3500 -18644.562 0.015 0.015
Chain 1: 3600 -17951.674 0.017 0.015
Chain 1: 3700 -18338.165 0.019 0.017
Chain 1: 3800 -17298.442 0.023 0.021
Chain 1: 3900 -17294.622 0.022 0.021
Chain 1: 4000 -17411.922 0.022 0.021
Chain 1: 4100 -17325.677 0.022 0.021
Chain 1: 4200 -17142.081 0.022 0.021
Chain 1: 4300 -17280.356 0.021 0.021
Chain 1: 4400 -17237.302 0.019 0.011
Chain 1: 4500 -17139.874 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001321 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.21 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12310.729 1.000 1.000
Chain 1: 200 -9211.430 0.668 1.000
Chain 1: 300 -7898.690 0.501 0.336
Chain 1: 400 -8071.451 0.381 0.336
Chain 1: 500 -7931.545 0.308 0.166
Chain 1: 600 -7854.643 0.259 0.166
Chain 1: 700 -7758.231 0.223 0.021
Chain 1: 800 -7781.775 0.196 0.021
Chain 1: 900 -7869.112 0.175 0.018
Chain 1: 1000 -7809.901 0.159 0.018
Chain 1: 1100 -7865.252 0.059 0.012
Chain 1: 1200 -7766.743 0.027 0.012
Chain 1: 1300 -7726.428 0.011 0.011
Chain 1: 1400 -7754.204 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00141 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.1 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56657.995 1.000 1.000
Chain 1: 200 -17312.776 1.636 2.273
Chain 1: 300 -8628.674 1.426 1.006
Chain 1: 400 -8295.248 1.080 1.006
Chain 1: 500 -8170.083 0.867 1.000
Chain 1: 600 -8837.783 0.735 1.000
Chain 1: 700 -8131.194 0.642 0.087
Chain 1: 800 -8045.750 0.563 0.087
Chain 1: 900 -7903.750 0.503 0.076
Chain 1: 1000 -7728.476 0.455 0.076
Chain 1: 1100 -7677.526 0.355 0.040
Chain 1: 1200 -7613.853 0.129 0.023
Chain 1: 1300 -7653.834 0.029 0.018
Chain 1: 1400 -7560.489 0.026 0.015
Chain 1: 1500 -7518.427 0.025 0.012
Chain 1: 1600 -7679.168 0.020 0.012
Chain 1: 1700 -7428.911 0.014 0.012
Chain 1: 1800 -7512.296 0.014 0.012
Chain 1: 1900 -7474.700 0.013 0.011
Chain 1: 2000 -7516.813 0.011 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002556 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.56 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86471.296 1.000 1.000
Chain 1: 200 -13370.456 3.234 5.467
Chain 1: 300 -9735.946 2.280 1.000
Chain 1: 400 -10564.600 1.730 1.000
Chain 1: 500 -8710.332 1.426 0.373
Chain 1: 600 -8221.162 1.199 0.373
Chain 1: 700 -8345.913 1.029 0.213
Chain 1: 800 -8988.497 0.910 0.213
Chain 1: 900 -8536.294 0.815 0.078
Chain 1: 1000 -8350.550 0.735 0.078
Chain 1: 1100 -8607.703 0.638 0.071 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8199.041 0.097 0.060
Chain 1: 1300 -8349.969 0.061 0.053
Chain 1: 1400 -8408.471 0.054 0.050
Chain 1: 1500 -8310.609 0.034 0.030
Chain 1: 1600 -8417.098 0.029 0.022
Chain 1: 1700 -8502.957 0.029 0.022
Chain 1: 1800 -8090.576 0.027 0.022
Chain 1: 1900 -8186.781 0.022 0.018
Chain 1: 2000 -8159.729 0.021 0.013
Chain 1: 2100 -8282.221 0.019 0.013
Chain 1: 2200 -8101.977 0.016 0.013
Chain 1: 2300 -8181.758 0.015 0.012
Chain 1: 2400 -8251.332 0.016 0.012
Chain 1: 2500 -8196.643 0.015 0.012
Chain 1: 2600 -8195.922 0.014 0.010
Chain 1: 2700 -8113.252 0.014 0.010
Chain 1: 2800 -8077.176 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002601 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 26.01 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8400039.469 1.000 1.000
Chain 1: 200 -1584709.217 2.650 4.301
Chain 1: 300 -891113.543 2.026 1.000
Chain 1: 400 -457876.109 1.756 1.000
Chain 1: 500 -357867.715 1.461 0.946
Chain 1: 600 -232935.010 1.307 0.946
Chain 1: 700 -119121.802 1.257 0.946
Chain 1: 800 -86310.463 1.147 0.946
Chain 1: 900 -66657.664 1.052 0.778
Chain 1: 1000 -51454.024 0.977 0.778
Chain 1: 1100 -38928.870 0.909 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38107.035 0.481 0.380
Chain 1: 1300 -26063.084 0.449 0.380
Chain 1: 1400 -25782.929 0.356 0.322
Chain 1: 1500 -22369.296 0.343 0.322
Chain 1: 1600 -21585.462 0.293 0.295
Chain 1: 1700 -20459.154 0.203 0.295
Chain 1: 1800 -20403.298 0.165 0.153
Chain 1: 1900 -20729.406 0.137 0.055
Chain 1: 2000 -19240.457 0.116 0.055
Chain 1: 2100 -19479.016 0.085 0.036
Chain 1: 2200 -19705.324 0.084 0.036
Chain 1: 2300 -19322.635 0.039 0.020
Chain 1: 2400 -19094.702 0.040 0.020
Chain 1: 2500 -18896.668 0.025 0.016
Chain 1: 2600 -18526.974 0.024 0.016
Chain 1: 2700 -18483.993 0.018 0.012
Chain 1: 2800 -18200.806 0.020 0.016
Chain 1: 2900 -18482.070 0.020 0.015
Chain 1: 3000 -18468.341 0.012 0.012
Chain 1: 3100 -18553.271 0.011 0.012
Chain 1: 3200 -18244.004 0.012 0.015
Chain 1: 3300 -18448.699 0.011 0.012
Chain 1: 3400 -17923.670 0.013 0.015
Chain 1: 3500 -18535.429 0.015 0.016
Chain 1: 3600 -17842.301 0.017 0.016
Chain 1: 3700 -18228.935 0.019 0.017
Chain 1: 3800 -17188.871 0.023 0.021
Chain 1: 3900 -17184.999 0.022 0.021
Chain 1: 4000 -17302.342 0.022 0.021
Chain 1: 4100 -17216.054 0.022 0.021
Chain 1: 4200 -17032.373 0.022 0.021
Chain 1: 4300 -17170.739 0.021 0.021
Chain 1: 4400 -17127.616 0.019 0.011
Chain 1: 4500 -17030.124 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001242 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.42 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -11840.419 1.000 1.000
Chain 1: 200 -8854.961 0.669 1.000
Chain 1: 300 -7854.968 0.488 0.337
Chain 1: 400 -7969.069 0.370 0.337
Chain 1: 500 -7766.096 0.301 0.127
Chain 1: 600 -7704.831 0.252 0.127
Chain 1: 700 -7650.327 0.217 0.026
Chain 1: 800 -7629.952 0.190 0.026
Chain 1: 900 -7579.895 0.170 0.014
Chain 1: 1000 -7695.606 0.154 0.015
Chain 1: 1100 -7742.580 0.055 0.014
Chain 1: 1200 -7657.067 0.022 0.011
Chain 1: 1300 -7618.558 0.010 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001481 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.81 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56381.710 1.000 1.000
Chain 1: 200 -16839.666 1.674 2.348
Chain 1: 300 -8480.429 1.445 1.000
Chain 1: 400 -8590.159 1.087 1.000
Chain 1: 500 -8055.303 0.883 0.986
Chain 1: 600 -8902.148 0.751 0.986
Chain 1: 700 -8117.703 0.658 0.097
Chain 1: 800 -8028.540 0.577 0.097
Chain 1: 900 -7900.055 0.515 0.095
Chain 1: 1000 -7683.883 0.466 0.095
Chain 1: 1100 -7647.335 0.367 0.066
Chain 1: 1200 -7804.267 0.134 0.028
Chain 1: 1300 -7710.911 0.036 0.020
Chain 1: 1400 -7778.476 0.036 0.020
Chain 1: 1500 -7608.181 0.032 0.020
Chain 1: 1600 -7523.289 0.023 0.016
Chain 1: 1700 -7494.302 0.014 0.012
Chain 1: 1800 -7563.873 0.014 0.012
Chain 1: 1900 -7532.017 0.012 0.011
Chain 1: 2000 -7586.466 0.010 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003698 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 36.98 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85876.823 1.000 1.000
Chain 1: 200 -12914.680 3.325 5.650
Chain 1: 300 -9437.175 2.339 1.000
Chain 1: 400 -10189.454 1.773 1.000
Chain 1: 500 -8282.288 1.464 0.368
Chain 1: 600 -8059.372 1.225 0.368
Chain 1: 700 -8349.626 1.055 0.230
Chain 1: 800 -8488.464 0.925 0.230
Chain 1: 900 -8337.565 0.824 0.074
Chain 1: 1000 -8082.427 0.745 0.074
Chain 1: 1100 -8237.455 0.647 0.035 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8082.174 0.084 0.032
Chain 1: 1300 -8235.771 0.049 0.028
Chain 1: 1400 -8149.543 0.043 0.019
Chain 1: 1500 -8111.705 0.020 0.019
Chain 1: 1600 -8107.210 0.017 0.019
Chain 1: 1700 -8056.914 0.014 0.018
Chain 1: 1800 -7935.752 0.014 0.018
Chain 1: 1900 -8045.654 0.014 0.015
Chain 1: 2000 -8011.248 0.011 0.014
Chain 1: 2100 -8159.473 0.011 0.014
Chain 1: 2200 -7937.383 0.012 0.014
Chain 1: 2300 -8018.929 0.011 0.011
Chain 1: 2400 -8084.838 0.011 0.010
Chain 1: 2500 -8046.947 0.011 0.010
Chain 1: 2600 -8040.091 0.011 0.010
Chain 1: 2700 -7952.525 0.011 0.011
Chain 1: 2800 -7940.056 0.010 0.010
Chain 1: 2900 -7943.306 0.009 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003721 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 37.21 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8400075.097 1.000 1.000
Chain 1: 200 -1585309.204 2.649 4.299
Chain 1: 300 -890954.059 2.026 1.000
Chain 1: 400 -457152.832 1.757 1.000
Chain 1: 500 -357235.640 1.461 0.949
Chain 1: 600 -232233.292 1.307 0.949
Chain 1: 700 -118542.934 1.258 0.949
Chain 1: 800 -85740.693 1.148 0.949
Chain 1: 900 -66099.150 1.054 0.779
Chain 1: 1000 -50894.278 0.978 0.779
Chain 1: 1100 -38376.688 0.911 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37547.141 0.483 0.383
Chain 1: 1300 -25529.422 0.452 0.383
Chain 1: 1400 -25246.070 0.359 0.326
Chain 1: 1500 -21840.371 0.346 0.326
Chain 1: 1600 -21057.421 0.296 0.299
Chain 1: 1700 -19935.534 0.206 0.297
Chain 1: 1800 -19880.039 0.168 0.156
Chain 1: 1900 -20205.174 0.140 0.056
Chain 1: 2000 -18720.474 0.118 0.056
Chain 1: 2100 -18958.565 0.086 0.037
Chain 1: 2200 -19183.915 0.085 0.037
Chain 1: 2300 -18802.386 0.040 0.020
Chain 1: 2400 -18574.903 0.040 0.020
Chain 1: 2500 -18376.767 0.026 0.016
Chain 1: 2600 -18008.106 0.024 0.016
Chain 1: 2700 -17965.488 0.019 0.013
Chain 1: 2800 -17682.691 0.020 0.016
Chain 1: 2900 -17963.490 0.020 0.016
Chain 1: 3000 -17949.789 0.012 0.013
Chain 1: 3100 -18034.563 0.011 0.012
Chain 1: 3200 -17725.975 0.012 0.016
Chain 1: 3300 -17930.172 0.011 0.012
Chain 1: 3400 -17406.290 0.013 0.016
Chain 1: 3500 -18016.261 0.015 0.016
Chain 1: 3600 -17325.537 0.017 0.016
Chain 1: 3700 -17710.357 0.019 0.017
Chain 1: 3800 -16673.959 0.024 0.022
Chain 1: 3900 -16670.217 0.022 0.022
Chain 1: 4000 -16787.526 0.023 0.022
Chain 1: 4100 -16701.403 0.023 0.022
Chain 1: 4200 -16518.581 0.022 0.022
Chain 1: 4300 -16656.343 0.022 0.022
Chain 1: 4400 -16613.880 0.019 0.011
Chain 1: 4500 -16516.545 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001242 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.42 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48274.761 1.000 1.000
Chain 1: 200 -16822.833 1.435 1.870
Chain 1: 300 -15046.159 0.996 1.000
Chain 1: 400 -15989.509 0.762 1.000
Chain 1: 500 -15785.359 0.612 0.118
Chain 1: 600 -10716.815 0.589 0.473
Chain 1: 700 -11644.590 0.516 0.118
Chain 1: 800 -14278.633 0.475 0.184
Chain 1: 900 -15272.411 0.429 0.118
Chain 1: 1000 -13288.122 0.401 0.149
Chain 1: 1100 -14182.543 0.307 0.118
Chain 1: 1200 -10483.845 0.156 0.118
Chain 1: 1300 -16064.514 0.179 0.149
Chain 1: 1400 -21680.921 0.199 0.184
Chain 1: 1500 -19165.384 0.211 0.184
Chain 1: 1600 -9300.686 0.269 0.184
Chain 1: 1700 -11187.258 0.278 0.184
Chain 1: 1800 -10672.509 0.265 0.169
Chain 1: 1900 -10357.694 0.261 0.169
Chain 1: 2000 -9807.071 0.252 0.169
Chain 1: 2100 -10514.256 0.252 0.169
Chain 1: 2200 -11133.259 0.222 0.131
Chain 1: 2300 -9007.673 0.211 0.131
Chain 1: 2400 -9518.163 0.191 0.067
Chain 1: 2500 -11528.179 0.195 0.067
Chain 1: 2600 -9602.004 0.109 0.067
Chain 1: 2700 -11234.575 0.107 0.067
Chain 1: 2800 -9795.837 0.117 0.145
Chain 1: 2900 -15464.237 0.150 0.147
Chain 1: 3000 -13678.047 0.158 0.147
Chain 1: 3100 -9093.496 0.201 0.174
Chain 1: 3200 -9350.323 0.199 0.174
Chain 1: 3300 -15879.674 0.216 0.174
Chain 1: 3400 -9527.865 0.277 0.201
Chain 1: 3500 -9316.266 0.262 0.201
Chain 1: 3600 -9872.578 0.248 0.147
Chain 1: 3700 -8542.762 0.249 0.156
Chain 1: 3800 -8545.597 0.234 0.156
Chain 1: 3900 -8841.647 0.201 0.131
Chain 1: 4000 -9247.153 0.192 0.056
Chain 1: 4100 -12614.659 0.168 0.056
Chain 1: 4200 -9076.623 0.205 0.156
Chain 1: 4300 -10180.299 0.174 0.108
Chain 1: 4400 -10664.253 0.112 0.056
Chain 1: 4500 -8534.667 0.135 0.108
Chain 1: 4600 -13660.351 0.167 0.156
Chain 1: 4700 -8617.214 0.210 0.250
Chain 1: 4800 -8211.442 0.215 0.250
Chain 1: 4900 -8828.929 0.218 0.250
Chain 1: 5000 -11839.835 0.239 0.254
Chain 1: 5100 -8400.557 0.254 0.254
Chain 1: 5200 -10251.048 0.233 0.250
Chain 1: 5300 -12759.386 0.242 0.250
Chain 1: 5400 -12854.525 0.238 0.250
Chain 1: 5500 -8749.395 0.260 0.254
Chain 1: 5600 -8580.546 0.224 0.197
Chain 1: 5700 -8614.510 0.166 0.181
Chain 1: 5800 -8278.978 0.165 0.181
Chain 1: 5900 -10108.302 0.176 0.181
Chain 1: 6000 -10826.281 0.157 0.181
Chain 1: 6100 -8100.990 0.150 0.181
Chain 1: 6200 -14306.386 0.175 0.181
Chain 1: 6300 -9233.158 0.211 0.181
Chain 1: 6400 -10688.469 0.224 0.181
Chain 1: 6500 -9919.165 0.184 0.136
Chain 1: 6600 -11511.259 0.196 0.138
Chain 1: 6700 -9735.887 0.214 0.181
Chain 1: 6800 -8457.399 0.225 0.181
Chain 1: 6900 -9164.131 0.215 0.151
Chain 1: 7000 -11858.292 0.231 0.182
Chain 1: 7100 -8800.798 0.232 0.182
Chain 1: 7200 -8273.514 0.195 0.151
Chain 1: 7300 -10740.963 0.163 0.151
Chain 1: 7400 -8088.454 0.182 0.182
Chain 1: 7500 -8033.362 0.175 0.182
Chain 1: 7600 -8295.904 0.165 0.182
Chain 1: 7700 -8405.203 0.148 0.151
Chain 1: 7800 -9922.842 0.148 0.153
Chain 1: 7900 -8018.413 0.164 0.227
Chain 1: 8000 -10670.709 0.166 0.230
Chain 1: 8100 -7997.884 0.165 0.230
Chain 1: 8200 -10510.849 0.182 0.238
Chain 1: 8300 -7943.524 0.191 0.239
Chain 1: 8400 -11123.494 0.187 0.239
Chain 1: 8500 -8250.770 0.221 0.249
Chain 1: 8600 -9178.094 0.228 0.249
Chain 1: 8700 -8180.531 0.239 0.249
Chain 1: 8800 -8450.802 0.227 0.249
Chain 1: 8900 -9995.269 0.219 0.249
Chain 1: 9000 -10060.896 0.195 0.239
Chain 1: 9100 -7892.535 0.189 0.239
Chain 1: 9200 -8013.419 0.166 0.155
Chain 1: 9300 -8353.180 0.138 0.122
Chain 1: 9400 -11214.297 0.135 0.122
Chain 1: 9500 -11608.482 0.104 0.101
Chain 1: 9600 -8569.296 0.129 0.122
Chain 1: 9700 -10581.918 0.136 0.155
Chain 1: 9800 -8012.756 0.165 0.190
Chain 1: 9900 -8343.069 0.153 0.190
Chain 1: 10000 -7866.666 0.159 0.190
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001463 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.63 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -45712.268 1.000 1.000
Chain 1: 200 -15076.546 1.516 2.032
Chain 1: 300 -8482.151 1.270 1.000
Chain 1: 400 -8298.255 0.958 1.000
Chain 1: 500 -8130.499 0.770 0.777
Chain 1: 600 -7858.796 0.648 0.777
Chain 1: 700 -7962.349 0.557 0.035
Chain 1: 800 -7519.839 0.495 0.059
Chain 1: 900 -7878.428 0.445 0.046
Chain 1: 1000 -7726.542 0.402 0.046
Chain 1: 1100 -7708.425 0.303 0.035
Chain 1: 1200 -7560.166 0.101 0.022
Chain 1: 1300 -7674.846 0.025 0.021
Chain 1: 1400 -7843.736 0.025 0.021
Chain 1: 1500 -7590.089 0.026 0.022
Chain 1: 1600 -7503.781 0.024 0.020
Chain 1: 1700 -7485.684 0.023 0.020
Chain 1: 1800 -7524.572 0.018 0.020
Chain 1: 1900 -7578.216 0.014 0.015
Chain 1: 2000 -7575.801 0.012 0.012
Chain 1: 2100 -7589.499 0.012 0.012
Chain 1: 2200 -7646.128 0.011 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003148 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.48 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86258.591 1.000 1.000
Chain 1: 200 -13062.541 3.302 5.604
Chain 1: 300 -9552.259 2.324 1.000
Chain 1: 400 -10347.137 1.762 1.000
Chain 1: 500 -8420.539 1.455 0.367
Chain 1: 600 -8175.078 1.218 0.367
Chain 1: 700 -8213.372 1.044 0.229
Chain 1: 800 -8580.569 0.919 0.229
Chain 1: 900 -8394.733 0.820 0.077
Chain 1: 1000 -8184.675 0.740 0.077
Chain 1: 1100 -8445.482 0.643 0.043 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8152.361 0.087 0.036
Chain 1: 1300 -8327.112 0.052 0.031
Chain 1: 1400 -8316.522 0.044 0.030
Chain 1: 1500 -8231.271 0.022 0.026
Chain 1: 1600 -8313.476 0.020 0.022
Chain 1: 1700 -8418.542 0.021 0.022
Chain 1: 1800 -8036.532 0.022 0.022
Chain 1: 1900 -8134.894 0.021 0.021
Chain 1: 2000 -8105.158 0.019 0.012
Chain 1: 2100 -8250.211 0.017 0.012
Chain 1: 2200 -8027.967 0.016 0.012
Chain 1: 2300 -8163.401 0.016 0.012
Chain 1: 2400 -8052.014 0.017 0.014
Chain 1: 2500 -8112.032 0.017 0.014
Chain 1: 2600 -8125.190 0.016 0.014
Chain 1: 2700 -8046.055 0.016 0.014
Chain 1: 2800 -8031.114 0.011 0.012
Chain 1: 2900 -8019.974 0.010 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002757 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 27.57 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8395974.803 1.000 1.000
Chain 1: 200 -1585066.382 2.648 4.297
Chain 1: 300 -890951.950 2.025 1.000
Chain 1: 400 -457294.370 1.756 1.000
Chain 1: 500 -357569.995 1.461 0.948
Chain 1: 600 -232507.878 1.307 0.948
Chain 1: 700 -118751.613 1.257 0.948
Chain 1: 800 -85935.854 1.148 0.948
Chain 1: 900 -66283.782 1.053 0.779
Chain 1: 1000 -51074.596 0.978 0.779
Chain 1: 1100 -38552.360 0.910 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37720.952 0.483 0.382
Chain 1: 1300 -25696.607 0.451 0.382
Chain 1: 1400 -25412.861 0.358 0.325
Chain 1: 1500 -22005.349 0.345 0.325
Chain 1: 1600 -21221.748 0.295 0.298
Chain 1: 1700 -20099.124 0.205 0.296
Chain 1: 1800 -20043.409 0.167 0.155
Chain 1: 1900 -20368.756 0.139 0.056
Chain 1: 2000 -18882.981 0.117 0.056
Chain 1: 2100 -19121.239 0.086 0.037
Chain 1: 2200 -19346.812 0.085 0.037
Chain 1: 2300 -18964.969 0.040 0.020
Chain 1: 2400 -18737.397 0.040 0.020
Chain 1: 2500 -18539.239 0.026 0.016
Chain 1: 2600 -18170.478 0.024 0.016
Chain 1: 2700 -18127.728 0.019 0.012
Chain 1: 2800 -17844.917 0.020 0.016
Chain 1: 2900 -18125.719 0.020 0.015
Chain 1: 3000 -18112.040 0.012 0.012
Chain 1: 3100 -18196.887 0.011 0.012
Chain 1: 3200 -17888.172 0.012 0.015
Chain 1: 3300 -18092.393 0.011 0.012
Chain 1: 3400 -17568.339 0.013 0.015
Chain 1: 3500 -18178.666 0.015 0.016
Chain 1: 3600 -17487.375 0.017 0.016
Chain 1: 3700 -17872.667 0.019 0.017
Chain 1: 3800 -16835.470 0.024 0.022
Chain 1: 3900 -16831.671 0.022 0.022
Chain 1: 4000 -16948.989 0.023 0.022
Chain 1: 4100 -16862.909 0.023 0.022
Chain 1: 4200 -16679.818 0.022 0.022
Chain 1: 4300 -16817.765 0.022 0.022
Chain 1: 4400 -16775.163 0.019 0.011
Chain 1: 4500 -16677.769 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001268 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.68 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12276.099 1.000 1.000
Chain 1: 200 -9201.038 0.667 1.000
Chain 1: 300 -7958.564 0.497 0.334
Chain 1: 400 -8133.768 0.378 0.334
Chain 1: 500 -7986.986 0.306 0.156
Chain 1: 600 -7911.947 0.257 0.156
Chain 1: 700 -7828.551 0.221 0.022
Chain 1: 800 -7868.180 0.194 0.022
Chain 1: 900 -7993.762 0.175 0.018
Chain 1: 1000 -7874.394 0.159 0.018
Chain 1: 1100 -7913.039 0.059 0.016
Chain 1: 1200 -7857.696 0.026 0.015
Chain 1: 1300 -7807.248 0.011 0.011
Chain 1: 1400 -7816.541 0.009 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001525 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.25 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57865.469 1.000 1.000
Chain 1: 200 -17534.121 1.650 2.300
Chain 1: 300 -8587.785 1.447 1.042
Chain 1: 400 -8165.528 1.098 1.042
Chain 1: 500 -8400.010 0.884 1.000
Chain 1: 600 -8562.440 0.740 1.000
Chain 1: 700 -8521.881 0.635 0.052
Chain 1: 800 -8131.354 0.562 0.052
Chain 1: 900 -7752.866 0.505 0.049
Chain 1: 1000 -7875.540 0.456 0.049
Chain 1: 1100 -7767.424 0.357 0.048
Chain 1: 1200 -7779.450 0.127 0.028
Chain 1: 1300 -7690.996 0.024 0.019
Chain 1: 1400 -7789.144 0.020 0.016
Chain 1: 1500 -7572.625 0.020 0.016
Chain 1: 1600 -7627.721 0.019 0.014
Chain 1: 1700 -7495.629 0.021 0.016
Chain 1: 1800 -7580.907 0.017 0.014
Chain 1: 1900 -7432.826 0.014 0.014
Chain 1: 2000 -7539.301 0.014 0.014
Chain 1: 2100 -7473.434 0.013 0.013
Chain 1: 2200 -7669.158 0.016 0.014
Chain 1: 2300 -7544.381 0.016 0.017
Chain 1: 2400 -7581.531 0.015 0.017
Chain 1: 2500 -7580.882 0.013 0.014
Chain 1: 2600 -7490.817 0.013 0.014
Chain 1: 2700 -7546.802 0.012 0.012
Chain 1: 2800 -7455.397 0.012 0.012
Chain 1: 2900 -7372.230 0.011 0.012
Chain 1: 3000 -7497.526 0.012 0.012
Chain 1: 3100 -7484.187 0.011 0.012
Chain 1: 3200 -7663.103 0.011 0.012
Chain 1: 3300 -7437.338 0.012 0.012
Chain 1: 3400 -7603.426 0.014 0.012
Chain 1: 3500 -7407.658 0.016 0.017
Chain 1: 3600 -7460.802 0.016 0.017
Chain 1: 3700 -7416.282 0.016 0.017
Chain 1: 3800 -7432.436 0.015 0.017
Chain 1: 3900 -7416.675 0.014 0.017
Chain 1: 4000 -7385.539 0.013 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002946 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.46 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86618.487 1.000 1.000
Chain 1: 200 -13343.632 3.246 5.491
Chain 1: 300 -9756.148 2.286 1.000
Chain 1: 400 -10650.790 1.736 1.000
Chain 1: 500 -8674.883 1.434 0.368
Chain 1: 600 -8410.473 1.200 0.368
Chain 1: 700 -8567.346 1.032 0.228
Chain 1: 800 -8975.577 0.908 0.228
Chain 1: 900 -8578.345 0.812 0.084
Chain 1: 1000 -8276.424 0.735 0.084
Chain 1: 1100 -8642.817 0.639 0.046 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8258.066 0.095 0.046
Chain 1: 1300 -8447.798 0.060 0.045
Chain 1: 1400 -8451.333 0.052 0.042
Chain 1: 1500 -8350.817 0.030 0.036
Chain 1: 1600 -8455.776 0.028 0.036
Chain 1: 1700 -8545.477 0.028 0.036
Chain 1: 1800 -8145.376 0.028 0.036
Chain 1: 1900 -8245.783 0.024 0.022
Chain 1: 2000 -8216.655 0.021 0.012
Chain 1: 2100 -8336.918 0.018 0.012
Chain 1: 2200 -8115.018 0.016 0.012
Chain 1: 2300 -8275.336 0.016 0.012
Chain 1: 2400 -8157.171 0.018 0.014
Chain 1: 2500 -8220.888 0.017 0.014
Chain 1: 2600 -8242.163 0.016 0.014
Chain 1: 2700 -8161.518 0.016 0.014
Chain 1: 2800 -8136.048 0.011 0.012
Chain 1: 2900 -8191.102 0.011 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003207 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.07 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8398528.910 1.000 1.000
Chain 1: 200 -1583740.794 2.651 4.303
Chain 1: 300 -889480.578 2.028 1.000
Chain 1: 400 -457216.938 1.757 1.000
Chain 1: 500 -357544.581 1.462 0.945
Chain 1: 600 -232649.688 1.307 0.945
Chain 1: 700 -118981.601 1.257 0.945
Chain 1: 800 -86203.925 1.148 0.945
Chain 1: 900 -66563.954 1.053 0.781
Chain 1: 1000 -51366.538 0.977 0.781
Chain 1: 1100 -38855.421 0.909 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38031.885 0.481 0.380
Chain 1: 1300 -26006.995 0.449 0.380
Chain 1: 1400 -25726.813 0.356 0.322
Chain 1: 1500 -22318.680 0.343 0.322
Chain 1: 1600 -21536.172 0.293 0.296
Chain 1: 1700 -20412.361 0.203 0.295
Chain 1: 1800 -20357.003 0.165 0.153
Chain 1: 1900 -20682.897 0.138 0.055
Chain 1: 2000 -19195.783 0.116 0.055
Chain 1: 2100 -19434.180 0.085 0.036
Chain 1: 2200 -19660.145 0.084 0.036
Chain 1: 2300 -19277.806 0.039 0.020
Chain 1: 2400 -19049.985 0.040 0.020
Chain 1: 2500 -18851.929 0.025 0.016
Chain 1: 2600 -18482.543 0.024 0.016
Chain 1: 2700 -18439.637 0.018 0.012
Chain 1: 2800 -18156.549 0.020 0.016
Chain 1: 2900 -18437.683 0.020 0.015
Chain 1: 3000 -18423.901 0.012 0.012
Chain 1: 3100 -18508.838 0.011 0.012
Chain 1: 3200 -18199.779 0.012 0.015
Chain 1: 3300 -18404.307 0.011 0.012
Chain 1: 3400 -17879.619 0.013 0.015
Chain 1: 3500 -18490.899 0.015 0.016
Chain 1: 3600 -17798.338 0.017 0.016
Chain 1: 3700 -18184.558 0.019 0.017
Chain 1: 3800 -17145.430 0.023 0.021
Chain 1: 3900 -17141.579 0.022 0.021
Chain 1: 4000 -17258.904 0.022 0.021
Chain 1: 4100 -17172.694 0.022 0.021
Chain 1: 4200 -16989.197 0.022 0.021
Chain 1: 4300 -17127.433 0.021 0.021
Chain 1: 4400 -17084.478 0.019 0.011
Chain 1: 4500 -16987.007 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001324 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.24 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12431.010 1.000 1.000
Chain 1: 200 -9184.370 0.677 1.000
Chain 1: 300 -8149.640 0.493 0.353
Chain 1: 400 -8225.340 0.372 0.353
Chain 1: 500 -8215.460 0.298 0.127
Chain 1: 600 -8003.385 0.253 0.127
Chain 1: 700 -7955.334 0.218 0.026
Chain 1: 800 -7953.344 0.190 0.026
Chain 1: 900 -8017.272 0.170 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00143 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.3 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -51594.487 1.000 1.000
Chain 1: 200 -16181.521 1.594 2.188
Chain 1: 300 -8694.981 1.350 1.000
Chain 1: 400 -8053.096 1.032 1.000
Chain 1: 500 -8488.490 0.836 0.861
Chain 1: 600 -9079.579 0.708 0.861
Chain 1: 700 -7983.944 0.626 0.137
Chain 1: 800 -8063.815 0.549 0.137
Chain 1: 900 -7996.093 0.489 0.080
Chain 1: 1000 -7855.889 0.442 0.080
Chain 1: 1100 -7731.076 0.344 0.065
Chain 1: 1200 -7751.220 0.125 0.051
Chain 1: 1300 -7752.174 0.039 0.018
Chain 1: 1400 -7715.023 0.031 0.016
Chain 1: 1500 -7633.757 0.027 0.011
Chain 1: 1600 -7820.557 0.023 0.011
Chain 1: 1700 -7550.138 0.013 0.011
Chain 1: 1800 -7646.874 0.013 0.013
Chain 1: 1900 -7652.396 0.013 0.013
Chain 1: 2000 -7661.592 0.011 0.011
Chain 1: 2100 -7687.739 0.010 0.005 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003145 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.45 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85857.222 1.000 1.000
Chain 1: 200 -13354.891 3.214 5.429
Chain 1: 300 -9817.390 2.263 1.000
Chain 1: 400 -10623.932 1.716 1.000
Chain 1: 500 -8743.155 1.416 0.360
Chain 1: 600 -8364.139 1.188 0.360
Chain 1: 700 -8696.933 1.023 0.215
Chain 1: 800 -8888.089 0.898 0.215
Chain 1: 900 -8654.081 0.801 0.076
Chain 1: 1000 -8407.290 0.724 0.076
Chain 1: 1100 -8713.963 0.628 0.045 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8383.197 0.089 0.039
Chain 1: 1300 -8384.818 0.053 0.038
Chain 1: 1400 -8387.429 0.045 0.035
Chain 1: 1500 -8420.073 0.024 0.029
Chain 1: 1600 -8426.683 0.020 0.027
Chain 1: 1700 -8356.283 0.017 0.022
Chain 1: 1800 -8240.978 0.016 0.014
Chain 1: 1900 -8358.063 0.015 0.014
Chain 1: 2000 -8318.281 0.012 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003557 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.57 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8403858.247 1.000 1.000
Chain 1: 200 -1584354.940 2.652 4.304
Chain 1: 300 -890980.366 2.027 1.000
Chain 1: 400 -457687.511 1.757 1.000
Chain 1: 500 -358028.635 1.462 0.947
Chain 1: 600 -232948.045 1.307 0.947
Chain 1: 700 -119099.963 1.257 0.947
Chain 1: 800 -86293.458 1.148 0.947
Chain 1: 900 -66622.338 1.053 0.778
Chain 1: 1000 -51409.129 0.977 0.778
Chain 1: 1100 -38882.437 0.909 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38050.103 0.481 0.380
Chain 1: 1300 -26014.709 0.450 0.380
Chain 1: 1400 -25730.986 0.356 0.322
Chain 1: 1500 -22321.265 0.343 0.322
Chain 1: 1600 -21537.472 0.293 0.296
Chain 1: 1700 -20413.126 0.203 0.295
Chain 1: 1800 -20357.138 0.166 0.153
Chain 1: 1900 -20682.710 0.138 0.055
Chain 1: 2000 -19195.818 0.116 0.055
Chain 1: 2100 -19433.986 0.085 0.036
Chain 1: 2200 -19660.002 0.084 0.036
Chain 1: 2300 -19277.740 0.039 0.020
Chain 1: 2400 -19050.060 0.040 0.020
Chain 1: 2500 -18852.037 0.025 0.016
Chain 1: 2600 -18482.899 0.024 0.016
Chain 1: 2700 -18439.995 0.018 0.012
Chain 1: 2800 -18157.183 0.020 0.016
Chain 1: 2900 -18438.067 0.020 0.015
Chain 1: 3000 -18424.335 0.012 0.012
Chain 1: 3100 -18509.248 0.011 0.012
Chain 1: 3200 -18200.319 0.012 0.015
Chain 1: 3300 -18404.681 0.011 0.012
Chain 1: 3400 -17880.344 0.013 0.015
Chain 1: 3500 -18491.151 0.015 0.016
Chain 1: 3600 -17799.180 0.017 0.016
Chain 1: 3700 -18185.008 0.019 0.017
Chain 1: 3800 -17146.835 0.023 0.021
Chain 1: 3900 -17143.016 0.022 0.021
Chain 1: 4000 -17260.318 0.022 0.021
Chain 1: 4100 -17174.246 0.022 0.021
Chain 1: 4200 -16990.892 0.022 0.021
Chain 1: 4300 -17129.001 0.021 0.021
Chain 1: 4400 -17086.200 0.019 0.011
Chain 1: 4500 -16988.794 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001301 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.01 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12477.649 1.000 1.000
Chain 1: 200 -9425.642 0.662 1.000
Chain 1: 300 -8116.733 0.495 0.324
Chain 1: 400 -8287.866 0.376 0.324
Chain 1: 500 -8110.256 0.306 0.161
Chain 1: 600 -8044.129 0.256 0.161
Chain 1: 700 -7954.012 0.221 0.022
Chain 1: 800 -7962.338 0.194 0.022
Chain 1: 900 -7929.907 0.172 0.021
Chain 1: 1000 -8066.354 0.157 0.021
Chain 1: 1100 -8064.720 0.057 0.017
Chain 1: 1200 -7977.861 0.026 0.011
Chain 1: 1300 -7928.492 0.010 0.011
Chain 1: 1400 -7945.677 0.008 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001434 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.34 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58074.333 1.000 1.000
Chain 1: 200 -17781.605 1.633 2.266
Chain 1: 300 -8709.402 1.436 1.042
Chain 1: 400 -8126.242 1.095 1.042
Chain 1: 500 -8581.679 0.886 1.000
Chain 1: 600 -8052.294 0.750 1.000
Chain 1: 700 -8112.970 0.644 0.072
Chain 1: 800 -8171.189 0.564 0.072
Chain 1: 900 -7894.736 0.505 0.066
Chain 1: 1000 -7791.618 0.456 0.066
Chain 1: 1100 -7824.182 0.357 0.053
Chain 1: 1200 -7564.157 0.133 0.035
Chain 1: 1300 -7743.211 0.032 0.034
Chain 1: 1400 -7630.353 0.026 0.023
Chain 1: 1500 -7568.250 0.021 0.015
Chain 1: 1600 -7533.669 0.015 0.013
Chain 1: 1700 -7536.562 0.015 0.013
Chain 1: 1800 -7576.064 0.014 0.013
Chain 1: 1900 -7577.381 0.011 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003375 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.75 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86654.943 1.000 1.000
Chain 1: 200 -13578.919 3.191 5.382
Chain 1: 300 -9934.447 2.249 1.000
Chain 1: 400 -10813.603 1.707 1.000
Chain 1: 500 -8917.140 1.408 0.367
Chain 1: 600 -8808.354 1.176 0.367
Chain 1: 700 -8741.443 1.009 0.213
Chain 1: 800 -8740.101 0.883 0.213
Chain 1: 900 -8734.750 0.785 0.081
Chain 1: 1000 -8614.562 0.708 0.081
Chain 1: 1100 -8716.519 0.609 0.014 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8405.626 0.074 0.014
Chain 1: 1300 -8634.435 0.040 0.014
Chain 1: 1400 -8641.836 0.032 0.012
Chain 1: 1500 -8496.792 0.013 0.012
Chain 1: 1600 -8609.796 0.013 0.013
Chain 1: 1700 -8691.985 0.013 0.013
Chain 1: 1800 -8276.808 0.018 0.014
Chain 1: 1900 -8373.890 0.019 0.014
Chain 1: 2000 -8347.465 0.018 0.013
Chain 1: 2100 -8470.603 0.018 0.015
Chain 1: 2200 -8289.534 0.017 0.015
Chain 1: 2300 -8368.558 0.015 0.013
Chain 1: 2400 -8438.286 0.016 0.013
Chain 1: 2500 -8383.954 0.015 0.012
Chain 1: 2600 -8383.754 0.013 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003236 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.36 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8422093.113 1.000 1.000
Chain 1: 200 -1589296.888 2.650 4.299
Chain 1: 300 -892713.314 2.027 1.000
Chain 1: 400 -458731.284 1.756 1.000
Chain 1: 500 -358551.632 1.461 0.946
Chain 1: 600 -233142.740 1.307 0.946
Chain 1: 700 -119290.477 1.257 0.946
Chain 1: 800 -86509.083 1.147 0.946
Chain 1: 900 -66842.018 1.052 0.780
Chain 1: 1000 -51641.002 0.976 0.780
Chain 1: 1100 -39125.810 0.908 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38302.163 0.481 0.379
Chain 1: 1300 -26265.629 0.448 0.379
Chain 1: 1400 -25985.562 0.355 0.320
Chain 1: 1500 -22575.246 0.342 0.320
Chain 1: 1600 -21792.732 0.292 0.294
Chain 1: 1700 -20667.182 0.202 0.294
Chain 1: 1800 -20611.599 0.164 0.151
Chain 1: 1900 -20937.682 0.136 0.054
Chain 1: 2000 -19449.541 0.115 0.054
Chain 1: 2100 -19687.760 0.084 0.036
Chain 1: 2200 -19914.211 0.083 0.036
Chain 1: 2300 -19531.448 0.039 0.020
Chain 1: 2400 -19303.541 0.039 0.020
Chain 1: 2500 -19105.619 0.025 0.016
Chain 1: 2600 -18735.755 0.023 0.016
Chain 1: 2700 -18692.732 0.018 0.012
Chain 1: 2800 -18409.609 0.019 0.015
Chain 1: 2900 -18690.856 0.019 0.015
Chain 1: 3000 -18677.019 0.012 0.012
Chain 1: 3100 -18762.009 0.011 0.012
Chain 1: 3200 -18452.690 0.012 0.015
Chain 1: 3300 -18657.431 0.011 0.012
Chain 1: 3400 -18132.365 0.012 0.015
Chain 1: 3500 -18744.208 0.015 0.015
Chain 1: 3600 -18050.943 0.017 0.015
Chain 1: 3700 -18437.669 0.018 0.017
Chain 1: 3800 -17397.482 0.023 0.021
Chain 1: 3900 -17393.641 0.021 0.021
Chain 1: 4000 -17510.943 0.022 0.021
Chain 1: 4100 -17424.695 0.022 0.021
Chain 1: 4200 -17240.976 0.021 0.021
Chain 1: 4300 -17379.334 0.021 0.021
Chain 1: 4400 -17336.162 0.018 0.011
Chain 1: 4500 -17238.718 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001317 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.17 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12865.654 1.000 1.000
Chain 1: 200 -9772.240 0.658 1.000
Chain 1: 300 -8343.204 0.496 0.317
Chain 1: 400 -8567.476 0.379 0.317
Chain 1: 500 -8431.782 0.306 0.171
Chain 1: 600 -8283.372 0.258 0.171
Chain 1: 700 -8359.608 0.222 0.026
Chain 1: 800 -8221.282 0.197 0.026
Chain 1: 900 -8383.030 0.177 0.019
Chain 1: 1000 -8260.568 0.161 0.019
Chain 1: 1100 -8305.775 0.061 0.018
Chain 1: 1200 -8213.909 0.031 0.017
Chain 1: 1300 -8152.763 0.014 0.016
Chain 1: 1400 -8161.797 0.012 0.015
Chain 1: 1500 -8251.341 0.011 0.011
Chain 1: 1600 -8170.640 0.011 0.011
Chain 1: 1700 -8137.294 0.010 0.011
Chain 1: 1800 -8109.656 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001394 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.94 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -59221.520 1.000 1.000
Chain 1: 200 -18206.699 1.626 2.253
Chain 1: 300 -9088.083 1.419 1.003
Chain 1: 400 -8379.927 1.085 1.003
Chain 1: 500 -8526.872 0.872 1.000
Chain 1: 600 -8505.197 0.727 1.000
Chain 1: 700 -8110.831 0.630 0.085
Chain 1: 800 -8121.765 0.551 0.085
Chain 1: 900 -8090.546 0.490 0.049
Chain 1: 1000 -7781.755 0.445 0.049
Chain 1: 1100 -7749.438 0.346 0.040
Chain 1: 1200 -7616.928 0.122 0.017
Chain 1: 1300 -7824.465 0.025 0.017
Chain 1: 1400 -7949.251 0.018 0.017
Chain 1: 1500 -7614.954 0.020 0.017
Chain 1: 1600 -7774.927 0.022 0.021
Chain 1: 1700 -7475.764 0.021 0.021
Chain 1: 1800 -7609.953 0.023 0.021
Chain 1: 1900 -7583.162 0.023 0.021
Chain 1: 2000 -7670.614 0.020 0.018
Chain 1: 2100 -7512.976 0.022 0.021
Chain 1: 2200 -7636.015 0.022 0.021
Chain 1: 2300 -7571.390 0.020 0.018
Chain 1: 2400 -7607.189 0.019 0.018
Chain 1: 2500 -7631.792 0.015 0.016
Chain 1: 2600 -7515.300 0.014 0.016
Chain 1: 2700 -7482.788 0.011 0.011
Chain 1: 2800 -7518.260 0.009 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003295 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.95 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87368.904 1.000 1.000
Chain 1: 200 -14074.397 3.104 5.208
Chain 1: 300 -10309.909 2.191 1.000
Chain 1: 400 -12023.684 1.679 1.000
Chain 1: 500 -8882.227 1.414 0.365
Chain 1: 600 -8729.301 1.181 0.365
Chain 1: 700 -8826.990 1.014 0.354
Chain 1: 800 -9178.934 0.892 0.354
Chain 1: 900 -9072.218 0.794 0.143
Chain 1: 1000 -9195.266 0.716 0.143
Chain 1: 1100 -9094.257 0.617 0.038 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8623.037 0.102 0.038
Chain 1: 1300 -8941.736 0.069 0.036
Chain 1: 1400 -8748.986 0.057 0.022
Chain 1: 1500 -8796.567 0.022 0.018
Chain 1: 1600 -8904.082 0.022 0.013
Chain 1: 1700 -8957.344 0.021 0.013
Chain 1: 1800 -8508.308 0.022 0.013
Chain 1: 1900 -8616.289 0.023 0.013
Chain 1: 2000 -8602.376 0.021 0.013
Chain 1: 2100 -8721.637 0.022 0.014
Chain 1: 2200 -8509.828 0.019 0.014
Chain 1: 2300 -8670.783 0.017 0.014
Chain 1: 2400 -8508.579 0.017 0.014
Chain 1: 2500 -8587.778 0.017 0.014
Chain 1: 2600 -8522.968 0.017 0.014
Chain 1: 2700 -8531.417 0.016 0.014
Chain 1: 2800 -8486.888 0.011 0.013
Chain 1: 2900 -8596.407 0.011 0.013
Chain 1: 3000 -8546.191 0.012 0.013
Chain 1: 3100 -8478.042 0.011 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003255 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.55 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8409091.198 1.000 1.000
Chain 1: 200 -1584414.037 2.654 4.307
Chain 1: 300 -891995.497 2.028 1.000
Chain 1: 400 -458593.180 1.757 1.000
Chain 1: 500 -358898.369 1.461 0.945
Chain 1: 600 -233641.449 1.307 0.945
Chain 1: 700 -119862.652 1.256 0.945
Chain 1: 800 -87052.609 1.146 0.945
Chain 1: 900 -67394.359 1.051 0.776
Chain 1: 1000 -52197.464 0.975 0.776
Chain 1: 1100 -39673.111 0.907 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38858.742 0.478 0.377
Chain 1: 1300 -26804.745 0.445 0.377
Chain 1: 1400 -26526.330 0.352 0.316
Chain 1: 1500 -23110.093 0.339 0.316
Chain 1: 1600 -22326.686 0.289 0.292
Chain 1: 1700 -21198.559 0.199 0.291
Chain 1: 1800 -21142.893 0.162 0.148
Chain 1: 1900 -21469.648 0.134 0.053
Chain 1: 2000 -19978.816 0.113 0.053
Chain 1: 2100 -20217.330 0.082 0.035
Chain 1: 2200 -20444.285 0.081 0.035
Chain 1: 2300 -20060.928 0.038 0.019
Chain 1: 2400 -19832.793 0.038 0.019
Chain 1: 2500 -19634.744 0.024 0.015
Chain 1: 2600 -19264.234 0.023 0.015
Chain 1: 2700 -19221.122 0.018 0.012
Chain 1: 2800 -18937.585 0.019 0.015
Chain 1: 2900 -19219.257 0.019 0.015
Chain 1: 3000 -19205.333 0.012 0.012
Chain 1: 3100 -19290.376 0.011 0.012
Chain 1: 3200 -18980.655 0.011 0.015
Chain 1: 3300 -19185.774 0.010 0.012
Chain 1: 3400 -18659.869 0.012 0.015
Chain 1: 3500 -19272.906 0.014 0.015
Chain 1: 3600 -18578.191 0.016 0.015
Chain 1: 3700 -18965.988 0.018 0.016
Chain 1: 3800 -17923.400 0.022 0.020
Chain 1: 3900 -17919.506 0.021 0.020
Chain 1: 4000 -18036.818 0.021 0.020
Chain 1: 4100 -17950.383 0.021 0.020
Chain 1: 4200 -17766.207 0.021 0.020
Chain 1: 4300 -17904.906 0.021 0.020
Chain 1: 4400 -17861.334 0.018 0.010
Chain 1: 4500 -17763.801 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001291 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.91 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12407.553 1.000 1.000
Chain 1: 200 -9180.435 0.676 1.000
Chain 1: 300 -7950.686 0.502 0.352
Chain 1: 400 -8179.515 0.384 0.352
Chain 1: 500 -8062.426 0.310 0.155
Chain 1: 600 -8069.146 0.258 0.155
Chain 1: 700 -7871.810 0.225 0.028
Chain 1: 800 -7821.186 0.198 0.028
Chain 1: 900 -7774.384 0.176 0.025
Chain 1: 1000 -7935.413 0.161 0.025
Chain 1: 1100 -8012.433 0.062 0.020
Chain 1: 1200 -7875.077 0.028 0.017
Chain 1: 1300 -7810.203 0.014 0.015
Chain 1: 1400 -7850.217 0.011 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001391 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.91 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57039.766 1.000 1.000
Chain 1: 200 -17360.087 1.643 2.286
Chain 1: 300 -8753.152 1.423 1.000
Chain 1: 400 -8327.158 1.080 1.000
Chain 1: 500 -8579.932 0.870 0.983
Chain 1: 600 -8658.863 0.726 0.983
Chain 1: 700 -7816.939 0.638 0.108
Chain 1: 800 -8153.343 0.563 0.108
Chain 1: 900 -8004.330 0.503 0.051
Chain 1: 1000 -7879.918 0.454 0.051
Chain 1: 1100 -7881.151 0.354 0.041
Chain 1: 1200 -7627.689 0.129 0.033
Chain 1: 1300 -7791.997 0.033 0.029
Chain 1: 1400 -7875.973 0.029 0.021
Chain 1: 1500 -7654.391 0.029 0.021
Chain 1: 1600 -7844.582 0.030 0.024
Chain 1: 1700 -7577.086 0.023 0.024
Chain 1: 1800 -7632.288 0.020 0.021
Chain 1: 1900 -7647.676 0.018 0.021
Chain 1: 2000 -7673.508 0.017 0.021
Chain 1: 2100 -7670.635 0.017 0.021
Chain 1: 2200 -7745.873 0.014 0.011
Chain 1: 2300 -7814.105 0.013 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003391 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.91 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86619.772 1.000 1.000
Chain 1: 200 -13472.045 3.215 5.430
Chain 1: 300 -9843.503 2.266 1.000
Chain 1: 400 -10877.348 1.723 1.000
Chain 1: 500 -8679.759 1.429 0.369
Chain 1: 600 -8317.377 1.198 0.369
Chain 1: 700 -8330.674 1.027 0.253
Chain 1: 800 -9052.207 0.909 0.253
Chain 1: 900 -8642.021 0.813 0.095
Chain 1: 1000 -8551.367 0.733 0.095
Chain 1: 1100 -8680.308 0.634 0.080 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8193.590 0.097 0.059
Chain 1: 1300 -8535.998 0.065 0.047
Chain 1: 1400 -8534.932 0.055 0.044
Chain 1: 1500 -8400.734 0.031 0.040
Chain 1: 1600 -8518.214 0.028 0.016
Chain 1: 1700 -8591.668 0.029 0.016
Chain 1: 1800 -8178.078 0.026 0.016
Chain 1: 1900 -8274.329 0.023 0.015
Chain 1: 2000 -8247.718 0.022 0.015
Chain 1: 2100 -8370.995 0.022 0.015
Chain 1: 2200 -8189.970 0.018 0.015
Chain 1: 2300 -8268.948 0.015 0.014
Chain 1: 2400 -8338.645 0.016 0.014
Chain 1: 2500 -8284.300 0.015 0.012
Chain 1: 2600 -8284.063 0.014 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003114 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.14 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8440532.579 1.000 1.000
Chain 1: 200 -1592337.315 2.650 4.301
Chain 1: 300 -891495.682 2.029 1.000
Chain 1: 400 -457634.986 1.759 1.000
Chain 1: 500 -357171.907 1.463 0.948
Chain 1: 600 -231951.015 1.309 0.948
Chain 1: 700 -118607.730 1.259 0.948
Chain 1: 800 -85970.058 1.149 0.948
Chain 1: 900 -66418.988 1.054 0.786
Chain 1: 1000 -51313.026 0.978 0.786
Chain 1: 1100 -38882.190 0.910 0.540 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38067.588 0.482 0.380
Chain 1: 1300 -26112.433 0.449 0.380
Chain 1: 1400 -25840.102 0.355 0.320
Chain 1: 1500 -22450.798 0.342 0.320
Chain 1: 1600 -21674.636 0.292 0.294
Chain 1: 1700 -20558.825 0.202 0.294
Chain 1: 1800 -20505.467 0.164 0.151
Chain 1: 1900 -20831.634 0.136 0.054
Chain 1: 2000 -19348.315 0.115 0.054
Chain 1: 2100 -19586.339 0.084 0.036
Chain 1: 2200 -19812.014 0.083 0.036
Chain 1: 2300 -19429.890 0.039 0.020
Chain 1: 2400 -19202.091 0.039 0.020
Chain 1: 2500 -19003.820 0.025 0.016
Chain 1: 2600 -18634.339 0.023 0.016
Chain 1: 2700 -18591.412 0.018 0.012
Chain 1: 2800 -18308.192 0.020 0.015
Chain 1: 2900 -18589.233 0.019 0.015
Chain 1: 3000 -18575.524 0.012 0.012
Chain 1: 3100 -18660.525 0.011 0.012
Chain 1: 3200 -18351.288 0.012 0.015
Chain 1: 3300 -18555.931 0.011 0.012
Chain 1: 3400 -18030.907 0.013 0.015
Chain 1: 3500 -18642.609 0.015 0.015
Chain 1: 3600 -17949.417 0.017 0.015
Chain 1: 3700 -18336.045 0.019 0.017
Chain 1: 3800 -17295.975 0.023 0.021
Chain 1: 3900 -17292.068 0.022 0.021
Chain 1: 4000 -17409.415 0.022 0.021
Chain 1: 4100 -17323.205 0.022 0.021
Chain 1: 4200 -17139.478 0.022 0.021
Chain 1: 4300 -17277.886 0.021 0.021
Chain 1: 4400 -17234.725 0.019 0.011
Chain 1: 4500 -17137.228 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001447 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.47 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12753.420 1.000 1.000
Chain 1: 200 -9549.743 0.668 1.000
Chain 1: 300 -8166.054 0.502 0.335
Chain 1: 400 -8407.054 0.383 0.335
Chain 1: 500 -8015.271 0.316 0.169
Chain 1: 600 -8152.784 0.267 0.169
Chain 1: 700 -8220.842 0.230 0.049
Chain 1: 800 -8056.197 0.204 0.049
Chain 1: 900 -8175.930 0.183 0.029
Chain 1: 1000 -8118.745 0.165 0.029
Chain 1: 1100 -8155.319 0.065 0.020
Chain 1: 1200 -8061.620 0.033 0.017
Chain 1: 1300 -8151.019 0.017 0.015
Chain 1: 1400 -8041.168 0.016 0.014
Chain 1: 1500 -8145.870 0.012 0.013
Chain 1: 1600 -8067.971 0.011 0.012
Chain 1: 1700 -8022.182 0.011 0.012
Chain 1: 1800 -7996.601 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001376 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.76 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57131.884 1.000 1.000
Chain 1: 200 -17765.176 1.608 2.216
Chain 1: 300 -8860.944 1.407 1.005
Chain 1: 400 -8248.560 1.074 1.005
Chain 1: 500 -9089.188 0.878 1.000
Chain 1: 600 -8777.656 0.737 1.000
Chain 1: 700 -8769.568 0.632 0.092
Chain 1: 800 -8024.932 0.565 0.093
Chain 1: 900 -7681.064 0.507 0.092
Chain 1: 1000 -7750.024 0.457 0.092
Chain 1: 1100 -7830.871 0.358 0.074
Chain 1: 1200 -7969.181 0.138 0.045
Chain 1: 1300 -7498.265 0.044 0.045
Chain 1: 1400 -8001.817 0.043 0.045
Chain 1: 1500 -7544.126 0.040 0.045
Chain 1: 1600 -7743.135 0.039 0.045
Chain 1: 1700 -7541.395 0.041 0.045
Chain 1: 1800 -7602.821 0.033 0.027
Chain 1: 1900 -7492.255 0.030 0.026
Chain 1: 2000 -7653.070 0.031 0.026
Chain 1: 2100 -7530.056 0.032 0.026
Chain 1: 2200 -7721.504 0.032 0.026
Chain 1: 2300 -7501.146 0.029 0.026
Chain 1: 2400 -7530.778 0.023 0.025
Chain 1: 2500 -7557.825 0.017 0.021
Chain 1: 2600 -7479.474 0.016 0.016
Chain 1: 2700 -7397.613 0.014 0.015
Chain 1: 2800 -7453.352 0.014 0.015
Chain 1: 2900 -7371.362 0.014 0.011
Chain 1: 3000 -7496.552 0.013 0.011
Chain 1: 3100 -7483.387 0.012 0.011
Chain 1: 3200 -7681.846 0.012 0.011
Chain 1: 3300 -7398.453 0.013 0.011
Chain 1: 3400 -7636.367 0.016 0.011
Chain 1: 3500 -7384.489 0.019 0.017
Chain 1: 3600 -7450.208 0.019 0.017
Chain 1: 3700 -7400.110 0.018 0.017
Chain 1: 3800 -7400.433 0.017 0.017
Chain 1: 3900 -7361.323 0.017 0.017
Chain 1: 4000 -7353.125 0.015 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002571 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.71 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87831.088 1.000 1.000
Chain 1: 200 -13829.150 3.176 5.351
Chain 1: 300 -10120.808 2.239 1.000
Chain 1: 400 -11288.221 1.705 1.000
Chain 1: 500 -9130.885 1.411 0.366
Chain 1: 600 -8534.398 1.188 0.366
Chain 1: 700 -8561.730 1.019 0.236
Chain 1: 800 -8896.816 0.896 0.236
Chain 1: 900 -8970.443 0.797 0.103
Chain 1: 1000 -8722.940 0.720 0.103
Chain 1: 1100 -8718.669 0.621 0.070 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8496.054 0.088 0.038
Chain 1: 1300 -8767.975 0.054 0.031
Chain 1: 1400 -8789.389 0.044 0.028
Chain 1: 1500 -8636.602 0.023 0.026
Chain 1: 1600 -8745.762 0.017 0.018
Chain 1: 1700 -8816.767 0.017 0.018
Chain 1: 1800 -8381.644 0.019 0.018
Chain 1: 1900 -8486.279 0.019 0.018
Chain 1: 2000 -8461.986 0.017 0.012
Chain 1: 2100 -8599.320 0.018 0.016
Chain 1: 2200 -8393.370 0.018 0.016
Chain 1: 2300 -8551.489 0.017 0.016
Chain 1: 2400 -8390.665 0.018 0.018
Chain 1: 2500 -8460.740 0.017 0.016
Chain 1: 2600 -8372.973 0.017 0.016
Chain 1: 2700 -8406.499 0.017 0.016
Chain 1: 2800 -8366.878 0.012 0.012
Chain 1: 2900 -8459.799 0.012 0.011
Chain 1: 3000 -8290.743 0.014 0.016
Chain 1: 3100 -8449.239 0.014 0.018
Chain 1: 3200 -8321.514 0.013 0.015
Chain 1: 3300 -8329.343 0.011 0.011
Chain 1: 3400 -8486.669 0.011 0.011
Chain 1: 3500 -8490.271 0.010 0.011
Chain 1: 3600 -8278.208 0.012 0.015
Chain 1: 3700 -8423.311 0.013 0.017
Chain 1: 3800 -8284.849 0.014 0.017
Chain 1: 3900 -8219.608 0.014 0.017
Chain 1: 4000 -8294.425 0.013 0.017
Chain 1: 4100 -8285.473 0.011 0.015
Chain 1: 4200 -8274.962 0.010 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004518 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 45.18 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8438870.201 1.000 1.000
Chain 1: 200 -1596919.332 2.642 4.284
Chain 1: 300 -893348.854 2.024 1.000
Chain 1: 400 -458280.624 1.755 1.000
Chain 1: 500 -357277.953 1.461 0.949
Chain 1: 600 -231907.400 1.307 0.949
Chain 1: 700 -118786.873 1.257 0.949
Chain 1: 800 -86118.652 1.147 0.949
Chain 1: 900 -66612.148 1.052 0.788
Chain 1: 1000 -51557.490 0.976 0.788
Chain 1: 1100 -39164.703 0.908 0.541 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38366.161 0.481 0.379
Chain 1: 1300 -26458.629 0.448 0.379
Chain 1: 1400 -26193.198 0.354 0.316
Chain 1: 1500 -22813.489 0.340 0.316
Chain 1: 1600 -22040.436 0.290 0.293
Chain 1: 1700 -20930.450 0.200 0.292
Chain 1: 1800 -20878.719 0.162 0.148
Chain 1: 1900 -21205.213 0.134 0.053
Chain 1: 2000 -19723.754 0.113 0.053
Chain 1: 2100 -19962.158 0.082 0.035
Chain 1: 2200 -20187.155 0.081 0.035
Chain 1: 2300 -19805.495 0.038 0.019
Chain 1: 2400 -19577.568 0.038 0.019
Chain 1: 2500 -19378.760 0.025 0.015
Chain 1: 2600 -19009.395 0.023 0.015
Chain 1: 2700 -18966.622 0.018 0.012
Chain 1: 2800 -18682.804 0.019 0.015
Chain 1: 2900 -18964.148 0.019 0.015
Chain 1: 3000 -18950.556 0.012 0.012
Chain 1: 3100 -19035.476 0.011 0.012
Chain 1: 3200 -18726.164 0.011 0.015
Chain 1: 3300 -18930.963 0.011 0.012
Chain 1: 3400 -18405.397 0.012 0.015
Chain 1: 3500 -19017.652 0.014 0.015
Chain 1: 3600 -18323.908 0.016 0.015
Chain 1: 3700 -18710.778 0.018 0.017
Chain 1: 3800 -17669.597 0.023 0.021
Chain 1: 3900 -17665.602 0.021 0.021
Chain 1: 4000 -17783.063 0.022 0.021
Chain 1: 4100 -17696.584 0.022 0.021
Chain 1: 4200 -17512.747 0.021 0.021
Chain 1: 4300 -17651.333 0.021 0.021
Chain 1: 4400 -17608.015 0.018 0.010
Chain 1: 4500 -17510.391 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001625 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 16.25 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49264.358 1.000 1.000
Chain 1: 200 -18111.041 1.360 1.720
Chain 1: 300 -16445.977 0.940 1.000
Chain 1: 400 -17261.910 0.717 1.000
Chain 1: 500 -15028.968 0.603 0.149
Chain 1: 600 -14227.834 0.512 0.149
Chain 1: 700 -16033.666 0.455 0.113
Chain 1: 800 -13471.656 0.422 0.149
Chain 1: 900 -12230.731 0.386 0.113
Chain 1: 1000 -12888.482 0.353 0.113
Chain 1: 1100 -11275.872 0.267 0.113
Chain 1: 1200 -12248.891 0.103 0.101
Chain 1: 1300 -17901.404 0.125 0.113
Chain 1: 1400 -11296.439 0.178 0.143
Chain 1: 1500 -10914.125 0.167 0.113
Chain 1: 1600 -11209.772 0.164 0.113
Chain 1: 1700 -11388.071 0.154 0.101
Chain 1: 1800 -11199.349 0.137 0.079
Chain 1: 1900 -10590.861 0.133 0.057
Chain 1: 2000 -20592.057 0.176 0.079
Chain 1: 2100 -11318.964 0.244 0.079
Chain 1: 2200 -11848.749 0.240 0.057
Chain 1: 2300 -9386.375 0.235 0.057
Chain 1: 2400 -9687.523 0.179 0.045
Chain 1: 2500 -9765.724 0.177 0.045
Chain 1: 2600 -9925.819 0.176 0.045
Chain 1: 2700 -10510.631 0.180 0.056
Chain 1: 2800 -10413.364 0.179 0.056
Chain 1: 2900 -9743.433 0.180 0.056
Chain 1: 3000 -10700.016 0.140 0.056
Chain 1: 3100 -9468.514 0.072 0.056
Chain 1: 3200 -12640.645 0.092 0.069
Chain 1: 3300 -16928.409 0.091 0.069
Chain 1: 3400 -13195.932 0.116 0.089
Chain 1: 3500 -9701.894 0.152 0.130
Chain 1: 3600 -11279.204 0.164 0.140
Chain 1: 3700 -17935.782 0.196 0.251
Chain 1: 3800 -9073.300 0.292 0.253
Chain 1: 3900 -10988.444 0.303 0.253
Chain 1: 4000 -8878.827 0.318 0.253
Chain 1: 4100 -11386.608 0.327 0.253
Chain 1: 4200 -9589.590 0.320 0.253
Chain 1: 4300 -14214.395 0.328 0.283
Chain 1: 4400 -13891.144 0.302 0.238
Chain 1: 4500 -11391.954 0.288 0.220
Chain 1: 4600 -8564.894 0.307 0.238
Chain 1: 4700 -8696.413 0.271 0.220
Chain 1: 4800 -8661.200 0.174 0.219
Chain 1: 4900 -9253.938 0.163 0.219
Chain 1: 5000 -13794.653 0.172 0.219
Chain 1: 5100 -10961.522 0.176 0.219
Chain 1: 5200 -8870.218 0.180 0.236
Chain 1: 5300 -9753.798 0.157 0.219
Chain 1: 5400 -8511.589 0.169 0.219
Chain 1: 5500 -12600.870 0.180 0.236
Chain 1: 5600 -10588.338 0.166 0.190
Chain 1: 5700 -10242.562 0.168 0.190
Chain 1: 5800 -8464.641 0.188 0.210
Chain 1: 5900 -14017.345 0.221 0.236
Chain 1: 6000 -8569.437 0.252 0.236
Chain 1: 6100 -8564.445 0.226 0.210
Chain 1: 6200 -8464.899 0.204 0.190
Chain 1: 6300 -8421.127 0.195 0.190
Chain 1: 6400 -14500.772 0.223 0.210
Chain 1: 6500 -10487.332 0.229 0.210
Chain 1: 6600 -12655.033 0.227 0.210
Chain 1: 6700 -8615.735 0.270 0.383
Chain 1: 6800 -8945.776 0.253 0.383
Chain 1: 6900 -11297.543 0.234 0.208
Chain 1: 7000 -8854.301 0.198 0.208
Chain 1: 7100 -8839.808 0.198 0.208
Chain 1: 7200 -8716.761 0.198 0.208
Chain 1: 7300 -8930.828 0.200 0.208
Chain 1: 7400 -8633.680 0.162 0.171
Chain 1: 7500 -10391.448 0.140 0.169
Chain 1: 7600 -8589.966 0.144 0.169
Chain 1: 7700 -8720.221 0.099 0.037
Chain 1: 7800 -8447.024 0.098 0.034
Chain 1: 7900 -8167.282 0.081 0.034
Chain 1: 8000 -12669.079 0.089 0.034
Chain 1: 8100 -10802.143 0.106 0.034
Chain 1: 8200 -8351.561 0.134 0.169
Chain 1: 8300 -10205.127 0.150 0.173
Chain 1: 8400 -8552.303 0.166 0.182
Chain 1: 8500 -11992.440 0.177 0.193
Chain 1: 8600 -8550.789 0.197 0.193
Chain 1: 8700 -8460.877 0.196 0.193
Chain 1: 8800 -8380.723 0.194 0.193
Chain 1: 8900 -8546.808 0.193 0.193
Chain 1: 9000 -10530.374 0.176 0.188
Chain 1: 9100 -8176.096 0.187 0.193
Chain 1: 9200 -8423.720 0.161 0.188
Chain 1: 9300 -8831.938 0.147 0.188
Chain 1: 9400 -12376.312 0.157 0.188
Chain 1: 9500 -9385.621 0.160 0.188
Chain 1: 9600 -10326.026 0.129 0.091
Chain 1: 9700 -12813.575 0.147 0.188
Chain 1: 9800 -8359.341 0.199 0.194
Chain 1: 9900 -10313.787 0.216 0.194
Chain 1: 10000 -8545.880 0.218 0.207
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001524 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.24 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58178.241 1.000 1.000
Chain 1: 200 -17793.276 1.635 2.270
Chain 1: 300 -8731.283 1.436 1.038
Chain 1: 400 -8108.720 1.096 1.038
Chain 1: 500 -8941.878 0.896 1.000
Chain 1: 600 -9420.724 0.755 1.000
Chain 1: 700 -7719.971 0.678 0.220
Chain 1: 800 -8037.529 0.599 0.220
Chain 1: 900 -7690.455 0.537 0.093
Chain 1: 1000 -8037.792 0.488 0.093
Chain 1: 1100 -7759.440 0.391 0.077
Chain 1: 1200 -7706.711 0.165 0.051
Chain 1: 1300 -7735.639 0.062 0.045
Chain 1: 1400 -7659.495 0.055 0.043
Chain 1: 1500 -7549.846 0.047 0.040
Chain 1: 1600 -7676.336 0.044 0.036
Chain 1: 1700 -7566.999 0.023 0.016
Chain 1: 1800 -7593.311 0.019 0.015
Chain 1: 1900 -7593.650 0.015 0.014
Chain 1: 2000 -7672.106 0.012 0.010
Chain 1: 2100 -7589.337 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00332 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.2 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86132.908 1.000 1.000
Chain 1: 200 -13636.566 3.158 5.316
Chain 1: 300 -9944.247 2.229 1.000
Chain 1: 400 -10970.099 1.695 1.000
Chain 1: 500 -8933.615 1.402 0.371
Chain 1: 600 -8427.311 1.178 0.371
Chain 1: 700 -8381.822 1.011 0.228
Chain 1: 800 -8603.180 0.888 0.228
Chain 1: 900 -8695.443 0.790 0.094
Chain 1: 1000 -8796.510 0.712 0.094
Chain 1: 1100 -8599.276 0.615 0.060 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8356.361 0.086 0.029
Chain 1: 1300 -8638.040 0.052 0.029
Chain 1: 1400 -8591.210 0.043 0.026
Chain 1: 1500 -8486.045 0.022 0.023
Chain 1: 1600 -8594.145 0.017 0.013
Chain 1: 1700 -8670.239 0.017 0.013
Chain 1: 1800 -8240.385 0.020 0.013
Chain 1: 1900 -8344.166 0.020 0.013
Chain 1: 2000 -8319.252 0.019 0.013
Chain 1: 2100 -8450.891 0.018 0.013
Chain 1: 2200 -8246.973 0.018 0.013
Chain 1: 2300 -8342.138 0.016 0.012
Chain 1: 2400 -8407.549 0.016 0.012
Chain 1: 2500 -8352.847 0.015 0.012
Chain 1: 2600 -8356.710 0.014 0.011
Chain 1: 2700 -8272.179 0.014 0.011
Chain 1: 2800 -8229.320 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003523 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.23 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8388735.020 1.000 1.000
Chain 1: 200 -1586229.427 2.644 4.288
Chain 1: 300 -891656.448 2.022 1.000
Chain 1: 400 -457857.089 1.754 1.000
Chain 1: 500 -358126.614 1.459 0.947
Chain 1: 600 -233141.850 1.305 0.947
Chain 1: 700 -119372.808 1.255 0.947
Chain 1: 800 -86560.363 1.145 0.947
Chain 1: 900 -66912.535 1.051 0.779
Chain 1: 1000 -51720.291 0.975 0.779
Chain 1: 1100 -39199.426 0.907 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38381.318 0.480 0.379
Chain 1: 1300 -26339.680 0.448 0.379
Chain 1: 1400 -26060.789 0.354 0.319
Chain 1: 1500 -22647.034 0.341 0.319
Chain 1: 1600 -21863.408 0.291 0.294
Chain 1: 1700 -20737.329 0.202 0.294
Chain 1: 1800 -20681.782 0.164 0.151
Chain 1: 1900 -21008.106 0.136 0.054
Chain 1: 2000 -19518.633 0.114 0.054
Chain 1: 2100 -19757.337 0.084 0.036
Chain 1: 2200 -19983.716 0.083 0.036
Chain 1: 2300 -19600.883 0.039 0.020
Chain 1: 2400 -19372.856 0.039 0.020
Chain 1: 2500 -19174.736 0.025 0.016
Chain 1: 2600 -18804.919 0.023 0.016
Chain 1: 2700 -18761.863 0.018 0.012
Chain 1: 2800 -18478.491 0.019 0.015
Chain 1: 2900 -18759.893 0.019 0.015
Chain 1: 3000 -18746.152 0.012 0.012
Chain 1: 3100 -18831.131 0.011 0.012
Chain 1: 3200 -18521.723 0.012 0.015
Chain 1: 3300 -18726.519 0.011 0.012
Chain 1: 3400 -18201.180 0.012 0.015
Chain 1: 3500 -18813.414 0.015 0.015
Chain 1: 3600 -18119.645 0.017 0.015
Chain 1: 3700 -18506.738 0.018 0.017
Chain 1: 3800 -17465.716 0.023 0.021
Chain 1: 3900 -17461.800 0.021 0.021
Chain 1: 4000 -17579.149 0.022 0.021
Chain 1: 4100 -17492.812 0.022 0.021
Chain 1: 4200 -17308.921 0.021 0.021
Chain 1: 4300 -17447.454 0.021 0.021
Chain 1: 4400 -17404.150 0.018 0.011
Chain 1: 4500 -17306.625 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001526 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.26 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49029.652 1.000 1.000
Chain 1: 200 -22570.746 1.086 1.172
Chain 1: 300 -17298.005 0.826 1.000
Chain 1: 400 -18739.225 0.638 1.000
Chain 1: 500 -12921.423 0.601 0.450
Chain 1: 600 -11905.239 0.515 0.450
Chain 1: 700 -14958.008 0.471 0.305
Chain 1: 800 -16332.576 0.422 0.305
Chain 1: 900 -15410.370 0.382 0.204
Chain 1: 1000 -11264.962 0.381 0.305
Chain 1: 1100 -9973.374 0.294 0.204
Chain 1: 1200 -13018.972 0.200 0.204
Chain 1: 1300 -13246.799 0.171 0.130
Chain 1: 1400 -11118.517 0.182 0.191
Chain 1: 1500 -11681.673 0.142 0.130
Chain 1: 1600 -10488.212 0.145 0.130
Chain 1: 1700 -11094.875 0.130 0.114
Chain 1: 1800 -13170.444 0.137 0.130
Chain 1: 1900 -12452.278 0.137 0.130
Chain 1: 2000 -9744.927 0.128 0.130
Chain 1: 2100 -10983.638 0.127 0.114
Chain 1: 2200 -11180.495 0.105 0.113
Chain 1: 2300 -10456.654 0.110 0.113
Chain 1: 2400 -11010.274 0.096 0.069
Chain 1: 2500 -14206.031 0.114 0.113
Chain 1: 2600 -10721.956 0.135 0.113
Chain 1: 2700 -11847.404 0.139 0.113
Chain 1: 2800 -10262.805 0.138 0.113
Chain 1: 2900 -11941.357 0.147 0.141
Chain 1: 3000 -9162.711 0.149 0.141
Chain 1: 3100 -10064.002 0.147 0.141
Chain 1: 3200 -16072.648 0.183 0.154
Chain 1: 3300 -19203.946 0.192 0.163
Chain 1: 3400 -14716.510 0.217 0.225
Chain 1: 3500 -9017.683 0.258 0.303
Chain 1: 3600 -9241.565 0.228 0.163
Chain 1: 3700 -9803.864 0.224 0.163
Chain 1: 3800 -13839.893 0.238 0.292
Chain 1: 3900 -9450.973 0.270 0.303
Chain 1: 4000 -11169.075 0.255 0.292
Chain 1: 4100 -9686.154 0.262 0.292
Chain 1: 4200 -8766.602 0.235 0.163
Chain 1: 4300 -10181.309 0.233 0.154
Chain 1: 4400 -9197.251 0.213 0.153
Chain 1: 4500 -9199.403 0.150 0.139
Chain 1: 4600 -13105.282 0.177 0.153
Chain 1: 4700 -11070.209 0.190 0.154
Chain 1: 4800 -8756.047 0.187 0.154
Chain 1: 4900 -10552.777 0.157 0.154
Chain 1: 5000 -8551.717 0.165 0.170
Chain 1: 5100 -8844.698 0.153 0.170
Chain 1: 5200 -12985.128 0.175 0.184
Chain 1: 5300 -8318.733 0.217 0.234
Chain 1: 5400 -8355.816 0.207 0.234
Chain 1: 5500 -8621.913 0.210 0.234
Chain 1: 5600 -8487.571 0.182 0.184
Chain 1: 5700 -13901.307 0.202 0.234
Chain 1: 5800 -14719.585 0.181 0.170
Chain 1: 5900 -13051.992 0.177 0.128
Chain 1: 6000 -8723.043 0.203 0.128
Chain 1: 6100 -9607.108 0.209 0.128
Chain 1: 6200 -8284.672 0.193 0.128
Chain 1: 6300 -10742.350 0.160 0.128
Chain 1: 6400 -12020.785 0.170 0.128
Chain 1: 6500 -12349.513 0.170 0.128
Chain 1: 6600 -8527.652 0.213 0.160
Chain 1: 6700 -12272.448 0.205 0.160
Chain 1: 6800 -8335.386 0.246 0.229
Chain 1: 6900 -8446.971 0.235 0.229
Chain 1: 7000 -10156.597 0.202 0.168
Chain 1: 7100 -16436.451 0.231 0.229
Chain 1: 7200 -8360.920 0.312 0.305
Chain 1: 7300 -10357.554 0.308 0.305
Chain 1: 7400 -8184.866 0.324 0.305
Chain 1: 7500 -10722.142 0.345 0.305
Chain 1: 7600 -8713.149 0.323 0.265
Chain 1: 7700 -8364.433 0.297 0.237
Chain 1: 7800 -8743.853 0.254 0.231
Chain 1: 7900 -8191.223 0.259 0.231
Chain 1: 8000 -8303.395 0.244 0.231
Chain 1: 8100 -9062.962 0.214 0.193
Chain 1: 8200 -9001.637 0.118 0.084
Chain 1: 8300 -8690.570 0.103 0.067
Chain 1: 8400 -12572.375 0.107 0.067
Chain 1: 8500 -8048.823 0.139 0.067
Chain 1: 8600 -8312.271 0.119 0.043
Chain 1: 8700 -8625.877 0.119 0.043
Chain 1: 8800 -8445.073 0.117 0.036
Chain 1: 8900 -8959.172 0.116 0.036
Chain 1: 9000 -11237.713 0.135 0.057
Chain 1: 9100 -8156.448 0.164 0.057
Chain 1: 9200 -8987.272 0.173 0.092
Chain 1: 9300 -9347.730 0.173 0.092
Chain 1: 9400 -8119.212 0.157 0.092
Chain 1: 9500 -8433.463 0.105 0.057
Chain 1: 9600 -8486.445 0.102 0.057
Chain 1: 9700 -8764.143 0.102 0.057
Chain 1: 9800 -9641.648 0.109 0.091
Chain 1: 9900 -8046.577 0.123 0.092
Chain 1: 10000 -8239.446 0.105 0.091
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001396 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.96 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58128.951 1.000 1.000
Chain 1: 200 -17874.895 1.626 2.252
Chain 1: 300 -8773.647 1.430 1.037
Chain 1: 400 -8212.436 1.089 1.037
Chain 1: 500 -8512.405 0.879 1.000
Chain 1: 600 -8446.212 0.733 1.000
Chain 1: 700 -8188.812 0.633 0.068
Chain 1: 800 -8375.752 0.557 0.068
Chain 1: 900 -8004.659 0.500 0.046
Chain 1: 1000 -7667.625 0.454 0.046
Chain 1: 1100 -7678.609 0.355 0.044
Chain 1: 1200 -8161.546 0.135 0.044
Chain 1: 1300 -7813.849 0.036 0.044
Chain 1: 1400 -7924.331 0.031 0.035
Chain 1: 1500 -7632.026 0.031 0.038
Chain 1: 1600 -7613.548 0.030 0.038
Chain 1: 1700 -7582.469 0.028 0.038
Chain 1: 1800 -7624.006 0.026 0.038
Chain 1: 1900 -7635.038 0.021 0.014
Chain 1: 2000 -7618.868 0.017 0.005 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002518 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.18 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86924.226 1.000 1.000
Chain 1: 200 -13661.565 3.181 5.363
Chain 1: 300 -9898.002 2.248 1.000
Chain 1: 400 -11188.756 1.715 1.000
Chain 1: 500 -8939.890 1.422 0.380
Chain 1: 600 -8337.326 1.197 0.380
Chain 1: 700 -8333.286 1.026 0.252
Chain 1: 800 -8571.613 0.901 0.252
Chain 1: 900 -8632.395 0.802 0.115
Chain 1: 1000 -8704.304 0.723 0.115
Chain 1: 1100 -8628.977 0.623 0.072 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8239.752 0.092 0.047
Chain 1: 1300 -8534.621 0.057 0.035
Chain 1: 1400 -8410.485 0.047 0.028
Chain 1: 1500 -8398.091 0.022 0.015
Chain 1: 1600 -8507.841 0.016 0.013
Chain 1: 1700 -8562.156 0.017 0.013
Chain 1: 1800 -8118.841 0.020 0.013
Chain 1: 1900 -8223.722 0.020 0.013
Chain 1: 2000 -8206.901 0.020 0.013
Chain 1: 2100 -8331.664 0.020 0.015
Chain 1: 2200 -8118.690 0.018 0.015
Chain 1: 2300 -8219.821 0.016 0.013
Chain 1: 2400 -8282.621 0.015 0.013
Chain 1: 2500 -8232.379 0.016 0.013
Chain 1: 2600 -8245.777 0.014 0.012
Chain 1: 2700 -8152.707 0.015 0.012
Chain 1: 2800 -8098.542 0.010 0.011
Chain 1: 2900 -8199.833 0.010 0.011
Chain 1: 3000 -8040.985 0.012 0.012
Chain 1: 3100 -8183.662 0.012 0.012
Chain 1: 3200 -8053.110 0.011 0.012
Chain 1: 3300 -8090.779 0.010 0.011
Chain 1: 3400 -8238.885 0.011 0.012
Chain 1: 3500 -8231.382 0.011 0.012
Chain 1: 3600 -8007.823 0.014 0.016
Chain 1: 3700 -8161.127 0.014 0.017
Chain 1: 3800 -8012.341 0.015 0.018
Chain 1: 3900 -7945.346 0.015 0.018
Chain 1: 4000 -8055.760 0.014 0.017
Chain 1: 4100 -8020.137 0.013 0.016
Chain 1: 4200 -8006.028 0.012 0.014
Chain 1: 4300 -8039.458 0.012 0.014
Chain 1: 4400 -7996.246 0.010 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003314 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.14 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8397389.603 1.000 1.000
Chain 1: 200 -1588761.163 2.643 4.285
Chain 1: 300 -892383.796 2.022 1.000
Chain 1: 400 -457797.188 1.754 1.000
Chain 1: 500 -357602.747 1.459 0.949
Chain 1: 600 -232518.935 1.306 0.949
Chain 1: 700 -119086.738 1.255 0.949
Chain 1: 800 -86306.307 1.146 0.949
Chain 1: 900 -66735.925 1.051 0.780
Chain 1: 1000 -51611.068 0.975 0.780
Chain 1: 1100 -39140.843 0.907 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38336.949 0.481 0.380
Chain 1: 1300 -26343.938 0.448 0.380
Chain 1: 1400 -26071.915 0.354 0.319
Chain 1: 1500 -22669.094 0.341 0.319
Chain 1: 1600 -21889.317 0.291 0.293
Chain 1: 1700 -20768.856 0.201 0.293
Chain 1: 1800 -20714.790 0.163 0.150
Chain 1: 1900 -21041.542 0.136 0.054
Chain 1: 2000 -19553.999 0.114 0.054
Chain 1: 2100 -19792.834 0.083 0.036
Chain 1: 2200 -20018.841 0.082 0.036
Chain 1: 2300 -19636.243 0.039 0.019
Chain 1: 2400 -19408.164 0.039 0.019
Chain 1: 2500 -19209.648 0.025 0.016
Chain 1: 2600 -18839.823 0.023 0.016
Chain 1: 2700 -18796.836 0.018 0.012
Chain 1: 2800 -18513.076 0.019 0.015
Chain 1: 2900 -18794.626 0.019 0.015
Chain 1: 3000 -18781.003 0.012 0.012
Chain 1: 3100 -18865.960 0.011 0.012
Chain 1: 3200 -18556.407 0.012 0.015
Chain 1: 3300 -18761.375 0.011 0.012
Chain 1: 3400 -18235.539 0.012 0.015
Chain 1: 3500 -18848.306 0.015 0.015
Chain 1: 3600 -18153.954 0.017 0.015
Chain 1: 3700 -18541.406 0.018 0.017
Chain 1: 3800 -17499.266 0.023 0.021
Chain 1: 3900 -17495.290 0.021 0.021
Chain 1: 4000 -17612.715 0.022 0.021
Chain 1: 4100 -17526.224 0.022 0.021
Chain 1: 4200 -17342.161 0.021 0.021
Chain 1: 4300 -17480.876 0.021 0.021
Chain 1: 4400 -17437.413 0.018 0.011
Chain 1: 4500 -17339.794 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001305 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.05 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49301.047 1.000 1.000
Chain 1: 200 -23164.854 1.064 1.128
Chain 1: 300 -17574.901 0.815 1.000
Chain 1: 400 -16781.326 0.623 1.000
Chain 1: 500 -17716.484 0.509 0.318
Chain 1: 600 -12155.232 0.501 0.458
Chain 1: 700 -17467.622 0.473 0.318
Chain 1: 800 -14350.204 0.441 0.318
Chain 1: 900 -12430.890 0.409 0.304
Chain 1: 1000 -12541.962 0.369 0.304
Chain 1: 1100 -11228.514 0.281 0.217
Chain 1: 1200 -25081.303 0.223 0.217
Chain 1: 1300 -10673.995 0.326 0.217
Chain 1: 1400 -12559.828 0.336 0.217
Chain 1: 1500 -12511.744 0.332 0.217
Chain 1: 1600 -15625.146 0.306 0.199
Chain 1: 1700 -10260.766 0.328 0.199
Chain 1: 1800 -11814.334 0.319 0.154
Chain 1: 1900 -10674.986 0.314 0.150
Chain 1: 2000 -11943.850 0.324 0.150
Chain 1: 2100 -10548.532 0.325 0.150
Chain 1: 2200 -19661.101 0.317 0.150
Chain 1: 2300 -9895.996 0.280 0.150
Chain 1: 2400 -12866.280 0.288 0.199
Chain 1: 2500 -10929.112 0.306 0.199
Chain 1: 2600 -10014.944 0.295 0.177
Chain 1: 2700 -9415.373 0.249 0.132
Chain 1: 2800 -11664.673 0.255 0.177
Chain 1: 2900 -10188.113 0.259 0.177
Chain 1: 3000 -9385.506 0.257 0.177
Chain 1: 3100 -10816.150 0.257 0.177
Chain 1: 3200 -11619.279 0.217 0.145
Chain 1: 3300 -13896.075 0.135 0.145
Chain 1: 3400 -12471.866 0.123 0.132
Chain 1: 3500 -10850.531 0.121 0.132
Chain 1: 3600 -10192.528 0.118 0.132
Chain 1: 3700 -9393.482 0.120 0.132
Chain 1: 3800 -9176.104 0.103 0.114
Chain 1: 3900 -9832.286 0.095 0.086
Chain 1: 4000 -9892.494 0.087 0.085
Chain 1: 4100 -9107.352 0.083 0.085
Chain 1: 4200 -12595.936 0.104 0.086
Chain 1: 4300 -14268.958 0.099 0.086
Chain 1: 4400 -8939.996 0.147 0.086
Chain 1: 4500 -11044.410 0.151 0.086
Chain 1: 4600 -9373.055 0.163 0.117
Chain 1: 4700 -8799.550 0.161 0.117
Chain 1: 4800 -13935.783 0.195 0.178
Chain 1: 4900 -10263.744 0.224 0.191
Chain 1: 5000 -10660.306 0.227 0.191
Chain 1: 5100 -9536.937 0.231 0.191
Chain 1: 5200 -16158.036 0.244 0.191
Chain 1: 5300 -14780.660 0.241 0.191
Chain 1: 5400 -9157.555 0.243 0.191
Chain 1: 5500 -10075.105 0.233 0.178
Chain 1: 5600 -9525.971 0.221 0.118
Chain 1: 5700 -9780.816 0.217 0.118
Chain 1: 5800 -9563.769 0.183 0.093
Chain 1: 5900 -9650.943 0.148 0.091
Chain 1: 6000 -10476.093 0.152 0.091
Chain 1: 6100 -14573.652 0.168 0.091
Chain 1: 6200 -9095.342 0.188 0.091
Chain 1: 6300 -10605.979 0.193 0.091
Chain 1: 6400 -9859.459 0.139 0.079
Chain 1: 6500 -9570.495 0.133 0.076
Chain 1: 6600 -8853.345 0.135 0.079
Chain 1: 6700 -8746.274 0.134 0.079
Chain 1: 6800 -8846.700 0.132 0.079
Chain 1: 6900 -13788.827 0.167 0.081
Chain 1: 7000 -12655.786 0.168 0.090
Chain 1: 7100 -8727.152 0.185 0.090
Chain 1: 7200 -8595.242 0.127 0.081
Chain 1: 7300 -11863.030 0.140 0.081
Chain 1: 7400 -14944.728 0.153 0.090
Chain 1: 7500 -10028.065 0.199 0.206
Chain 1: 7600 -8769.192 0.205 0.206
Chain 1: 7700 -10196.416 0.218 0.206
Chain 1: 7800 -11882.527 0.231 0.206
Chain 1: 7900 -8941.189 0.228 0.206
Chain 1: 8000 -9161.159 0.222 0.206
Chain 1: 8100 -9278.273 0.178 0.144
Chain 1: 8200 -10181.035 0.185 0.144
Chain 1: 8300 -8567.334 0.176 0.144
Chain 1: 8400 -8572.481 0.156 0.142
Chain 1: 8500 -10151.449 0.122 0.142
Chain 1: 8600 -13300.814 0.132 0.142
Chain 1: 8700 -9281.986 0.161 0.156
Chain 1: 8800 -8700.085 0.154 0.156
Chain 1: 8900 -10659.096 0.139 0.156
Chain 1: 9000 -9129.204 0.153 0.168
Chain 1: 9100 -8948.086 0.154 0.168
Chain 1: 9200 -10398.439 0.159 0.168
Chain 1: 9300 -8586.037 0.161 0.168
Chain 1: 9400 -8859.189 0.165 0.168
Chain 1: 9500 -9934.784 0.160 0.168
Chain 1: 9600 -11535.887 0.150 0.139
Chain 1: 9700 -10686.166 0.115 0.139
Chain 1: 9800 -10805.862 0.109 0.139
Chain 1: 9900 -9887.361 0.100 0.108
Chain 1: 10000 -9116.347 0.092 0.093
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001422 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.22 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -47849.987 1.000 1.000
Chain 1: 200 -16140.497 1.482 1.965
Chain 1: 300 -8801.514 1.266 1.000
Chain 1: 400 -8651.550 0.954 1.000
Chain 1: 500 -8845.312 0.768 0.834
Chain 1: 600 -9157.837 0.645 0.834
Chain 1: 700 -8164.362 0.570 0.122
Chain 1: 800 -8350.426 0.502 0.122
Chain 1: 900 -8116.085 0.449 0.034
Chain 1: 1000 -8040.818 0.405 0.034
Chain 1: 1100 -7857.298 0.308 0.029
Chain 1: 1200 -8002.870 0.113 0.023
Chain 1: 1300 -7663.213 0.034 0.023
Chain 1: 1400 -8042.530 0.037 0.029
Chain 1: 1500 -7722.862 0.039 0.034
Chain 1: 1600 -7886.799 0.038 0.029
Chain 1: 1700 -7657.638 0.029 0.029
Chain 1: 1800 -7716.970 0.027 0.029
Chain 1: 1900 -7701.957 0.024 0.023
Chain 1: 2000 -7787.870 0.025 0.023
Chain 1: 2100 -7704.131 0.023 0.021
Chain 1: 2200 -7840.616 0.023 0.021
Chain 1: 2300 -7701.574 0.021 0.018
Chain 1: 2400 -7761.112 0.017 0.017
Chain 1: 2500 -7700.165 0.013 0.011
Chain 1: 2600 -7646.921 0.012 0.011
Chain 1: 2700 -7572.746 0.010 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002553 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.53 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85765.443 1.000 1.000
Chain 1: 200 -13882.804 3.089 5.178
Chain 1: 300 -10235.473 2.178 1.000
Chain 1: 400 -11070.121 1.652 1.000
Chain 1: 500 -9092.833 1.365 0.356
Chain 1: 600 -8679.501 1.146 0.356
Chain 1: 700 -8746.038 0.983 0.217
Chain 1: 800 -9259.608 0.867 0.217
Chain 1: 900 -8923.899 0.775 0.075
Chain 1: 1000 -8779.106 0.699 0.075
Chain 1: 1100 -9064.972 0.602 0.055 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8671.546 0.089 0.048
Chain 1: 1300 -8989.275 0.057 0.045
Chain 1: 1400 -8938.851 0.050 0.038
Chain 1: 1500 -8781.769 0.030 0.035
Chain 1: 1600 -8893.397 0.027 0.032
Chain 1: 1700 -8974.073 0.027 0.032
Chain 1: 1800 -8550.518 0.026 0.032
Chain 1: 1900 -8651.489 0.024 0.018
Chain 1: 2000 -8626.074 0.022 0.018
Chain 1: 2100 -8751.392 0.020 0.014
Chain 1: 2200 -8554.602 0.018 0.014
Chain 1: 2300 -8646.319 0.016 0.013
Chain 1: 2400 -8715.124 0.016 0.013
Chain 1: 2500 -8661.395 0.015 0.012
Chain 1: 2600 -8662.761 0.014 0.011
Chain 1: 2700 -8579.454 0.014 0.011
Chain 1: 2800 -8539.357 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003175 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.75 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8378877.428 1.000 1.000
Chain 1: 200 -1581020.526 2.650 4.300
Chain 1: 300 -890529.006 2.025 1.000
Chain 1: 400 -457940.297 1.755 1.000
Chain 1: 500 -358790.733 1.459 0.945
Chain 1: 600 -233823.516 1.305 0.945
Chain 1: 700 -119888.465 1.254 0.945
Chain 1: 800 -87034.509 1.145 0.945
Chain 1: 900 -67331.496 1.050 0.775
Chain 1: 1000 -52092.688 0.974 0.775
Chain 1: 1100 -39532.505 0.906 0.534 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38706.139 0.478 0.377
Chain 1: 1300 -26623.400 0.446 0.377
Chain 1: 1400 -26339.224 0.353 0.318
Chain 1: 1500 -22915.922 0.340 0.318
Chain 1: 1600 -22129.434 0.290 0.293
Chain 1: 1700 -20998.365 0.201 0.293
Chain 1: 1800 -20941.531 0.163 0.149
Chain 1: 1900 -21267.832 0.135 0.054
Chain 1: 2000 -19776.220 0.114 0.054
Chain 1: 2100 -20014.763 0.083 0.036
Chain 1: 2200 -20241.673 0.082 0.036
Chain 1: 2300 -19858.446 0.039 0.019
Chain 1: 2400 -19630.422 0.039 0.019
Chain 1: 2500 -19432.589 0.025 0.015
Chain 1: 2600 -19062.514 0.023 0.015
Chain 1: 2700 -19019.434 0.018 0.012
Chain 1: 2800 -18736.250 0.019 0.015
Chain 1: 2900 -19017.667 0.019 0.015
Chain 1: 3000 -19003.796 0.012 0.012
Chain 1: 3100 -19088.787 0.011 0.012
Chain 1: 3200 -18779.395 0.011 0.015
Chain 1: 3300 -18984.199 0.011 0.012
Chain 1: 3400 -18458.965 0.012 0.015
Chain 1: 3500 -19071.142 0.014 0.015
Chain 1: 3600 -18377.494 0.016 0.015
Chain 1: 3700 -18764.545 0.018 0.016
Chain 1: 3800 -17723.758 0.022 0.021
Chain 1: 3900 -17719.923 0.021 0.021
Chain 1: 4000 -17837.201 0.022 0.021
Chain 1: 4100 -17750.916 0.022 0.021
Chain 1: 4200 -17567.079 0.021 0.021
Chain 1: 4300 -17705.528 0.021 0.021
Chain 1: 4400 -17662.268 0.018 0.010
Chain 1: 4500 -17564.806 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001305 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.05 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12790.573 1.000 1.000
Chain 1: 200 -9696.548 0.660 1.000
Chain 1: 300 -8288.568 0.496 0.319
Chain 1: 400 -8488.181 0.378 0.319
Chain 1: 500 -8381.091 0.305 0.170
Chain 1: 600 -8216.771 0.258 0.170
Chain 1: 700 -8089.326 0.223 0.024
Chain 1: 800 -8071.943 0.195 0.024
Chain 1: 900 -8271.554 0.176 0.024
Chain 1: 1000 -8199.185 0.160 0.024
Chain 1: 1100 -8165.742 0.060 0.020
Chain 1: 1200 -8162.722 0.028 0.016
Chain 1: 1300 -8057.254 0.012 0.013
Chain 1: 1400 -8063.319 0.010 0.013
Chain 1: 1500 -8178.072 0.010 0.013
Chain 1: 1600 -8103.462 0.009 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001378 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.78 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -59020.265 1.000 1.000
Chain 1: 200 -18253.463 1.617 2.233
Chain 1: 300 -9102.341 1.413 1.005
Chain 1: 400 -8355.904 1.082 1.005
Chain 1: 500 -7846.871 0.879 1.000
Chain 1: 600 -8582.074 0.746 1.000
Chain 1: 700 -8163.872 0.647 0.089
Chain 1: 800 -8339.253 0.569 0.089
Chain 1: 900 -7724.670 0.514 0.086
Chain 1: 1000 -7644.596 0.464 0.086
Chain 1: 1100 -7685.819 0.365 0.080
Chain 1: 1200 -7528.769 0.143 0.065
Chain 1: 1300 -7779.211 0.046 0.051
Chain 1: 1400 -7871.033 0.038 0.032
Chain 1: 1500 -7550.256 0.036 0.032
Chain 1: 1600 -7716.947 0.030 0.022
Chain 1: 1700 -7517.243 0.027 0.022
Chain 1: 1800 -7569.687 0.026 0.022
Chain 1: 1900 -7560.332 0.018 0.021
Chain 1: 2000 -7617.911 0.018 0.021
Chain 1: 2100 -7586.720 0.018 0.021
Chain 1: 2200 -7844.912 0.019 0.022
Chain 1: 2300 -7597.310 0.019 0.022
Chain 1: 2400 -7508.705 0.019 0.022
Chain 1: 2500 -7400.736 0.016 0.015
Chain 1: 2600 -7516.159 0.015 0.015
Chain 1: 2700 -7489.420 0.013 0.012
Chain 1: 2800 -7510.000 0.013 0.012
Chain 1: 2900 -7360.334 0.015 0.015
Chain 1: 3000 -7519.633 0.016 0.015
Chain 1: 3100 -7512.665 0.016 0.015
Chain 1: 3200 -7738.093 0.015 0.015
Chain 1: 3300 -7409.738 0.016 0.015
Chain 1: 3400 -7687.861 0.019 0.020
Chain 1: 3500 -7432.007 0.021 0.021
Chain 1: 3600 -7490.403 0.020 0.021
Chain 1: 3700 -7441.722 0.020 0.021
Chain 1: 3800 -7450.785 0.020 0.021
Chain 1: 3900 -7418.272 0.019 0.021
Chain 1: 4000 -7387.199 0.017 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002561 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.61 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86305.240 1.000 1.000
Chain 1: 200 -14022.598 3.077 5.155
Chain 1: 300 -10266.488 2.174 1.000
Chain 1: 400 -12037.315 1.667 1.000
Chain 1: 500 -8867.192 1.405 0.366
Chain 1: 600 -9865.814 1.188 0.366
Chain 1: 700 -8755.943 1.036 0.358
Chain 1: 800 -9353.640 0.915 0.358
Chain 1: 900 -9058.365 0.817 0.147
Chain 1: 1000 -9188.657 0.736 0.147
Chain 1: 1100 -8978.438 0.639 0.127 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8558.676 0.128 0.101
Chain 1: 1300 -8884.940 0.095 0.064
Chain 1: 1400 -8828.464 0.081 0.049
Chain 1: 1500 -8742.608 0.046 0.037
Chain 1: 1600 -8841.429 0.037 0.033
Chain 1: 1700 -8899.399 0.025 0.023
Chain 1: 1800 -8446.966 0.024 0.023
Chain 1: 1900 -8555.179 0.022 0.014
Chain 1: 2000 -8554.939 0.021 0.013
Chain 1: 2100 -8723.522 0.021 0.013
Chain 1: 2200 -8451.215 0.019 0.013
Chain 1: 2300 -8631.705 0.017 0.013
Chain 1: 2400 -8450.902 0.019 0.019
Chain 1: 2500 -8527.649 0.019 0.019
Chain 1: 2600 -8438.685 0.019 0.019
Chain 1: 2700 -8471.755 0.018 0.019
Chain 1: 2800 -8423.513 0.014 0.013
Chain 1: 2900 -8534.941 0.014 0.013
Chain 1: 3000 -8473.181 0.014 0.013
Chain 1: 3100 -8415.788 0.013 0.011
Chain 1: 3200 -8388.875 0.010 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003587 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.87 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8416397.601 1.000 1.000
Chain 1: 200 -1584929.568 2.655 4.310
Chain 1: 300 -892198.990 2.029 1.000
Chain 1: 400 -459073.515 1.758 1.000
Chain 1: 500 -359447.978 1.461 0.943
Chain 1: 600 -234205.416 1.307 0.943
Chain 1: 700 -120100.955 1.256 0.943
Chain 1: 800 -87257.185 1.146 0.943
Chain 1: 900 -67524.028 1.051 0.776
Chain 1: 1000 -52279.887 0.975 0.776
Chain 1: 1100 -39717.831 0.907 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38892.385 0.478 0.376
Chain 1: 1300 -26796.871 0.445 0.376
Chain 1: 1400 -26513.789 0.352 0.316
Chain 1: 1500 -23088.064 0.339 0.316
Chain 1: 1600 -22301.848 0.289 0.292
Chain 1: 1700 -21168.657 0.200 0.292
Chain 1: 1800 -21111.655 0.162 0.148
Chain 1: 1900 -21438.445 0.135 0.054
Chain 1: 2000 -19944.870 0.113 0.054
Chain 1: 2100 -20183.440 0.083 0.035
Chain 1: 2200 -20411.033 0.082 0.035
Chain 1: 2300 -20027.092 0.038 0.019
Chain 1: 2400 -19798.858 0.038 0.019
Chain 1: 2500 -19601.148 0.025 0.015
Chain 1: 2600 -19230.353 0.023 0.015
Chain 1: 2700 -19186.999 0.018 0.012
Chain 1: 2800 -18903.684 0.019 0.015
Chain 1: 2900 -19185.318 0.019 0.015
Chain 1: 3000 -19171.347 0.012 0.012
Chain 1: 3100 -19256.498 0.011 0.012
Chain 1: 3200 -18946.616 0.011 0.015
Chain 1: 3300 -19151.774 0.011 0.012
Chain 1: 3400 -18625.838 0.012 0.015
Chain 1: 3500 -19239.067 0.014 0.015
Chain 1: 3600 -18543.961 0.016 0.015
Chain 1: 3700 -18932.164 0.018 0.016
Chain 1: 3800 -17889.158 0.022 0.021
Chain 1: 3900 -17885.264 0.021 0.021
Chain 1: 4000 -18002.538 0.021 0.021
Chain 1: 4100 -17916.194 0.022 0.021
Chain 1: 4200 -17731.824 0.021 0.021
Chain 1: 4300 -17870.613 0.021 0.021
Chain 1: 4400 -17826.946 0.018 0.010
Chain 1: 4500 -17729.413 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001492 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.92 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48679.645 1.000 1.000
Chain 1: 200 -17890.084 1.361 1.721
Chain 1: 300 -35163.073 1.071 1.000
Chain 1: 400 -15921.216 1.105 1.209
Chain 1: 500 -17536.639 0.903 1.000
Chain 1: 600 -15264.277 0.777 1.000
Chain 1: 700 -15027.253 0.668 0.491
Chain 1: 800 -16193.452 0.594 0.491
Chain 1: 900 -11415.657 0.574 0.419
Chain 1: 1000 -13465.539 0.532 0.419
Chain 1: 1100 -15964.155 0.448 0.157
Chain 1: 1200 -13237.353 0.296 0.157
Chain 1: 1300 -12555.376 0.252 0.152
Chain 1: 1400 -10392.241 0.152 0.152
Chain 1: 1500 -18052.447 0.186 0.157
Chain 1: 1600 -9818.974 0.255 0.206
Chain 1: 1700 -9411.922 0.257 0.206
Chain 1: 1800 -11501.025 0.268 0.206
Chain 1: 1900 -10162.712 0.240 0.182
Chain 1: 2000 -12738.784 0.245 0.202
Chain 1: 2100 -19099.644 0.262 0.206
Chain 1: 2200 -10869.447 0.317 0.208
Chain 1: 2300 -9899.484 0.322 0.208
Chain 1: 2400 -9747.501 0.303 0.202
Chain 1: 2500 -9895.340 0.262 0.182
Chain 1: 2600 -10111.882 0.180 0.132
Chain 1: 2700 -15226.116 0.209 0.182
Chain 1: 2800 -8910.666 0.262 0.202
Chain 1: 2900 -9031.335 0.250 0.202
Chain 1: 3000 -9839.514 0.238 0.098
Chain 1: 3100 -9788.287 0.205 0.082
Chain 1: 3200 -9861.152 0.130 0.021
Chain 1: 3300 -8785.243 0.133 0.021
Chain 1: 3400 -9438.900 0.138 0.069
Chain 1: 3500 -8974.254 0.142 0.069
Chain 1: 3600 -9691.224 0.147 0.074
Chain 1: 3700 -9644.385 0.114 0.069
Chain 1: 3800 -8519.485 0.056 0.069
Chain 1: 3900 -10102.445 0.071 0.074
Chain 1: 4000 -10242.262 0.064 0.069
Chain 1: 4100 -9736.017 0.068 0.069
Chain 1: 4200 -11497.998 0.083 0.074
Chain 1: 4300 -15526.372 0.097 0.074
Chain 1: 4400 -8926.583 0.164 0.132
Chain 1: 4500 -9465.266 0.164 0.132
Chain 1: 4600 -9053.718 0.161 0.132
Chain 1: 4700 -12813.558 0.190 0.153
Chain 1: 4800 -8493.632 0.228 0.157
Chain 1: 4900 -8918.630 0.217 0.153
Chain 1: 5000 -13516.824 0.250 0.259
Chain 1: 5100 -8461.976 0.304 0.293
Chain 1: 5200 -8769.473 0.292 0.293
Chain 1: 5300 -9443.144 0.274 0.293
Chain 1: 5400 -10413.511 0.209 0.093
Chain 1: 5500 -9173.786 0.217 0.135
Chain 1: 5600 -8288.658 0.223 0.135
Chain 1: 5700 -14270.046 0.235 0.135
Chain 1: 5800 -8429.422 0.254 0.135
Chain 1: 5900 -9689.908 0.262 0.135
Chain 1: 6000 -8315.598 0.245 0.135
Chain 1: 6100 -11651.908 0.214 0.135
Chain 1: 6200 -8588.108 0.246 0.165
Chain 1: 6300 -10103.595 0.254 0.165
Chain 1: 6400 -8195.813 0.268 0.233
Chain 1: 6500 -8902.972 0.262 0.233
Chain 1: 6600 -8269.788 0.259 0.233
Chain 1: 6700 -8769.603 0.223 0.165
Chain 1: 6800 -8065.138 0.162 0.150
Chain 1: 6900 -8170.892 0.150 0.150
Chain 1: 7000 -9409.521 0.147 0.132
Chain 1: 7100 -8108.467 0.134 0.132
Chain 1: 7200 -11292.480 0.127 0.132
Chain 1: 7300 -9962.396 0.125 0.132
Chain 1: 7400 -8176.613 0.124 0.132
Chain 1: 7500 -8172.770 0.116 0.132
Chain 1: 7600 -8712.297 0.115 0.132
Chain 1: 7700 -10173.449 0.123 0.134
Chain 1: 7800 -8077.539 0.140 0.144
Chain 1: 7900 -8097.631 0.139 0.144
Chain 1: 8000 -9628.284 0.142 0.159
Chain 1: 8100 -9344.836 0.129 0.144
Chain 1: 8200 -7985.526 0.118 0.144
Chain 1: 8300 -10953.647 0.132 0.159
Chain 1: 8400 -9476.174 0.125 0.156
Chain 1: 8500 -8049.078 0.143 0.159
Chain 1: 8600 -8180.983 0.139 0.159
Chain 1: 8700 -7921.881 0.127 0.159
Chain 1: 8800 -8925.770 0.113 0.156
Chain 1: 8900 -12439.609 0.141 0.159
Chain 1: 9000 -10278.581 0.146 0.170
Chain 1: 9100 -8100.164 0.170 0.177
Chain 1: 9200 -11629.026 0.183 0.210
Chain 1: 9300 -8139.153 0.199 0.210
Chain 1: 9400 -8579.887 0.188 0.210
Chain 1: 9500 -7826.879 0.180 0.210
Chain 1: 9600 -8270.915 0.184 0.210
Chain 1: 9700 -9561.023 0.194 0.210
Chain 1: 9800 -8176.110 0.200 0.210
Chain 1: 9900 -10676.778 0.195 0.210
Chain 1: 10000 -7861.942 0.210 0.234
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001378 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.78 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58103.627 1.000 1.000
Chain 1: 200 -17538.694 1.656 2.313
Chain 1: 300 -8581.058 1.452 1.044
Chain 1: 400 -8152.208 1.102 1.044
Chain 1: 500 -8064.200 0.884 1.000
Chain 1: 600 -8619.276 0.747 1.000
Chain 1: 700 -8121.557 0.649 0.064
Chain 1: 800 -8146.216 0.569 0.064
Chain 1: 900 -7710.392 0.512 0.061
Chain 1: 1000 -7751.757 0.461 0.061
Chain 1: 1100 -7666.481 0.362 0.057
Chain 1: 1200 -7550.889 0.132 0.053
Chain 1: 1300 -7603.653 0.029 0.015
Chain 1: 1400 -7776.477 0.026 0.015
Chain 1: 1500 -7544.107 0.028 0.022
Chain 1: 1600 -7603.335 0.022 0.015
Chain 1: 1700 -7479.803 0.018 0.015
Chain 1: 1800 -7513.189 0.018 0.015
Chain 1: 1900 -7529.230 0.012 0.011
Chain 1: 2000 -7560.752 0.012 0.011
Chain 1: 2100 -7569.812 0.011 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00299 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.9 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86055.445 1.000 1.000
Chain 1: 200 -13302.288 3.235 5.469
Chain 1: 300 -9668.345 2.282 1.000
Chain 1: 400 -10676.777 1.735 1.000
Chain 1: 500 -8443.479 1.441 0.376
Chain 1: 600 -8138.245 1.207 0.376
Chain 1: 700 -8368.165 1.038 0.264
Chain 1: 800 -8524.751 0.911 0.264
Chain 1: 900 -8506.102 0.810 0.094
Chain 1: 1000 -8210.905 0.733 0.094
Chain 1: 1100 -8465.661 0.636 0.038 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8109.779 0.093 0.038
Chain 1: 1300 -8329.017 0.058 0.036
Chain 1: 1400 -8353.929 0.049 0.030
Chain 1: 1500 -8226.600 0.024 0.027
Chain 1: 1600 -8332.178 0.022 0.026
Chain 1: 1700 -8413.917 0.020 0.018
Chain 1: 1800 -7995.263 0.023 0.026
Chain 1: 1900 -8093.843 0.024 0.026
Chain 1: 2000 -8067.594 0.021 0.015
Chain 1: 2100 -8191.686 0.019 0.015
Chain 1: 2200 -8005.039 0.017 0.015
Chain 1: 2300 -8088.337 0.016 0.013
Chain 1: 2400 -8157.808 0.016 0.013
Chain 1: 2500 -8103.718 0.015 0.012
Chain 1: 2600 -8104.060 0.014 0.010
Chain 1: 2700 -8021.247 0.014 0.010
Chain 1: 2800 -7982.782 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003466 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.66 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8412200.597 1.000 1.000
Chain 1: 200 -1588173.848 2.648 4.297
Chain 1: 300 -890874.036 2.026 1.000
Chain 1: 400 -457098.936 1.757 1.000
Chain 1: 500 -357217.203 1.462 0.949
Chain 1: 600 -232028.211 1.308 0.949
Chain 1: 700 -118627.456 1.258 0.949
Chain 1: 800 -85935.213 1.148 0.949
Chain 1: 900 -66354.707 1.053 0.783
Chain 1: 1000 -51229.410 0.977 0.783
Chain 1: 1100 -38768.157 0.910 0.540 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37954.354 0.482 0.380
Chain 1: 1300 -25971.169 0.450 0.380
Chain 1: 1400 -25695.708 0.356 0.321
Chain 1: 1500 -22299.072 0.343 0.321
Chain 1: 1600 -21520.602 0.293 0.295
Chain 1: 1700 -20401.538 0.203 0.295
Chain 1: 1800 -20347.526 0.165 0.152
Chain 1: 1900 -20673.788 0.137 0.055
Chain 1: 2000 -19188.604 0.115 0.055
Chain 1: 2100 -19426.695 0.085 0.036
Chain 1: 2200 -19652.671 0.084 0.036
Chain 1: 2300 -19270.321 0.039 0.020
Chain 1: 2400 -19042.506 0.039 0.020
Chain 1: 2500 -18844.243 0.025 0.016
Chain 1: 2600 -18474.606 0.024 0.016
Chain 1: 2700 -18431.656 0.018 0.012
Chain 1: 2800 -18148.371 0.020 0.016
Chain 1: 2900 -18429.598 0.020 0.015
Chain 1: 3000 -18415.788 0.012 0.012
Chain 1: 3100 -18500.791 0.011 0.012
Chain 1: 3200 -18191.477 0.012 0.015
Chain 1: 3300 -18396.229 0.011 0.012
Chain 1: 3400 -17871.051 0.013 0.015
Chain 1: 3500 -18482.940 0.015 0.016
Chain 1: 3600 -17789.639 0.017 0.016
Chain 1: 3700 -18176.390 0.019 0.017
Chain 1: 3800 -17136.014 0.023 0.021
Chain 1: 3900 -17132.149 0.022 0.021
Chain 1: 4000 -17249.489 0.022 0.021
Chain 1: 4100 -17163.218 0.022 0.021
Chain 1: 4200 -16979.485 0.022 0.021
Chain 1: 4300 -17117.897 0.021 0.021
Chain 1: 4400 -17074.709 0.019 0.011
Chain 1: 4500 -16977.235 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001274 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.74 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48899.295 1.000 1.000
Chain 1: 200 -20048.235 1.220 1.439
Chain 1: 300 -19960.186 0.814 1.000
Chain 1: 400 -16754.825 0.659 1.000
Chain 1: 500 -11367.453 0.622 0.474
Chain 1: 600 -12284.119 0.531 0.474
Chain 1: 700 -14226.696 0.474 0.191
Chain 1: 800 -12946.783 0.427 0.191
Chain 1: 900 -13903.351 0.388 0.137
Chain 1: 1000 -12481.240 0.360 0.137
Chain 1: 1100 -10227.310 0.282 0.137
Chain 1: 1200 -11996.116 0.153 0.137
Chain 1: 1300 -10747.665 0.164 0.137
Chain 1: 1400 -12151.711 0.157 0.116
Chain 1: 1500 -10881.053 0.121 0.116
Chain 1: 1600 -21034.170 0.162 0.117
Chain 1: 1700 -17484.056 0.168 0.117
Chain 1: 1800 -11884.716 0.206 0.147
Chain 1: 1900 -9950.898 0.218 0.194
Chain 1: 2000 -9418.561 0.212 0.194
Chain 1: 2100 -9144.367 0.193 0.147
Chain 1: 2200 -9322.750 0.181 0.117
Chain 1: 2300 -9087.541 0.172 0.117
Chain 1: 2400 -9772.723 0.167 0.117
Chain 1: 2500 -10855.277 0.165 0.100
Chain 1: 2600 -15336.861 0.146 0.100
Chain 1: 2700 -10494.155 0.172 0.100
Chain 1: 2800 -9212.112 0.139 0.100
Chain 1: 2900 -15580.546 0.160 0.100
Chain 1: 3000 -8640.241 0.235 0.139
Chain 1: 3100 -9181.309 0.238 0.139
Chain 1: 3200 -8806.852 0.240 0.139
Chain 1: 3300 -9988.945 0.249 0.139
Chain 1: 3400 -8775.228 0.256 0.139
Chain 1: 3500 -8779.270 0.246 0.139
Chain 1: 3600 -15736.299 0.261 0.139
Chain 1: 3700 -9777.923 0.276 0.139
Chain 1: 3800 -8800.090 0.273 0.138
Chain 1: 3900 -9635.053 0.241 0.118
Chain 1: 4000 -9142.341 0.166 0.111
Chain 1: 4100 -8628.303 0.166 0.111
Chain 1: 4200 -10241.310 0.178 0.118
Chain 1: 4300 -8577.709 0.185 0.138
Chain 1: 4400 -9069.703 0.177 0.111
Chain 1: 4500 -8862.087 0.179 0.111
Chain 1: 4600 -8504.843 0.139 0.087
Chain 1: 4700 -14687.576 0.120 0.087
Chain 1: 4800 -8438.278 0.183 0.087
Chain 1: 4900 -13449.955 0.212 0.158
Chain 1: 5000 -8304.255 0.268 0.194
Chain 1: 5100 -8932.013 0.270 0.194
Chain 1: 5200 -9257.929 0.257 0.194
Chain 1: 5300 -9121.687 0.239 0.070
Chain 1: 5400 -8255.642 0.244 0.105
Chain 1: 5500 -11534.855 0.271 0.284
Chain 1: 5600 -9412.029 0.289 0.284
Chain 1: 5700 -8909.197 0.252 0.226
Chain 1: 5800 -8400.207 0.184 0.105
Chain 1: 5900 -10709.905 0.169 0.105
Chain 1: 6000 -8740.106 0.129 0.105
Chain 1: 6100 -8174.091 0.129 0.105
Chain 1: 6200 -10759.643 0.150 0.216
Chain 1: 6300 -10805.997 0.149 0.216
Chain 1: 6400 -9133.108 0.156 0.216
Chain 1: 6500 -8194.598 0.140 0.183
Chain 1: 6600 -8742.037 0.123 0.115
Chain 1: 6700 -8224.833 0.124 0.115
Chain 1: 6800 -15115.971 0.163 0.183
Chain 1: 6900 -9111.475 0.208 0.183
Chain 1: 7000 -8430.288 0.193 0.115
Chain 1: 7100 -8121.015 0.190 0.115
Chain 1: 7200 -8996.616 0.176 0.097
Chain 1: 7300 -10731.821 0.192 0.115
Chain 1: 7400 -9554.400 0.186 0.115
Chain 1: 7500 -8453.464 0.187 0.123
Chain 1: 7600 -9462.216 0.192 0.123
Chain 1: 7700 -8256.711 0.200 0.130
Chain 1: 7800 -8528.589 0.157 0.123
Chain 1: 7900 -9918.761 0.106 0.123
Chain 1: 8000 -8455.555 0.115 0.130
Chain 1: 8100 -8478.255 0.111 0.130
Chain 1: 8200 -11283.337 0.126 0.140
Chain 1: 8300 -8059.217 0.150 0.140
Chain 1: 8400 -9605.906 0.154 0.146
Chain 1: 8500 -8319.180 0.156 0.155
Chain 1: 8600 -8788.087 0.151 0.155
Chain 1: 8700 -8459.343 0.140 0.155
Chain 1: 8800 -7950.917 0.144 0.155
Chain 1: 8900 -10502.094 0.154 0.161
Chain 1: 9000 -8537.789 0.160 0.161
Chain 1: 9100 -8466.512 0.160 0.161
Chain 1: 9200 -9149.872 0.143 0.155
Chain 1: 9300 -10745.411 0.118 0.148
Chain 1: 9400 -9668.896 0.113 0.111
Chain 1: 9500 -10430.571 0.105 0.075
Chain 1: 9600 -8319.511 0.125 0.111
Chain 1: 9700 -10297.027 0.140 0.148
Chain 1: 9800 -8096.907 0.161 0.192
Chain 1: 9900 -8214.285 0.138 0.148
Chain 1: 10000 -8324.895 0.116 0.111
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001402 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.02 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61551.496 1.000 1.000
Chain 1: 200 -17672.707 1.741 2.483
Chain 1: 300 -8789.502 1.498 1.011
Chain 1: 400 -8290.708 1.138 1.011
Chain 1: 500 -8130.545 0.915 1.000
Chain 1: 600 -8690.932 0.773 1.000
Chain 1: 700 -7775.191 0.679 0.118
Chain 1: 800 -8500.836 0.605 0.118
Chain 1: 900 -7803.215 0.548 0.089
Chain 1: 1000 -7747.708 0.494 0.089
Chain 1: 1100 -7716.156 0.394 0.085
Chain 1: 1200 -7561.172 0.148 0.064
Chain 1: 1300 -7727.459 0.049 0.060
Chain 1: 1400 -7810.212 0.044 0.022
Chain 1: 1500 -7596.390 0.045 0.028
Chain 1: 1600 -7655.797 0.039 0.022
Chain 1: 1700 -7509.387 0.029 0.020
Chain 1: 1800 -7575.211 0.022 0.019
Chain 1: 1900 -7546.148 0.013 0.011
Chain 1: 2000 -7631.801 0.014 0.011
Chain 1: 2100 -7592.043 0.014 0.011
Chain 1: 2200 -7673.962 0.013 0.011
Chain 1: 2300 -7591.254 0.012 0.011
Chain 1: 2400 -7629.308 0.011 0.011
Chain 1: 2500 -7556.365 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003579 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.79 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86875.740 1.000 1.000
Chain 1: 200 -13321.906 3.261 5.521
Chain 1: 300 -9681.514 2.299 1.000
Chain 1: 400 -10571.848 1.745 1.000
Chain 1: 500 -8650.075 1.441 0.376
Chain 1: 600 -8104.335 1.212 0.376
Chain 1: 700 -8137.948 1.039 0.222
Chain 1: 800 -8572.677 0.916 0.222
Chain 1: 900 -8526.219 0.815 0.084
Chain 1: 1000 -8120.495 0.738 0.084
Chain 1: 1100 -8426.086 0.642 0.067 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8023.298 0.095 0.051
Chain 1: 1300 -8369.048 0.061 0.050
Chain 1: 1400 -8170.205 0.055 0.050
Chain 1: 1500 -8208.787 0.033 0.041
Chain 1: 1600 -8196.294 0.027 0.036
Chain 1: 1700 -8098.978 0.028 0.036
Chain 1: 1800 -7997.737 0.024 0.024
Chain 1: 1900 -8120.551 0.025 0.024
Chain 1: 2000 -8084.313 0.020 0.015
Chain 1: 2100 -8215.421 0.018 0.015
Chain 1: 2200 -8031.270 0.016 0.015
Chain 1: 2300 -8109.652 0.012 0.013
Chain 1: 2400 -8179.184 0.011 0.012
Chain 1: 2500 -8124.388 0.011 0.012
Chain 1: 2600 -8123.683 0.011 0.012
Chain 1: 2700 -8041.130 0.011 0.010
Chain 1: 2800 -8003.751 0.010 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004384 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 43.84 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8429270.425 1.000 1.000
Chain 1: 200 -1587991.734 2.654 4.308
Chain 1: 300 -891208.547 2.030 1.000
Chain 1: 400 -457364.129 1.760 1.000
Chain 1: 500 -357355.531 1.464 0.949
Chain 1: 600 -232358.282 1.309 0.949
Chain 1: 700 -118821.140 1.259 0.949
Chain 1: 800 -86077.399 1.149 0.949
Chain 1: 900 -66466.910 1.054 0.782
Chain 1: 1000 -51300.913 0.978 0.782
Chain 1: 1100 -38812.538 0.910 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37994.614 0.482 0.380
Chain 1: 1300 -25987.622 0.450 0.380
Chain 1: 1400 -25710.301 0.356 0.322
Chain 1: 1500 -22306.811 0.343 0.322
Chain 1: 1600 -21526.175 0.293 0.296
Chain 1: 1700 -20404.344 0.203 0.295
Chain 1: 1800 -20349.563 0.165 0.153
Chain 1: 1900 -20675.692 0.137 0.055
Chain 1: 2000 -19189.297 0.116 0.055
Chain 1: 2100 -19427.555 0.085 0.036
Chain 1: 2200 -19653.548 0.084 0.036
Chain 1: 2300 -19271.195 0.039 0.020
Chain 1: 2400 -19043.344 0.040 0.020
Chain 1: 2500 -18845.159 0.025 0.016
Chain 1: 2600 -18475.546 0.024 0.016
Chain 1: 2700 -18432.643 0.018 0.012
Chain 1: 2800 -18149.379 0.020 0.016
Chain 1: 2900 -18430.625 0.020 0.015
Chain 1: 3000 -18416.852 0.012 0.012
Chain 1: 3100 -18501.786 0.011 0.012
Chain 1: 3200 -18192.546 0.012 0.015
Chain 1: 3300 -18397.257 0.011 0.012
Chain 1: 3400 -17872.195 0.013 0.015
Chain 1: 3500 -18483.910 0.015 0.016
Chain 1: 3600 -17790.861 0.017 0.016
Chain 1: 3700 -18177.399 0.019 0.017
Chain 1: 3800 -17137.405 0.023 0.021
Chain 1: 3900 -17133.537 0.022 0.021
Chain 1: 4000 -17250.880 0.022 0.021
Chain 1: 4100 -17164.583 0.022 0.021
Chain 1: 4200 -16980.951 0.022 0.021
Chain 1: 4300 -17119.291 0.021 0.021
Chain 1: 4400 -17076.180 0.019 0.011
Chain 1: 4500 -16978.703 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001333 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.33 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12992.330 1.000 1.000
Chain 1: 200 -9891.546 0.657 1.000
Chain 1: 300 -8484.429 0.493 0.313
Chain 1: 400 -8636.854 0.374 0.313
Chain 1: 500 -8353.908 0.306 0.166
Chain 1: 600 -8398.394 0.256 0.166
Chain 1: 700 -8310.798 0.221 0.034
Chain 1: 800 -8323.596 0.194 0.034
Chain 1: 900 -8345.223 0.172 0.018
Chain 1: 1000 -8413.863 0.156 0.018
Chain 1: 1100 -8448.372 0.056 0.011
Chain 1: 1200 -8342.794 0.026 0.011
Chain 1: 1300 -8279.035 0.010 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001393 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.93 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -46528.491 1.000 1.000
Chain 1: 200 -16139.861 1.441 1.883
Chain 1: 300 -9097.887 1.219 1.000
Chain 1: 400 -8724.061 0.925 1.000
Chain 1: 500 -9129.148 0.749 0.774
Chain 1: 600 -8402.628 0.638 0.774
Chain 1: 700 -8000.458 0.554 0.086
Chain 1: 800 -8129.665 0.487 0.086
Chain 1: 900 -8016.800 0.435 0.050
Chain 1: 1000 -8197.439 0.393 0.050
Chain 1: 1100 -8033.278 0.295 0.044
Chain 1: 1200 -7732.371 0.111 0.043
Chain 1: 1300 -8030.339 0.037 0.039
Chain 1: 1400 -8143.116 0.034 0.037
Chain 1: 1500 -7799.581 0.034 0.037
Chain 1: 1600 -7949.548 0.028 0.022
Chain 1: 1700 -7815.372 0.024 0.020
Chain 1: 1800 -7850.508 0.023 0.020
Chain 1: 1900 -7819.627 0.022 0.020
Chain 1: 2000 -7907.582 0.021 0.019
Chain 1: 2100 -7813.913 0.020 0.017
Chain 1: 2200 -7976.742 0.018 0.017
Chain 1: 2300 -7818.597 0.017 0.017
Chain 1: 2400 -7800.783 0.015 0.017
Chain 1: 2500 -7834.260 0.011 0.012
Chain 1: 2600 -7745.830 0.011 0.011
Chain 1: 2700 -7663.367 0.010 0.011
Chain 1: 2800 -7844.040 0.012 0.011
Chain 1: 2900 -7598.709 0.015 0.012
Chain 1: 3000 -7755.184 0.016 0.020
Chain 1: 3100 -7738.560 0.015 0.020
Chain 1: 3200 -7955.481 0.015 0.020
Chain 1: 3300 -7665.093 0.017 0.020
Chain 1: 3400 -7907.640 0.020 0.023
Chain 1: 3500 -7653.983 0.023 0.027
Chain 1: 3600 -7718.896 0.023 0.027
Chain 1: 3700 -7670.196 0.022 0.027
Chain 1: 3800 -7669.856 0.020 0.027
Chain 1: 3900 -7629.507 0.017 0.020
Chain 1: 4000 -7621.360 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003175 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.75 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87501.618 1.000 1.000
Chain 1: 200 -14130.957 3.096 5.192
Chain 1: 300 -10389.872 2.184 1.000
Chain 1: 400 -11881.638 1.669 1.000
Chain 1: 500 -9269.333 1.392 0.360
Chain 1: 600 -9717.109 1.168 0.360
Chain 1: 700 -9129.275 1.010 0.282
Chain 1: 800 -8673.682 0.890 0.282
Chain 1: 900 -8730.813 0.792 0.126
Chain 1: 1000 -9129.340 0.717 0.126
Chain 1: 1100 -9202.298 0.618 0.064 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8759.043 0.104 0.053
Chain 1: 1300 -9033.397 0.071 0.051
Chain 1: 1400 -9009.504 0.059 0.046
Chain 1: 1500 -8914.160 0.032 0.044
Chain 1: 1600 -9016.151 0.028 0.030
Chain 1: 1700 -9078.789 0.022 0.011
Chain 1: 1800 -8641.259 0.022 0.011
Chain 1: 1900 -8746.176 0.023 0.012
Chain 1: 2000 -8723.731 0.019 0.011
Chain 1: 2100 -8700.576 0.018 0.011
Chain 1: 2200 -8667.749 0.013 0.011
Chain 1: 2300 -8801.727 0.012 0.011
Chain 1: 2400 -8644.101 0.013 0.011
Chain 1: 2500 -8715.993 0.013 0.011
Chain 1: 2600 -8629.345 0.013 0.010
Chain 1: 2700 -8665.903 0.013 0.010
Chain 1: 2800 -8624.461 0.008 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003275 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.75 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8393870.922 1.000 1.000
Chain 1: 200 -1583809.216 2.650 4.300
Chain 1: 300 -891894.138 2.025 1.000
Chain 1: 400 -458488.884 1.755 1.000
Chain 1: 500 -358702.375 1.460 0.945
Chain 1: 600 -233777.149 1.306 0.945
Chain 1: 700 -119960.820 1.255 0.945
Chain 1: 800 -87115.904 1.145 0.945
Chain 1: 900 -67466.144 1.050 0.776
Chain 1: 1000 -52262.315 0.974 0.776
Chain 1: 1100 -39726.465 0.906 0.534 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38910.073 0.478 0.377
Chain 1: 1300 -26848.758 0.445 0.377
Chain 1: 1400 -26568.616 0.352 0.316
Chain 1: 1500 -23149.522 0.339 0.316
Chain 1: 1600 -22364.519 0.289 0.291
Chain 1: 1700 -21235.902 0.199 0.291
Chain 1: 1800 -21179.751 0.162 0.148
Chain 1: 1900 -21506.348 0.134 0.053
Chain 1: 2000 -20015.192 0.112 0.053
Chain 1: 2100 -20253.915 0.082 0.035
Chain 1: 2200 -20480.669 0.081 0.035
Chain 1: 2300 -20097.482 0.038 0.019
Chain 1: 2400 -19869.407 0.038 0.019
Chain 1: 2500 -19671.295 0.024 0.015
Chain 1: 2600 -19301.120 0.023 0.015
Chain 1: 2700 -19258.051 0.018 0.012
Chain 1: 2800 -18974.608 0.019 0.015
Chain 1: 2900 -19256.138 0.019 0.015
Chain 1: 3000 -19242.387 0.012 0.012
Chain 1: 3100 -19327.358 0.011 0.011
Chain 1: 3200 -19017.788 0.011 0.015
Chain 1: 3300 -19222.747 0.010 0.011
Chain 1: 3400 -18697.099 0.012 0.015
Chain 1: 3500 -19309.750 0.014 0.015
Chain 1: 3600 -18615.557 0.016 0.015
Chain 1: 3700 -19002.973 0.018 0.016
Chain 1: 3800 -17961.168 0.022 0.020
Chain 1: 3900 -17957.275 0.021 0.020
Chain 1: 4000 -18074.613 0.021 0.020
Chain 1: 4100 -17988.209 0.021 0.020
Chain 1: 4200 -17804.199 0.021 0.020
Chain 1: 4300 -17942.810 0.021 0.020
Chain 1: 4400 -17899.391 0.018 0.010
Chain 1: 4500 -17801.863 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001284 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.84 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12841.378 1.000 1.000
Chain 1: 200 -9816.549 0.654 1.000
Chain 1: 300 -8484.444 0.488 0.308
Chain 1: 400 -8619.662 0.370 0.308
Chain 1: 500 -8619.588 0.296 0.157
Chain 1: 600 -8398.805 0.251 0.157
Chain 1: 700 -8305.543 0.217 0.026
Chain 1: 800 -8328.911 0.190 0.026
Chain 1: 900 -8444.828 0.171 0.016
Chain 1: 1000 -8346.763 0.155 0.016
Chain 1: 1100 -8378.864 0.055 0.014
Chain 1: 1200 -8348.133 0.025 0.012
Chain 1: 1300 -8264.107 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001432 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.32 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58710.469 1.000 1.000
Chain 1: 200 -18177.868 1.615 2.230
Chain 1: 300 -8916.381 1.423 1.039
Chain 1: 400 -8134.074 1.091 1.039
Chain 1: 500 -8767.588 0.887 1.000
Chain 1: 600 -8550.414 0.744 1.000
Chain 1: 700 -7795.477 0.651 0.097
Chain 1: 800 -8142.870 0.575 0.097
Chain 1: 900 -7902.852 0.515 0.096
Chain 1: 1000 -7924.315 0.463 0.096
Chain 1: 1100 -7871.826 0.364 0.072
Chain 1: 1200 -7825.696 0.142 0.043
Chain 1: 1300 -7890.714 0.039 0.030
Chain 1: 1400 -7919.589 0.029 0.025
Chain 1: 1500 -7612.004 0.026 0.025
Chain 1: 1600 -7816.379 0.026 0.026
Chain 1: 1700 -7629.814 0.019 0.024
Chain 1: 1800 -7659.643 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003493 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.93 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86716.324 1.000 1.000
Chain 1: 200 -13975.061 3.103 5.205
Chain 1: 300 -10309.691 2.187 1.000
Chain 1: 400 -11437.052 1.665 1.000
Chain 1: 500 -9292.868 1.378 0.356
Chain 1: 600 -8889.275 1.156 0.356
Chain 1: 700 -8865.493 0.991 0.231
Chain 1: 800 -9505.883 0.876 0.231
Chain 1: 900 -9097.708 0.783 0.099
Chain 1: 1000 -9034.645 0.706 0.099
Chain 1: 1100 -9129.641 0.607 0.067 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8646.168 0.092 0.056
Chain 1: 1300 -8997.307 0.060 0.045
Chain 1: 1400 -8998.329 0.050 0.045
Chain 1: 1500 -8866.556 0.029 0.039
Chain 1: 1600 -8974.877 0.025 0.015
Chain 1: 1700 -9051.997 0.026 0.015
Chain 1: 1800 -8627.955 0.024 0.015
Chain 1: 1900 -8728.736 0.021 0.012
Chain 1: 2000 -8703.363 0.020 0.012
Chain 1: 2100 -8829.008 0.021 0.014
Chain 1: 2200 -8631.403 0.018 0.014
Chain 1: 2300 -8723.658 0.015 0.012
Chain 1: 2400 -8792.364 0.015 0.012
Chain 1: 2500 -8738.668 0.015 0.012
Chain 1: 2600 -8740.148 0.013 0.011
Chain 1: 2700 -8656.783 0.014 0.011
Chain 1: 2800 -8616.514 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003464 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.64 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8401727.369 1.000 1.000
Chain 1: 200 -1585430.286 2.650 4.299
Chain 1: 300 -892309.303 2.025 1.000
Chain 1: 400 -459064.489 1.755 1.000
Chain 1: 500 -359098.708 1.460 0.944
Chain 1: 600 -233858.935 1.306 0.944
Chain 1: 700 -119871.613 1.255 0.944
Chain 1: 800 -87047.857 1.145 0.944
Chain 1: 900 -67355.352 1.050 0.777
Chain 1: 1000 -52127.067 0.975 0.777
Chain 1: 1100 -39583.913 0.906 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38759.606 0.479 0.377
Chain 1: 1300 -26692.861 0.446 0.377
Chain 1: 1400 -26411.276 0.353 0.317
Chain 1: 1500 -22992.347 0.340 0.317
Chain 1: 1600 -22207.598 0.290 0.292
Chain 1: 1700 -21078.137 0.200 0.292
Chain 1: 1800 -21021.749 0.163 0.149
Chain 1: 1900 -21347.992 0.135 0.054
Chain 1: 2000 -19857.333 0.113 0.054
Chain 1: 2100 -20095.863 0.083 0.035
Chain 1: 2200 -20322.689 0.082 0.035
Chain 1: 2300 -19939.525 0.038 0.019
Chain 1: 2400 -19711.512 0.038 0.019
Chain 1: 2500 -19513.716 0.025 0.015
Chain 1: 2600 -19143.674 0.023 0.015
Chain 1: 2700 -19100.517 0.018 0.012
Chain 1: 2800 -18817.387 0.019 0.015
Chain 1: 2900 -19098.738 0.019 0.015
Chain 1: 3000 -19084.885 0.012 0.012
Chain 1: 3100 -19169.928 0.011 0.012
Chain 1: 3200 -18860.481 0.011 0.015
Chain 1: 3300 -19065.297 0.011 0.012
Chain 1: 3400 -18540.090 0.012 0.015
Chain 1: 3500 -19152.211 0.014 0.015
Chain 1: 3600 -18458.544 0.016 0.015
Chain 1: 3700 -18845.638 0.018 0.016
Chain 1: 3800 -17804.881 0.022 0.021
Chain 1: 3900 -17801.017 0.021 0.021
Chain 1: 4000 -17918.312 0.022 0.021
Chain 1: 4100 -17832.069 0.022 0.021
Chain 1: 4200 -17648.183 0.021 0.021
Chain 1: 4300 -17786.650 0.021 0.021
Chain 1: 4400 -17743.391 0.018 0.010
Chain 1: 4500 -17645.909 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001284 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.84 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12101.170 1.000 1.000
Chain 1: 200 -9101.503 0.665 1.000
Chain 1: 300 -7794.915 0.499 0.330
Chain 1: 400 -7931.835 0.379 0.330
Chain 1: 500 -7822.174 0.306 0.168
Chain 1: 600 -7746.294 0.256 0.168
Chain 1: 700 -7658.458 0.221 0.017
Chain 1: 800 -7702.138 0.194 0.017
Chain 1: 900 -7821.828 0.175 0.015
Chain 1: 1000 -7722.965 0.158 0.015
Chain 1: 1100 -7709.306 0.059 0.014
Chain 1: 1200 -7669.161 0.026 0.013
Chain 1: 1300 -7626.975 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001398 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.98 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56361.486 1.000 1.000
Chain 1: 200 -17119.646 1.646 2.292
Chain 1: 300 -8605.142 1.427 1.000
Chain 1: 400 -9115.821 1.084 1.000
Chain 1: 500 -8629.401 0.879 0.989
Chain 1: 600 -8490.538 0.735 0.989
Chain 1: 700 -7809.027 0.643 0.087
Chain 1: 800 -8091.209 0.567 0.087
Chain 1: 900 -7688.196 0.509 0.056
Chain 1: 1000 -7868.278 0.461 0.056
Chain 1: 1100 -7673.503 0.363 0.056
Chain 1: 1200 -7587.791 0.135 0.052
Chain 1: 1300 -7721.117 0.038 0.035
Chain 1: 1400 -7867.027 0.034 0.025
Chain 1: 1500 -7631.215 0.032 0.025
Chain 1: 1600 -7575.608 0.031 0.025
Chain 1: 1700 -7516.199 0.023 0.023
Chain 1: 1800 -7587.516 0.020 0.019
Chain 1: 1900 -7631.852 0.016 0.017
Chain 1: 2000 -7633.815 0.013 0.011
Chain 1: 2100 -7657.425 0.011 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003141 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.41 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86672.284 1.000 1.000
Chain 1: 200 -13212.534 3.280 5.560
Chain 1: 300 -9600.956 2.312 1.000
Chain 1: 400 -10323.161 1.751 1.000
Chain 1: 500 -8541.510 1.443 0.376
Chain 1: 600 -8188.528 1.210 0.376
Chain 1: 700 -8230.429 1.038 0.209
Chain 1: 800 -8831.333 0.916 0.209
Chain 1: 900 -8340.971 0.821 0.070
Chain 1: 1000 -8186.266 0.741 0.070
Chain 1: 1100 -8447.559 0.644 0.068 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8081.872 0.092 0.059
Chain 1: 1300 -8293.883 0.057 0.045
Chain 1: 1400 -8294.893 0.050 0.043
Chain 1: 1500 -8183.202 0.031 0.031
Chain 1: 1600 -8287.963 0.028 0.026
Chain 1: 1700 -8376.351 0.028 0.026
Chain 1: 1800 -7970.396 0.027 0.026
Chain 1: 1900 -8067.971 0.022 0.019
Chain 1: 2000 -8039.725 0.021 0.014
Chain 1: 2100 -8159.914 0.019 0.014
Chain 1: 2200 -7953.686 0.017 0.014
Chain 1: 2300 -8104.075 0.016 0.014
Chain 1: 2400 -8110.966 0.016 0.014
Chain 1: 2500 -8081.477 0.015 0.013
Chain 1: 2600 -8080.215 0.014 0.012
Chain 1: 2700 -7992.621 0.014 0.012
Chain 1: 2800 -7958.583 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003509 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.09 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8397444.290 1.000 1.000
Chain 1: 200 -1583120.107 2.652 4.304
Chain 1: 300 -890781.049 2.027 1.000
Chain 1: 400 -457205.946 1.757 1.000
Chain 1: 500 -357818.003 1.462 0.948
Chain 1: 600 -232756.097 1.307 0.948
Chain 1: 700 -119018.719 1.257 0.948
Chain 1: 800 -86189.683 1.148 0.948
Chain 1: 900 -66522.709 1.053 0.777
Chain 1: 1000 -51305.257 0.977 0.777
Chain 1: 1100 -38767.742 0.910 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37941.852 0.481 0.381
Chain 1: 1300 -25890.554 0.450 0.381
Chain 1: 1400 -25607.079 0.357 0.323
Chain 1: 1500 -22192.049 0.344 0.323
Chain 1: 1600 -21407.475 0.294 0.297
Chain 1: 1700 -20280.935 0.204 0.296
Chain 1: 1800 -20224.824 0.166 0.154
Chain 1: 1900 -20550.786 0.138 0.056
Chain 1: 2000 -19062.005 0.116 0.056
Chain 1: 2100 -19300.462 0.085 0.037
Chain 1: 2200 -19526.695 0.084 0.037
Chain 1: 2300 -19144.129 0.040 0.020
Chain 1: 2400 -18916.314 0.040 0.020
Chain 1: 2500 -18718.224 0.026 0.016
Chain 1: 2600 -18348.714 0.024 0.016
Chain 1: 2700 -18305.796 0.019 0.012
Chain 1: 2800 -18022.697 0.020 0.016
Chain 1: 2900 -18303.897 0.020 0.015
Chain 1: 3000 -18290.059 0.012 0.012
Chain 1: 3100 -18375.019 0.011 0.012
Chain 1: 3200 -18065.857 0.012 0.015
Chain 1: 3300 -18270.484 0.011 0.012
Chain 1: 3400 -17745.644 0.013 0.015
Chain 1: 3500 -18357.121 0.015 0.016
Chain 1: 3600 -17664.366 0.017 0.016
Chain 1: 3700 -18050.772 0.019 0.017
Chain 1: 3800 -17011.275 0.023 0.021
Chain 1: 3900 -17007.446 0.022 0.021
Chain 1: 4000 -17124.746 0.022 0.021
Chain 1: 4100 -17038.510 0.023 0.021
Chain 1: 4200 -16854.975 0.022 0.021
Chain 1: 4300 -16993.243 0.022 0.021
Chain 1: 4400 -16950.243 0.019 0.011
Chain 1: 4500 -16852.777 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001246 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.46 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13369.228 1.000 1.000
Chain 1: 200 -10008.249 0.668 1.000
Chain 1: 300 -8591.896 0.500 0.336
Chain 1: 400 -8334.243 0.383 0.336
Chain 1: 500 -8381.462 0.307 0.165
Chain 1: 600 -8224.054 0.259 0.165
Chain 1: 700 -8130.249 0.224 0.031
Chain 1: 800 -8378.502 0.200 0.031
Chain 1: 900 -8118.961 0.181 0.031
Chain 1: 1000 -8145.249 0.163 0.031
Chain 1: 1100 -8230.855 0.064 0.030
Chain 1: 1200 -8136.592 0.032 0.019
Chain 1: 1300 -8130.330 0.015 0.012
Chain 1: 1400 -8122.469 0.012 0.012
Chain 1: 1500 -8209.169 0.013 0.012
Chain 1: 1600 -8159.716 0.012 0.011
Chain 1: 1700 -8102.810 0.011 0.010
Chain 1: 1800 -8077.982 0.009 0.007 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001402 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.02 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -47613.483 1.000 1.000
Chain 1: 200 -16075.334 1.481 1.962
Chain 1: 300 -8718.945 1.269 1.000
Chain 1: 400 -8559.974 0.956 1.000
Chain 1: 500 -8538.562 0.765 0.844
Chain 1: 600 -8409.218 0.640 0.844
Chain 1: 700 -8049.890 0.555 0.045
Chain 1: 800 -7987.689 0.487 0.045
Chain 1: 900 -7837.432 0.435 0.019
Chain 1: 1000 -7973.618 0.393 0.019
Chain 1: 1100 -7617.537 0.298 0.019
Chain 1: 1200 -7903.426 0.105 0.019
Chain 1: 1300 -7855.660 0.021 0.019
Chain 1: 1400 -7734.807 0.021 0.017
Chain 1: 1500 -7641.451 0.022 0.017
Chain 1: 1600 -7794.209 0.023 0.019
Chain 1: 1700 -7572.847 0.021 0.019
Chain 1: 1800 -7624.800 0.021 0.019
Chain 1: 1900 -7627.713 0.019 0.017
Chain 1: 2000 -7704.150 0.018 0.016
Chain 1: 2100 -7624.610 0.015 0.012
Chain 1: 2200 -7753.451 0.013 0.012
Chain 1: 2300 -7606.091 0.014 0.016
Chain 1: 2400 -7676.133 0.013 0.012
Chain 1: 2500 -7593.970 0.013 0.011
Chain 1: 2600 -7554.409 0.012 0.010
Chain 1: 2700 -7497.529 0.010 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003549 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.49 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86611.166 1.000 1.000
Chain 1: 200 -13788.578 3.141 5.281
Chain 1: 300 -10125.242 2.214 1.000
Chain 1: 400 -10974.671 1.680 1.000
Chain 1: 500 -9126.072 1.385 0.362
Chain 1: 600 -8884.513 1.158 0.362
Chain 1: 700 -8575.183 0.998 0.203
Chain 1: 800 -9125.054 0.881 0.203
Chain 1: 900 -8880.443 0.786 0.077
Chain 1: 1000 -8666.933 0.710 0.077
Chain 1: 1100 -8913.992 0.613 0.060 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8467.141 0.090 0.053
Chain 1: 1300 -8677.869 0.056 0.036
Chain 1: 1400 -8814.029 0.050 0.028
Chain 1: 1500 -8686.182 0.031 0.028
Chain 1: 1600 -8804.883 0.030 0.028
Chain 1: 1700 -8875.379 0.027 0.025
Chain 1: 1800 -8455.461 0.026 0.025
Chain 1: 1900 -8554.203 0.024 0.024
Chain 1: 2000 -8528.541 0.022 0.015
Chain 1: 2100 -8653.525 0.021 0.015
Chain 1: 2200 -8459.811 0.018 0.015
Chain 1: 2300 -8549.011 0.016 0.014
Chain 1: 2400 -8618.115 0.016 0.013
Chain 1: 2500 -8564.302 0.015 0.012
Chain 1: 2600 -8565.220 0.013 0.010
Chain 1: 2700 -8482.126 0.014 0.010
Chain 1: 2800 -8442.676 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003378 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.78 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8397353.624 1.000 1.000
Chain 1: 200 -1585727.396 2.648 4.296
Chain 1: 300 -891519.831 2.025 1.000
Chain 1: 400 -457966.975 1.755 1.000
Chain 1: 500 -358078.143 1.460 0.947
Chain 1: 600 -233110.284 1.306 0.947
Chain 1: 700 -119479.635 1.255 0.947
Chain 1: 800 -86648.531 1.146 0.947
Chain 1: 900 -67022.485 1.051 0.779
Chain 1: 1000 -51834.659 0.975 0.779
Chain 1: 1100 -39321.740 0.907 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38504.109 0.480 0.379
Chain 1: 1300 -26477.959 0.447 0.379
Chain 1: 1400 -26198.901 0.354 0.318
Chain 1: 1500 -22788.758 0.341 0.318
Chain 1: 1600 -22005.829 0.291 0.293
Chain 1: 1700 -20882.102 0.201 0.293
Chain 1: 1800 -20826.898 0.163 0.150
Chain 1: 1900 -21153.123 0.135 0.054
Chain 1: 2000 -19665.099 0.114 0.054
Chain 1: 2100 -19903.687 0.083 0.036
Chain 1: 2200 -20129.701 0.082 0.036
Chain 1: 2300 -19747.255 0.039 0.019
Chain 1: 2400 -19519.338 0.039 0.019
Chain 1: 2500 -19321.018 0.025 0.015
Chain 1: 2600 -18951.423 0.023 0.015
Chain 1: 2700 -18908.557 0.018 0.012
Chain 1: 2800 -18625.120 0.019 0.015
Chain 1: 2900 -18906.494 0.019 0.015
Chain 1: 3000 -18892.775 0.012 0.012
Chain 1: 3100 -18977.666 0.011 0.012
Chain 1: 3200 -18668.432 0.011 0.015
Chain 1: 3300 -18873.148 0.011 0.012
Chain 1: 3400 -18347.932 0.012 0.015
Chain 1: 3500 -18959.871 0.015 0.015
Chain 1: 3600 -18266.611 0.016 0.015
Chain 1: 3700 -18653.275 0.018 0.017
Chain 1: 3800 -17612.872 0.023 0.021
Chain 1: 3900 -17608.993 0.021 0.021
Chain 1: 4000 -17726.358 0.022 0.021
Chain 1: 4100 -17639.971 0.022 0.021
Chain 1: 4200 -17456.295 0.021 0.021
Chain 1: 4300 -17594.705 0.021 0.021
Chain 1: 4400 -17551.546 0.018 0.011
Chain 1: 4500 -17454.032 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001383 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.83 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12362.475 1.000 1.000
Chain 1: 200 -9154.341 0.675 1.000
Chain 1: 300 -8096.470 0.494 0.350
Chain 1: 400 -8063.397 0.371 0.350
Chain 1: 500 -7962.939 0.300 0.131
Chain 1: 600 -7897.741 0.251 0.131
Chain 1: 700 -7808.870 0.217 0.013
Chain 1: 800 -7817.082 0.190 0.013
Chain 1: 900 -7716.593 0.170 0.013
Chain 1: 1000 -7874.352 0.155 0.013
Chain 1: 1100 -7844.481 0.056 0.013
Chain 1: 1200 -7835.139 0.021 0.011
Chain 1: 1300 -7779.444 0.008 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001429 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.29 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -60017.108 1.000 1.000
Chain 1: 200 -17827.530 1.683 2.367
Chain 1: 300 -8659.722 1.475 1.059
Chain 1: 400 -8301.581 1.117 1.059
Chain 1: 500 -8057.390 0.900 1.000
Chain 1: 600 -8418.679 0.757 1.000
Chain 1: 700 -7883.869 0.658 0.068
Chain 1: 800 -8201.390 0.581 0.068
Chain 1: 900 -7888.239 0.521 0.043
Chain 1: 1000 -7801.866 0.470 0.043
Chain 1: 1100 -7709.516 0.371 0.043
Chain 1: 1200 -7567.922 0.136 0.040
Chain 1: 1300 -7638.984 0.031 0.039
Chain 1: 1400 -7903.550 0.030 0.033
Chain 1: 1500 -7559.191 0.032 0.039
Chain 1: 1600 -7537.879 0.028 0.033
Chain 1: 1700 -7490.358 0.022 0.019
Chain 1: 1800 -7523.916 0.018 0.012
Chain 1: 1900 -7525.516 0.014 0.011
Chain 1: 2000 -7554.322 0.014 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00313 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.3 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86092.239 1.000 1.000
Chain 1: 200 -13385.622 3.216 5.432
Chain 1: 300 -9772.301 2.267 1.000
Chain 1: 400 -10493.743 1.718 1.000
Chain 1: 500 -8744.889 1.414 0.370
Chain 1: 600 -8260.281 1.188 0.370
Chain 1: 700 -8330.785 1.020 0.200
Chain 1: 800 -9020.881 0.902 0.200
Chain 1: 900 -8540.853 0.808 0.076
Chain 1: 1000 -8396.181 0.729 0.076
Chain 1: 1100 -8595.841 0.631 0.069 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8302.272 0.091 0.059
Chain 1: 1300 -8494.235 0.057 0.056
Chain 1: 1400 -8495.259 0.050 0.035
Chain 1: 1500 -8345.837 0.032 0.023
Chain 1: 1600 -8456.740 0.027 0.023
Chain 1: 1700 -8545.760 0.027 0.023
Chain 1: 1800 -8138.440 0.025 0.023
Chain 1: 1900 -8234.265 0.020 0.018
Chain 1: 2000 -8206.682 0.019 0.018
Chain 1: 2100 -8327.471 0.018 0.015
Chain 1: 2200 -8155.358 0.016 0.015
Chain 1: 2300 -8271.787 0.016 0.014
Chain 1: 2400 -8283.328 0.016 0.014
Chain 1: 2500 -8244.970 0.014 0.013
Chain 1: 2600 -8244.827 0.013 0.012
Chain 1: 2700 -8159.978 0.013 0.012
Chain 1: 2800 -8124.674 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003458 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.58 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8412238.546 1.000 1.000
Chain 1: 200 -1583500.999 2.656 4.312
Chain 1: 300 -891288.883 2.030 1.000
Chain 1: 400 -458178.714 1.759 1.000
Chain 1: 500 -358646.469 1.462 0.945
Chain 1: 600 -233351.364 1.308 0.945
Chain 1: 700 -119333.488 1.258 0.945
Chain 1: 800 -86497.790 1.148 0.945
Chain 1: 900 -66783.028 1.053 0.777
Chain 1: 1000 -51540.165 0.977 0.777
Chain 1: 1100 -38989.238 0.910 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38158.623 0.481 0.380
Chain 1: 1300 -26089.745 0.449 0.380
Chain 1: 1400 -25804.811 0.356 0.322
Chain 1: 1500 -22386.634 0.343 0.322
Chain 1: 1600 -21601.540 0.293 0.296
Chain 1: 1700 -20472.531 0.203 0.295
Chain 1: 1800 -20415.951 0.166 0.153
Chain 1: 1900 -20741.938 0.138 0.055
Chain 1: 2000 -19252.149 0.116 0.055
Chain 1: 2100 -19490.380 0.085 0.036
Chain 1: 2200 -19717.088 0.084 0.036
Chain 1: 2300 -19334.129 0.039 0.020
Chain 1: 2400 -19106.273 0.040 0.020
Chain 1: 2500 -18908.456 0.025 0.016
Chain 1: 2600 -18538.593 0.024 0.016
Chain 1: 2700 -18495.541 0.018 0.012
Chain 1: 2800 -18212.569 0.020 0.016
Chain 1: 2900 -18493.798 0.020 0.015
Chain 1: 3000 -18479.850 0.012 0.012
Chain 1: 3100 -18564.868 0.011 0.012
Chain 1: 3200 -18255.585 0.012 0.015
Chain 1: 3300 -18460.291 0.011 0.012
Chain 1: 3400 -17935.378 0.013 0.015
Chain 1: 3500 -18547.038 0.015 0.016
Chain 1: 3600 -17853.989 0.017 0.016
Chain 1: 3700 -18240.660 0.019 0.017
Chain 1: 3800 -17200.798 0.023 0.021
Chain 1: 3900 -17196.985 0.022 0.021
Chain 1: 4000 -17314.253 0.022 0.021
Chain 1: 4100 -17228.066 0.022 0.021
Chain 1: 4200 -17044.390 0.022 0.021
Chain 1: 4300 -17182.692 0.021 0.021
Chain 1: 4400 -17139.611 0.019 0.011
Chain 1: 4500 -17042.170 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001351 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.51 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12896.397 1.000 1.000
Chain 1: 200 -9784.884 0.659 1.000
Chain 1: 300 -8645.335 0.483 0.318
Chain 1: 400 -8293.676 0.373 0.318
Chain 1: 500 -8473.334 0.303 0.132
Chain 1: 600 -8184.945 0.258 0.132
Chain 1: 700 -8041.076 0.224 0.042
Chain 1: 800 -8085.972 0.197 0.042
Chain 1: 900 -8180.650 0.176 0.035
Chain 1: 1000 -8229.812 0.159 0.035
Chain 1: 1100 -8177.190 0.060 0.021
Chain 1: 1200 -8084.208 0.029 0.018
Chain 1: 1300 -8022.767 0.017 0.012
Chain 1: 1400 -8045.419 0.013 0.012
Chain 1: 1500 -8075.196 0.011 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001502 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.02 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -63973.343 1.000 1.000
Chain 1: 200 -18849.612 1.697 2.394
Chain 1: 300 -9061.763 1.491 1.080
Chain 1: 400 -10048.308 1.143 1.080
Chain 1: 500 -9091.712 0.935 1.000
Chain 1: 600 -9218.800 0.782 1.000
Chain 1: 700 -8665.166 0.679 0.105
Chain 1: 800 -8394.252 0.598 0.105
Chain 1: 900 -7911.882 0.539 0.098
Chain 1: 1000 -7834.396 0.486 0.098
Chain 1: 1100 -7677.686 0.388 0.064
Chain 1: 1200 -7557.014 0.150 0.061
Chain 1: 1300 -7905.724 0.046 0.044
Chain 1: 1400 -7595.742 0.041 0.041
Chain 1: 1500 -7509.438 0.031 0.032
Chain 1: 1600 -7730.300 0.033 0.032
Chain 1: 1700 -7765.905 0.027 0.029
Chain 1: 1800 -7739.862 0.024 0.020
Chain 1: 1900 -7567.441 0.020 0.020
Chain 1: 2000 -7548.551 0.019 0.020
Chain 1: 2100 -7521.161 0.018 0.016
Chain 1: 2200 -7765.668 0.019 0.023
Chain 1: 2300 -7562.233 0.018 0.023
Chain 1: 2400 -7668.382 0.015 0.014
Chain 1: 2500 -7568.719 0.015 0.014
Chain 1: 2600 -7493.744 0.013 0.013
Chain 1: 2700 -7531.986 0.013 0.013
Chain 1: 2800 -7599.591 0.014 0.013
Chain 1: 2900 -7318.956 0.015 0.013
Chain 1: 3000 -7493.712 0.017 0.014
Chain 1: 3100 -7476.914 0.017 0.014
Chain 1: 3200 -7687.023 0.017 0.014
Chain 1: 3300 -7368.797 0.019 0.014
Chain 1: 3400 -7616.145 0.020 0.023
Chain 1: 3500 -7392.629 0.022 0.027
Chain 1: 3600 -7455.395 0.022 0.027
Chain 1: 3700 -7412.747 0.022 0.027
Chain 1: 3800 -7382.635 0.022 0.027
Chain 1: 3900 -7355.039 0.018 0.023
Chain 1: 4000 -7351.216 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003078 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.78 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86863.060 1.000 1.000
Chain 1: 200 -14087.547 3.083 5.166
Chain 1: 300 -10285.142 2.179 1.000
Chain 1: 400 -12239.138 1.674 1.000
Chain 1: 500 -8678.620 1.421 0.410
Chain 1: 600 -8973.479 1.190 0.410
Chain 1: 700 -8573.463 1.026 0.370
Chain 1: 800 -8872.315 0.902 0.370
Chain 1: 900 -9013.331 0.804 0.160
Chain 1: 1000 -9024.402 0.724 0.160
Chain 1: 1100 -8946.607 0.624 0.047 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8396.536 0.114 0.047
Chain 1: 1300 -8862.153 0.083 0.047
Chain 1: 1400 -8760.878 0.068 0.034
Chain 1: 1500 -8716.092 0.027 0.033
Chain 1: 1600 -8834.127 0.025 0.016
Chain 1: 1700 -8872.422 0.021 0.013
Chain 1: 1800 -8409.623 0.023 0.013
Chain 1: 1900 -8520.304 0.023 0.013
Chain 1: 2000 -8540.364 0.023 0.013
Chain 1: 2100 -8622.917 0.023 0.013
Chain 1: 2200 -8402.547 0.019 0.013
Chain 1: 2300 -8615.859 0.017 0.013
Chain 1: 2400 -8411.131 0.018 0.013
Chain 1: 2500 -8488.217 0.018 0.013
Chain 1: 2600 -8395.498 0.018 0.013
Chain 1: 2700 -8433.108 0.018 0.013
Chain 1: 2800 -8385.571 0.013 0.011
Chain 1: 2900 -8499.084 0.013 0.011
Chain 1: 3000 -8407.716 0.014 0.011
Chain 1: 3100 -8375.619 0.013 0.011
Chain 1: 3200 -8346.142 0.011 0.011
Chain 1: 3300 -8611.647 0.012 0.011
Chain 1: 3400 -8660.001 0.010 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003401 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.01 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8435073.711 1.000 1.000
Chain 1: 200 -1585660.889 2.660 4.320
Chain 1: 300 -890598.278 2.033 1.000
Chain 1: 400 -457909.684 1.761 1.000
Chain 1: 500 -358021.705 1.465 0.945
Chain 1: 600 -233062.368 1.310 0.945
Chain 1: 700 -119552.472 1.259 0.945
Chain 1: 800 -86857.625 1.148 0.945
Chain 1: 900 -67257.670 1.053 0.780
Chain 1: 1000 -52114.140 0.977 0.780
Chain 1: 1100 -39640.235 0.908 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38831.014 0.478 0.376
Chain 1: 1300 -26817.900 0.445 0.376
Chain 1: 1400 -26544.917 0.352 0.315
Chain 1: 1500 -23139.863 0.338 0.315
Chain 1: 1600 -22360.691 0.288 0.291
Chain 1: 1700 -21236.717 0.199 0.291
Chain 1: 1800 -21182.315 0.161 0.147
Chain 1: 1900 -21509.350 0.134 0.053
Chain 1: 2000 -20020.410 0.112 0.053
Chain 1: 2100 -20258.709 0.082 0.035
Chain 1: 2200 -20485.621 0.081 0.035
Chain 1: 2300 -20102.249 0.038 0.019
Chain 1: 2400 -19874.057 0.038 0.019
Chain 1: 2500 -19675.984 0.024 0.015
Chain 1: 2600 -19305.229 0.023 0.015
Chain 1: 2700 -19262.014 0.018 0.012
Chain 1: 2800 -18978.390 0.019 0.015
Chain 1: 2900 -19260.070 0.019 0.015
Chain 1: 3000 -19246.160 0.012 0.012
Chain 1: 3100 -19331.276 0.011 0.011
Chain 1: 3200 -19021.337 0.011 0.015
Chain 1: 3300 -19226.603 0.010 0.011
Chain 1: 3400 -18700.375 0.012 0.015
Chain 1: 3500 -19313.861 0.014 0.015
Chain 1: 3600 -18618.452 0.016 0.015
Chain 1: 3700 -19006.757 0.018 0.016
Chain 1: 3800 -17963.151 0.022 0.020
Chain 1: 3900 -17959.205 0.021 0.020
Chain 1: 4000 -18076.534 0.021 0.020
Chain 1: 4100 -17990.090 0.021 0.020
Chain 1: 4200 -17805.653 0.021 0.020
Chain 1: 4300 -17944.538 0.021 0.020
Chain 1: 4400 -17900.761 0.018 0.010
Chain 1: 4500 -17803.189 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001132 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 11.32 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48479.941 1.000 1.000
Chain 1: 200 -19822.309 1.223 1.446
Chain 1: 300 -17785.476 0.853 1.000
Chain 1: 400 -12857.553 0.736 1.000
Chain 1: 500 -12110.503 0.601 0.383
Chain 1: 600 -15469.379 0.537 0.383
Chain 1: 700 -14633.238 0.468 0.217
Chain 1: 800 -19932.862 0.443 0.266
Chain 1: 900 -14906.289 0.431 0.266
Chain 1: 1000 -12376.194 0.409 0.266
Chain 1: 1100 -11355.591 0.318 0.217
Chain 1: 1200 -15222.907 0.199 0.217
Chain 1: 1300 -12142.984 0.212 0.254
Chain 1: 1400 -10256.790 0.192 0.217
Chain 1: 1500 -9788.266 0.191 0.217
Chain 1: 1600 -10809.520 0.179 0.204
Chain 1: 1700 -9751.227 0.184 0.204
Chain 1: 1800 -9688.162 0.158 0.184
Chain 1: 1900 -10349.353 0.131 0.109
Chain 1: 2000 -17601.481 0.151 0.109
Chain 1: 2100 -9259.255 0.233 0.184
Chain 1: 2200 -12715.476 0.234 0.184
Chain 1: 2300 -9304.860 0.246 0.184
Chain 1: 2400 -10119.036 0.235 0.109
Chain 1: 2500 -17858.421 0.274 0.272
Chain 1: 2600 -10161.305 0.340 0.367
Chain 1: 2700 -9051.153 0.342 0.367
Chain 1: 2800 -10943.311 0.358 0.367
Chain 1: 2900 -8818.739 0.376 0.367
Chain 1: 3000 -9143.352 0.338 0.272
Chain 1: 3100 -15010.586 0.287 0.272
Chain 1: 3200 -9325.253 0.321 0.367
Chain 1: 3300 -17711.038 0.332 0.391
Chain 1: 3400 -9886.087 0.403 0.433
Chain 1: 3500 -13395.000 0.386 0.391
Chain 1: 3600 -9840.927 0.346 0.361
Chain 1: 3700 -8947.918 0.344 0.361
Chain 1: 3800 -9080.391 0.328 0.361
Chain 1: 3900 -8945.285 0.305 0.361
Chain 1: 4000 -10280.143 0.315 0.361
Chain 1: 4100 -8690.758 0.294 0.262
Chain 1: 4200 -13276.655 0.268 0.262
Chain 1: 4300 -8868.991 0.270 0.262
Chain 1: 4400 -8828.775 0.191 0.183
Chain 1: 4500 -8626.101 0.167 0.130
Chain 1: 4600 -12355.570 0.161 0.130
Chain 1: 4700 -10736.611 0.167 0.151
Chain 1: 4800 -8526.714 0.191 0.183
Chain 1: 4900 -8662.299 0.191 0.183
Chain 1: 5000 -9308.423 0.185 0.183
Chain 1: 5100 -9147.293 0.168 0.151
Chain 1: 5200 -14283.521 0.170 0.151
Chain 1: 5300 -9410.913 0.172 0.151
Chain 1: 5400 -8559.577 0.181 0.151
Chain 1: 5500 -8450.023 0.180 0.151
Chain 1: 5600 -9390.756 0.160 0.100
Chain 1: 5700 -8776.281 0.152 0.099
Chain 1: 5800 -13049.853 0.159 0.099
Chain 1: 5900 -9231.835 0.199 0.100
Chain 1: 6000 -12065.524 0.215 0.235
Chain 1: 6100 -11469.137 0.219 0.235
Chain 1: 6200 -9916.799 0.198 0.157
Chain 1: 6300 -10771.456 0.155 0.100
Chain 1: 6400 -12020.341 0.155 0.104
Chain 1: 6500 -8533.151 0.195 0.157
Chain 1: 6600 -8370.059 0.187 0.157
Chain 1: 6700 -8499.060 0.181 0.157
Chain 1: 6800 -8384.993 0.150 0.104
Chain 1: 6900 -11012.647 0.132 0.104
Chain 1: 7000 -11910.602 0.116 0.079
Chain 1: 7100 -9118.913 0.142 0.104
Chain 1: 7200 -10285.213 0.137 0.104
Chain 1: 7300 -10814.515 0.134 0.104
Chain 1: 7400 -8582.046 0.150 0.113
Chain 1: 7500 -9618.513 0.120 0.108
Chain 1: 7600 -8406.273 0.132 0.113
Chain 1: 7700 -12085.500 0.161 0.144
Chain 1: 7800 -8759.261 0.198 0.239
Chain 1: 7900 -11006.415 0.194 0.204
Chain 1: 8000 -8233.756 0.221 0.260
Chain 1: 8100 -8839.337 0.197 0.204
Chain 1: 8200 -8321.827 0.192 0.204
Chain 1: 8300 -8230.614 0.188 0.204
Chain 1: 8400 -9778.686 0.178 0.158
Chain 1: 8500 -10087.506 0.170 0.158
Chain 1: 8600 -8941.320 0.168 0.158
Chain 1: 8700 -8403.009 0.144 0.128
Chain 1: 8800 -8271.892 0.108 0.069
Chain 1: 8900 -8477.902 0.090 0.064
Chain 1: 9000 -8394.147 0.057 0.062
Chain 1: 9100 -8234.393 0.052 0.031
Chain 1: 9200 -11829.460 0.077 0.031
Chain 1: 9300 -8094.663 0.122 0.064
Chain 1: 9400 -8178.249 0.107 0.031
Chain 1: 9500 -8039.739 0.105 0.024
Chain 1: 9600 -8689.596 0.100 0.024
Chain 1: 9700 -8640.911 0.094 0.019
Chain 1: 9800 -8428.113 0.095 0.024
Chain 1: 9900 -8693.410 0.096 0.025
Chain 1: 10000 -8172.890 0.101 0.031
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001372 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.72 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56821.559 1.000 1.000
Chain 1: 200 -17230.714 1.649 2.298
Chain 1: 300 -8691.330 1.427 1.000
Chain 1: 400 -8278.692 1.083 1.000
Chain 1: 500 -8413.913 0.869 0.983
Chain 1: 600 -8496.769 0.726 0.983
Chain 1: 700 -7945.201 0.632 0.069
Chain 1: 800 -8090.333 0.555 0.069
Chain 1: 900 -7970.997 0.495 0.050
Chain 1: 1000 -7885.597 0.447 0.050
Chain 1: 1100 -7911.298 0.347 0.018
Chain 1: 1200 -7752.070 0.120 0.018
Chain 1: 1300 -7776.503 0.022 0.016
Chain 1: 1400 -7928.018 0.019 0.016
Chain 1: 1500 -7699.522 0.020 0.018
Chain 1: 1600 -7655.800 0.019 0.018
Chain 1: 1700 -7597.252 0.013 0.015
Chain 1: 1800 -7657.985 0.012 0.011
Chain 1: 1900 -7631.657 0.011 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003637 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 36.37 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86349.962 1.000 1.000
Chain 1: 200 -13276.509 3.252 5.504
Chain 1: 300 -9756.004 2.288 1.000
Chain 1: 400 -10568.039 1.735 1.000
Chain 1: 500 -8614.556 1.434 0.361
Chain 1: 600 -8333.987 1.200 0.361
Chain 1: 700 -8730.554 1.035 0.227
Chain 1: 800 -8806.466 0.907 0.227
Chain 1: 900 -8628.641 0.809 0.077
Chain 1: 1000 -8359.096 0.731 0.077
Chain 1: 1100 -8586.507 0.634 0.045 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8322.386 0.086 0.034
Chain 1: 1300 -8355.578 0.051 0.032
Chain 1: 1400 -8350.089 0.043 0.032
Chain 1: 1500 -8383.950 0.021 0.026
Chain 1: 1600 -8389.361 0.017 0.021
Chain 1: 1700 -8324.988 0.014 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003092 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.92 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8419135.992 1.000 1.000
Chain 1: 200 -1587471.386 2.652 4.303
Chain 1: 300 -891510.007 2.028 1.000
Chain 1: 400 -457542.075 1.758 1.000
Chain 1: 500 -357461.005 1.463 0.948
Chain 1: 600 -232385.455 1.308 0.948
Chain 1: 700 -118791.900 1.258 0.948
Chain 1: 800 -86025.403 1.148 0.948
Chain 1: 900 -66407.971 1.054 0.781
Chain 1: 1000 -51227.825 0.978 0.781
Chain 1: 1100 -38732.029 0.910 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37906.483 0.482 0.381
Chain 1: 1300 -25905.896 0.450 0.381
Chain 1: 1400 -25625.197 0.357 0.323
Chain 1: 1500 -22224.188 0.344 0.323
Chain 1: 1600 -21443.002 0.294 0.296
Chain 1: 1700 -20322.933 0.204 0.295
Chain 1: 1800 -20267.985 0.166 0.153
Chain 1: 1900 -20593.425 0.138 0.055
Chain 1: 2000 -19109.124 0.116 0.055
Chain 1: 2100 -19347.175 0.085 0.036
Chain 1: 2200 -19572.632 0.084 0.036
Chain 1: 2300 -19190.934 0.040 0.020
Chain 1: 2400 -18963.371 0.040 0.020
Chain 1: 2500 -18765.151 0.025 0.016
Chain 1: 2600 -18396.256 0.024 0.016
Chain 1: 2700 -18353.539 0.018 0.012
Chain 1: 2800 -18070.616 0.020 0.016
Chain 1: 2900 -18351.500 0.020 0.015
Chain 1: 3000 -18337.787 0.012 0.012
Chain 1: 3100 -18422.628 0.011 0.012
Chain 1: 3200 -18113.833 0.012 0.015
Chain 1: 3300 -18318.173 0.011 0.012
Chain 1: 3400 -17793.929 0.013 0.015
Chain 1: 3500 -18404.412 0.015 0.016
Chain 1: 3600 -17712.980 0.017 0.016
Chain 1: 3700 -18098.336 0.019 0.017
Chain 1: 3800 -17060.828 0.023 0.021
Chain 1: 3900 -17057.034 0.022 0.021
Chain 1: 4000 -17174.369 0.022 0.021
Chain 1: 4100 -17088.212 0.022 0.021
Chain 1: 4200 -16905.114 0.022 0.021
Chain 1: 4300 -17043.074 0.021 0.021
Chain 1: 4400 -17000.412 0.019 0.011
Chain 1: 4500 -16903.022 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001341 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.41 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -50037.111 1.000 1.000
Chain 1: 200 -20992.048 1.192 1.384
Chain 1: 300 -18959.761 0.830 1.000
Chain 1: 400 -21773.815 0.655 1.000
Chain 1: 500 -14926.097 0.616 0.459
Chain 1: 600 -13749.186 0.527 0.459
Chain 1: 700 -18560.377 0.489 0.259
Chain 1: 800 -11922.218 0.498 0.459
Chain 1: 900 -15401.494 0.467 0.259
Chain 1: 1000 -13790.527 0.432 0.259
Chain 1: 1100 -13252.643 0.336 0.226
Chain 1: 1200 -12900.935 0.201 0.129
Chain 1: 1300 -12772.392 0.191 0.129
Chain 1: 1400 -11277.943 0.191 0.133
Chain 1: 1500 -11644.301 0.149 0.117
Chain 1: 1600 -12260.408 0.145 0.117
Chain 1: 1700 -11013.658 0.130 0.113
Chain 1: 1800 -18283.168 0.115 0.113
Chain 1: 1900 -10543.792 0.165 0.113
Chain 1: 2000 -20057.666 0.201 0.113
Chain 1: 2100 -10417.937 0.290 0.133
Chain 1: 2200 -10950.050 0.292 0.133
Chain 1: 2300 -17477.085 0.328 0.373
Chain 1: 2400 -9860.575 0.392 0.398
Chain 1: 2500 -10103.921 0.391 0.398
Chain 1: 2600 -9814.828 0.389 0.398
Chain 1: 2700 -10598.414 0.385 0.398
Chain 1: 2800 -9985.233 0.352 0.373
Chain 1: 2900 -15545.342 0.314 0.358
Chain 1: 3000 -13584.613 0.281 0.144
Chain 1: 3100 -10865.781 0.214 0.144
Chain 1: 3200 -14117.476 0.232 0.230
Chain 1: 3300 -16420.447 0.208 0.144
Chain 1: 3400 -9497.918 0.204 0.144
Chain 1: 3500 -10174.086 0.208 0.144
Chain 1: 3600 -9659.067 0.211 0.144
Chain 1: 3700 -9714.142 0.204 0.144
Chain 1: 3800 -12294.556 0.219 0.210
Chain 1: 3900 -10630.276 0.199 0.157
Chain 1: 4000 -10132.144 0.189 0.157
Chain 1: 4100 -9599.846 0.170 0.140
Chain 1: 4200 -11444.036 0.163 0.140
Chain 1: 4300 -11997.985 0.153 0.066
Chain 1: 4400 -9959.591 0.101 0.066
Chain 1: 4500 -9406.487 0.100 0.059
Chain 1: 4600 -9450.563 0.095 0.059
Chain 1: 4700 -10283.957 0.103 0.081
Chain 1: 4800 -15243.144 0.114 0.081
Chain 1: 4900 -9617.974 0.157 0.081
Chain 1: 5000 -17451.828 0.197 0.161
Chain 1: 5100 -9043.422 0.285 0.205
Chain 1: 5200 -9187.342 0.270 0.205
Chain 1: 5300 -12230.314 0.290 0.249
Chain 1: 5400 -9214.052 0.303 0.325
Chain 1: 5500 -9105.246 0.298 0.325
Chain 1: 5600 -9447.349 0.301 0.325
Chain 1: 5700 -9921.076 0.298 0.325
Chain 1: 5800 -9037.638 0.275 0.249
Chain 1: 5900 -14677.744 0.255 0.249
Chain 1: 6000 -15044.996 0.212 0.098
Chain 1: 6100 -9892.802 0.171 0.098
Chain 1: 6200 -9911.348 0.170 0.098
Chain 1: 6300 -9448.650 0.150 0.049
Chain 1: 6400 -11452.799 0.135 0.049
Chain 1: 6500 -9681.837 0.152 0.098
Chain 1: 6600 -9792.779 0.150 0.098
Chain 1: 6700 -8950.760 0.154 0.098
Chain 1: 6800 -9930.875 0.154 0.099
Chain 1: 6900 -8765.688 0.129 0.099
Chain 1: 7000 -10101.550 0.140 0.132
Chain 1: 7100 -8838.717 0.102 0.132
Chain 1: 7200 -9431.968 0.108 0.132
Chain 1: 7300 -9100.131 0.107 0.132
Chain 1: 7400 -9284.315 0.091 0.099
Chain 1: 7500 -11437.061 0.092 0.099
Chain 1: 7600 -9177.561 0.115 0.132
Chain 1: 7700 -14877.217 0.144 0.133
Chain 1: 7800 -13564.880 0.144 0.133
Chain 1: 7900 -9579.454 0.172 0.143
Chain 1: 8000 -8985.822 0.166 0.143
Chain 1: 8100 -9326.112 0.155 0.097
Chain 1: 8200 -11247.410 0.166 0.171
Chain 1: 8300 -8843.236 0.190 0.188
Chain 1: 8400 -9517.118 0.195 0.188
Chain 1: 8500 -8832.476 0.184 0.171
Chain 1: 8600 -9012.702 0.161 0.097
Chain 1: 8700 -8762.979 0.125 0.078
Chain 1: 8800 -10715.154 0.134 0.078
Chain 1: 8900 -11822.936 0.102 0.078
Chain 1: 9000 -12368.193 0.100 0.078
Chain 1: 9100 -9097.697 0.132 0.094
Chain 1: 9200 -9735.024 0.121 0.078
Chain 1: 9300 -9719.950 0.094 0.071
Chain 1: 9400 -9440.794 0.090 0.065
Chain 1: 9500 -8850.998 0.089 0.065
Chain 1: 9600 -9538.897 0.094 0.067
Chain 1: 9700 -9355.067 0.093 0.067
Chain 1: 9800 -9037.268 0.079 0.065
Chain 1: 9900 -12897.369 0.099 0.065
Chain 1: 10000 -9063.051 0.137 0.067
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001481 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.81 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -64419.336 1.000 1.000
Chain 1: 200 -19034.432 1.692 2.384
Chain 1: 300 -9191.403 1.485 1.071
Chain 1: 400 -8168.314 1.145 1.071
Chain 1: 500 -8348.280 0.920 1.000
Chain 1: 600 -8116.684 0.772 1.000
Chain 1: 700 -9135.161 0.677 0.125
Chain 1: 800 -8589.319 0.601 0.125
Chain 1: 900 -8025.537 0.542 0.111
Chain 1: 1000 -8154.863 0.489 0.111
Chain 1: 1100 -7851.469 0.393 0.070
Chain 1: 1200 -7714.942 0.156 0.064
Chain 1: 1300 -7848.140 0.051 0.039
Chain 1: 1400 -7803.554 0.039 0.029
Chain 1: 1500 -7583.844 0.040 0.029
Chain 1: 1600 -7761.912 0.039 0.029
Chain 1: 1700 -7708.217 0.029 0.023
Chain 1: 1800 -7711.083 0.022 0.018
Chain 1: 1900 -7607.966 0.017 0.017
Chain 1: 2000 -7571.896 0.016 0.017
Chain 1: 2100 -7626.547 0.013 0.014
Chain 1: 2200 -7877.115 0.014 0.014
Chain 1: 2300 -7647.993 0.015 0.014
Chain 1: 2400 -7596.908 0.015 0.014
Chain 1: 2500 -7642.507 0.013 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003825 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 38.25 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86229.843 1.000 1.000
Chain 1: 200 -14274.244 3.020 5.041
Chain 1: 300 -10554.474 2.131 1.000
Chain 1: 400 -11887.366 1.626 1.000
Chain 1: 500 -9542.001 1.350 0.352
Chain 1: 600 -9243.732 1.131 0.352
Chain 1: 700 -9625.875 0.975 0.246
Chain 1: 800 -8919.811 0.863 0.246
Chain 1: 900 -8819.124 0.768 0.112
Chain 1: 1000 -9580.290 0.699 0.112
Chain 1: 1100 -9157.716 0.604 0.079 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -9395.085 0.102 0.079
Chain 1: 1300 -8925.321 0.072 0.053
Chain 1: 1400 -8973.698 0.062 0.046
Chain 1: 1500 -8946.526 0.037 0.040
Chain 1: 1600 -8949.993 0.034 0.040
Chain 1: 1700 -8830.389 0.032 0.025
Chain 1: 1800 -8887.639 0.024 0.014
Chain 1: 1900 -8766.999 0.025 0.014
Chain 1: 2000 -8831.103 0.017 0.014
Chain 1: 2100 -8985.875 0.014 0.014
Chain 1: 2200 -8762.499 0.015 0.014
Chain 1: 2300 -8892.781 0.011 0.014
Chain 1: 2400 -8761.893 0.012 0.014
Chain 1: 2500 -8831.191 0.012 0.014
Chain 1: 2600 -8748.775 0.013 0.014
Chain 1: 2700 -8777.070 0.012 0.014
Chain 1: 2800 -8731.377 0.012 0.014
Chain 1: 2900 -8833.008 0.012 0.012
Chain 1: 3000 -8730.128 0.012 0.012
Chain 1: 3100 -8718.665 0.011 0.012
Chain 1: 3200 -8693.981 0.008 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003311 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.11 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8423262.417 1.000 1.000
Chain 1: 200 -1586678.972 2.654 4.309
Chain 1: 300 -890287.738 2.030 1.000
Chain 1: 400 -457883.276 1.759 1.000
Chain 1: 500 -358031.804 1.463 0.944
Chain 1: 600 -233199.971 1.308 0.944
Chain 1: 700 -119727.763 1.257 0.944
Chain 1: 800 -87036.217 1.147 0.944
Chain 1: 900 -67440.573 1.051 0.782
Chain 1: 1000 -52296.276 0.975 0.782
Chain 1: 1100 -39821.453 0.907 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39008.624 0.478 0.376
Chain 1: 1300 -26996.358 0.444 0.376
Chain 1: 1400 -26721.884 0.351 0.313
Chain 1: 1500 -23317.353 0.337 0.313
Chain 1: 1600 -22537.653 0.287 0.291
Chain 1: 1700 -21414.050 0.198 0.290
Chain 1: 1800 -21359.475 0.161 0.146
Chain 1: 1900 -21686.285 0.133 0.052
Chain 1: 2000 -20197.933 0.111 0.052
Chain 1: 2100 -20436.191 0.081 0.035
Chain 1: 2200 -20662.923 0.080 0.035
Chain 1: 2300 -20279.740 0.038 0.019
Chain 1: 2400 -20051.654 0.038 0.019
Chain 1: 2500 -19853.638 0.024 0.015
Chain 1: 2600 -19483.239 0.023 0.015
Chain 1: 2700 -19440.069 0.018 0.012
Chain 1: 2800 -19156.634 0.019 0.015
Chain 1: 2900 -19438.111 0.019 0.014
Chain 1: 3000 -19424.287 0.011 0.012
Chain 1: 3100 -19509.361 0.011 0.011
Chain 1: 3200 -19199.661 0.011 0.014
Chain 1: 3300 -19404.702 0.010 0.011
Chain 1: 3400 -18878.894 0.012 0.014
Chain 1: 3500 -19491.818 0.014 0.015
Chain 1: 3600 -18797.108 0.016 0.015
Chain 1: 3700 -19184.913 0.018 0.016
Chain 1: 3800 -18142.464 0.022 0.020
Chain 1: 3900 -18138.550 0.021 0.020
Chain 1: 4000 -18255.870 0.021 0.020
Chain 1: 4100 -18169.519 0.021 0.020
Chain 1: 4200 -17985.309 0.021 0.020
Chain 1: 4300 -18124.036 0.020 0.020
Chain 1: 4400 -18080.454 0.018 0.010
Chain 1: 4500 -17982.930 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001772 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 17.72 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12276.873 1.000 1.000
Chain 1: 200 -9138.272 0.672 1.000
Chain 1: 300 -7949.284 0.498 0.343
Chain 1: 400 -8086.097 0.377 0.343
Chain 1: 500 -8075.790 0.302 0.150
Chain 1: 600 -7860.803 0.256 0.150
Chain 1: 700 -7766.079 0.222 0.027
Chain 1: 800 -7808.489 0.195 0.027
Chain 1: 900 -7932.519 0.175 0.017
Chain 1: 1000 -7810.055 0.159 0.017
Chain 1: 1100 -7852.142 0.059 0.016
Chain 1: 1200 -7791.106 0.026 0.016
Chain 1: 1300 -7728.892 0.012 0.012
Chain 1: 1400 -7768.859 0.010 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001388 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.88 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61777.239 1.000 1.000
Chain 1: 200 -17726.900 1.742 2.485
Chain 1: 300 -8843.409 1.496 1.005
Chain 1: 400 -8353.271 1.137 1.005
Chain 1: 500 -8320.110 0.910 1.000
Chain 1: 600 -8135.334 0.762 1.000
Chain 1: 700 -8103.370 0.654 0.059
Chain 1: 800 -7988.713 0.574 0.059
Chain 1: 900 -7691.539 0.515 0.039
Chain 1: 1000 -7934.015 0.466 0.039
Chain 1: 1100 -7829.068 0.368 0.031
Chain 1: 1200 -7693.237 0.121 0.023
Chain 1: 1300 -7735.843 0.021 0.018
Chain 1: 1400 -7678.659 0.016 0.014
Chain 1: 1500 -7571.614 0.017 0.014
Chain 1: 1600 -7615.786 0.015 0.014
Chain 1: 1700 -7530.937 0.016 0.014
Chain 1: 1800 -7602.525 0.015 0.013
Chain 1: 1900 -7632.443 0.012 0.011
Chain 1: 2000 -7612.569 0.009 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003393 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.93 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86705.230 1.000 1.000
Chain 1: 200 -13417.311 3.231 5.462
Chain 1: 300 -9769.847 2.279 1.000
Chain 1: 400 -10837.970 1.734 1.000
Chain 1: 500 -8744.009 1.435 0.373
Chain 1: 600 -8399.393 1.202 0.373
Chain 1: 700 -8212.111 1.034 0.239
Chain 1: 800 -9036.817 0.916 0.239
Chain 1: 900 -8621.750 0.820 0.099
Chain 1: 1000 -8500.564 0.739 0.099
Chain 1: 1100 -8601.457 0.640 0.091 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8109.998 0.100 0.061
Chain 1: 1300 -8450.759 0.067 0.048
Chain 1: 1400 -8428.245 0.057 0.041
Chain 1: 1500 -8336.147 0.034 0.040
Chain 1: 1600 -8443.112 0.032 0.023
Chain 1: 1700 -8513.138 0.030 0.014
Chain 1: 1800 -8095.720 0.026 0.014
Chain 1: 1900 -8193.304 0.022 0.013
Chain 1: 2000 -8167.581 0.021 0.012
Chain 1: 2100 -8292.046 0.022 0.013
Chain 1: 2200 -8102.733 0.018 0.013
Chain 1: 2300 -8188.263 0.015 0.012
Chain 1: 2400 -8257.620 0.016 0.012
Chain 1: 2500 -8203.624 0.015 0.012
Chain 1: 2600 -8204.154 0.014 0.010
Chain 1: 2700 -8121.264 0.014 0.010
Chain 1: 2800 -8082.427 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003354 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.54 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8444997.290 1.000 1.000
Chain 1: 200 -1592936.509 2.651 4.302
Chain 1: 300 -891638.405 2.029 1.000
Chain 1: 400 -457513.009 1.759 1.000
Chain 1: 500 -356860.707 1.464 0.949
Chain 1: 600 -231684.673 1.310 0.949
Chain 1: 700 -118449.977 1.259 0.949
Chain 1: 800 -85826.041 1.149 0.949
Chain 1: 900 -66297.697 1.054 0.787
Chain 1: 1000 -51211.702 0.978 0.787
Chain 1: 1100 -38797.738 0.910 0.540 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37989.543 0.482 0.380
Chain 1: 1300 -26049.015 0.450 0.380
Chain 1: 1400 -25779.648 0.356 0.320
Chain 1: 1500 -22393.589 0.343 0.320
Chain 1: 1600 -21618.839 0.292 0.295
Chain 1: 1700 -20504.325 0.202 0.295
Chain 1: 1800 -20451.538 0.164 0.151
Chain 1: 1900 -20777.762 0.136 0.054
Chain 1: 2000 -19295.058 0.115 0.054
Chain 1: 2100 -19533.144 0.084 0.036
Chain 1: 2200 -19758.671 0.083 0.036
Chain 1: 2300 -19376.661 0.039 0.020
Chain 1: 2400 -19148.829 0.039 0.020
Chain 1: 2500 -18950.510 0.025 0.016
Chain 1: 2600 -18581.008 0.024 0.016
Chain 1: 2700 -18538.109 0.018 0.012
Chain 1: 2800 -18254.759 0.020 0.016
Chain 1: 2900 -18535.881 0.020 0.015
Chain 1: 3000 -18522.181 0.012 0.012
Chain 1: 3100 -18607.174 0.011 0.012
Chain 1: 3200 -18297.903 0.012 0.015
Chain 1: 3300 -18502.612 0.011 0.012
Chain 1: 3400 -17977.478 0.013 0.015
Chain 1: 3500 -18589.285 0.015 0.016
Chain 1: 3600 -17895.974 0.017 0.016
Chain 1: 3700 -18282.652 0.019 0.017
Chain 1: 3800 -17242.383 0.023 0.021
Chain 1: 3900 -17238.461 0.022 0.021
Chain 1: 4000 -17355.820 0.022 0.021
Chain 1: 4100 -17269.555 0.022 0.021
Chain 1: 4200 -17085.822 0.022 0.021
Chain 1: 4300 -17224.248 0.021 0.021
Chain 1: 4400 -17181.055 0.019 0.011
Chain 1: 4500 -17083.540 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001916 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 19.16 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12334.340 1.000 1.000
Chain 1: 200 -9127.771 0.676 1.000
Chain 1: 300 -7994.152 0.498 0.351
Chain 1: 400 -8097.544 0.376 0.351
Chain 1: 500 -8173.859 0.303 0.142
Chain 1: 600 -7897.832 0.258 0.142
Chain 1: 700 -7856.479 0.222 0.035
Chain 1: 800 -7824.578 0.195 0.035
Chain 1: 900 -7803.172 0.174 0.013
Chain 1: 1000 -7863.283 0.157 0.013
Chain 1: 1100 -7964.733 0.058 0.013
Chain 1: 1200 -7914.280 0.024 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001495 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.95 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -63144.473 1.000 1.000
Chain 1: 200 -18083.416 1.746 2.492
Chain 1: 300 -8703.796 1.523 1.078
Chain 1: 400 -8457.478 1.150 1.078
Chain 1: 500 -8341.054 0.923 1.000
Chain 1: 600 -8516.352 0.772 1.000
Chain 1: 700 -7674.909 0.678 0.110
Chain 1: 800 -8138.646 0.600 0.110
Chain 1: 900 -7842.104 0.538 0.057
Chain 1: 1000 -7713.774 0.485 0.057
Chain 1: 1100 -7630.753 0.387 0.038
Chain 1: 1200 -7808.934 0.140 0.029
Chain 1: 1300 -7707.012 0.033 0.023
Chain 1: 1400 -7571.215 0.032 0.021
Chain 1: 1500 -7536.913 0.031 0.021
Chain 1: 1600 -7713.013 0.031 0.023
Chain 1: 1700 -7459.334 0.024 0.023
Chain 1: 1800 -7540.315 0.019 0.018
Chain 1: 1900 -7512.957 0.016 0.017
Chain 1: 2000 -7552.622 0.015 0.013
Chain 1: 2100 -7516.954 0.014 0.013
Chain 1: 2200 -7627.647 0.013 0.013
Chain 1: 2300 -7541.256 0.013 0.011
Chain 1: 2400 -7599.109 0.012 0.011
Chain 1: 2500 -7450.684 0.013 0.011
Chain 1: 2600 -7506.571 0.012 0.011
Chain 1: 2700 -7450.619 0.009 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004175 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 41.75 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86246.114 1.000 1.000
Chain 1: 200 -13377.567 3.224 5.447
Chain 1: 300 -9775.826 2.272 1.000
Chain 1: 400 -10605.805 1.723 1.000
Chain 1: 500 -8741.315 1.421 0.368
Chain 1: 600 -8263.538 1.194 0.368
Chain 1: 700 -8345.800 1.025 0.213
Chain 1: 800 -9031.661 0.906 0.213
Chain 1: 900 -8567.978 0.812 0.078
Chain 1: 1000 -8392.664 0.733 0.078
Chain 1: 1100 -8607.508 0.635 0.076 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8259.561 0.095 0.058
Chain 1: 1300 -8465.344 0.060 0.054
Chain 1: 1400 -8469.496 0.052 0.042
Chain 1: 1500 -8357.187 0.032 0.025
Chain 1: 1600 -8461.754 0.028 0.024
Chain 1: 1700 -8550.088 0.028 0.024
Chain 1: 1800 -8143.560 0.025 0.024
Chain 1: 1900 -8240.781 0.021 0.021
Chain 1: 2000 -8212.707 0.019 0.013
Chain 1: 2100 -8333.001 0.018 0.013
Chain 1: 2200 -8134.902 0.016 0.013
Chain 1: 2300 -8279.035 0.016 0.013
Chain 1: 2400 -8285.823 0.016 0.013
Chain 1: 2500 -8254.184 0.015 0.012
Chain 1: 2600 -8252.575 0.014 0.012
Chain 1: 2700 -8165.744 0.014 0.012
Chain 1: 2800 -8131.358 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003158 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.58 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8428295.459 1.000 1.000
Chain 1: 200 -1586890.296 2.656 4.311
Chain 1: 300 -890152.124 2.031 1.000
Chain 1: 400 -457107.772 1.760 1.000
Chain 1: 500 -357069.346 1.464 0.947
Chain 1: 600 -232218.678 1.310 0.947
Chain 1: 700 -118728.723 1.259 0.947
Chain 1: 800 -86040.608 1.149 0.947
Chain 1: 900 -66447.358 1.054 0.783
Chain 1: 1000 -51299.792 0.979 0.783
Chain 1: 1100 -38831.820 0.911 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38011.350 0.482 0.380
Chain 1: 1300 -26025.270 0.449 0.380
Chain 1: 1400 -25749.466 0.356 0.321
Chain 1: 1500 -22352.013 0.343 0.321
Chain 1: 1600 -21573.054 0.293 0.295
Chain 1: 1700 -20453.718 0.203 0.295
Chain 1: 1800 -20399.434 0.165 0.152
Chain 1: 1900 -20725.380 0.137 0.055
Chain 1: 2000 -19240.445 0.115 0.055
Chain 1: 2100 -19478.706 0.084 0.036
Chain 1: 2200 -19704.446 0.083 0.036
Chain 1: 2300 -19322.283 0.039 0.020
Chain 1: 2400 -19094.479 0.039 0.020
Chain 1: 2500 -18896.298 0.025 0.016
Chain 1: 2600 -18526.967 0.024 0.016
Chain 1: 2700 -18484.024 0.018 0.012
Chain 1: 2800 -18200.921 0.020 0.016
Chain 1: 2900 -18481.935 0.020 0.015
Chain 1: 3000 -18468.201 0.012 0.012
Chain 1: 3100 -18553.190 0.011 0.012
Chain 1: 3200 -18244.040 0.012 0.015
Chain 1: 3300 -18448.586 0.011 0.012
Chain 1: 3400 -17923.819 0.013 0.015
Chain 1: 3500 -18535.219 0.015 0.016
Chain 1: 3600 -17842.371 0.017 0.016
Chain 1: 3700 -18228.799 0.019 0.017
Chain 1: 3800 -17189.308 0.023 0.021
Chain 1: 3900 -17185.397 0.022 0.021
Chain 1: 4000 -17302.734 0.022 0.021
Chain 1: 4100 -17216.583 0.022 0.021
Chain 1: 4200 -17032.931 0.022 0.021
Chain 1: 4300 -17171.273 0.021 0.021
Chain 1: 4400 -17128.230 0.019 0.011
Chain 1: 4500 -17030.722 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001483 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.83 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12895.279 1.000 1.000
Chain 1: 200 -9774.407 0.660 1.000
Chain 1: 300 -8343.941 0.497 0.319
Chain 1: 400 -8579.795 0.380 0.319
Chain 1: 500 -8482.310 0.306 0.171
Chain 1: 600 -8444.641 0.256 0.171
Chain 1: 700 -8160.094 0.224 0.035
Chain 1: 800 -8162.736 0.196 0.035
Chain 1: 900 -8254.060 0.176 0.027
Chain 1: 1000 -8195.185 0.159 0.027
Chain 1: 1100 -8261.589 0.060 0.011
Chain 1: 1200 -8184.995 0.029 0.011
Chain 1: 1300 -8141.058 0.012 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001541 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.41 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57370.223 1.000 1.000
Chain 1: 200 -18047.971 1.589 2.179
Chain 1: 300 -8979.431 1.396 1.010
Chain 1: 400 -8205.043 1.071 1.010
Chain 1: 500 -9260.328 0.879 1.000
Chain 1: 600 -8753.292 0.742 1.000
Chain 1: 700 -8322.895 0.644 0.114
Chain 1: 800 -8268.844 0.564 0.114
Chain 1: 900 -7875.233 0.507 0.094
Chain 1: 1000 -7567.590 0.460 0.094
Chain 1: 1100 -7557.724 0.361 0.058
Chain 1: 1200 -8304.980 0.152 0.058
Chain 1: 1300 -7925.413 0.055 0.052
Chain 1: 1400 -7871.524 0.047 0.050
Chain 1: 1500 -7559.802 0.039 0.048
Chain 1: 1600 -7606.835 0.034 0.041
Chain 1: 1700 -7516.565 0.030 0.041
Chain 1: 1800 -7590.093 0.031 0.041
Chain 1: 1900 -7596.209 0.026 0.012
Chain 1: 2000 -7681.573 0.023 0.011
Chain 1: 2100 -7597.653 0.024 0.011
Chain 1: 2200 -7838.568 0.018 0.011
Chain 1: 2300 -7580.952 0.016 0.011
Chain 1: 2400 -7695.996 0.017 0.012
Chain 1: 2500 -7684.003 0.013 0.011
Chain 1: 2600 -7553.836 0.014 0.012
Chain 1: 2700 -7460.655 0.014 0.012
Chain 1: 2800 -7430.948 0.014 0.012
Chain 1: 2900 -7407.922 0.014 0.012
Chain 1: 3000 -7571.834 0.015 0.015
Chain 1: 3100 -7530.892 0.015 0.015
Chain 1: 3200 -7764.326 0.014 0.015
Chain 1: 3300 -7465.494 0.015 0.015
Chain 1: 3400 -7712.940 0.017 0.017
Chain 1: 3500 -7461.084 0.020 0.022
Chain 1: 3600 -7520.184 0.019 0.022
Chain 1: 3700 -7474.037 0.018 0.022
Chain 1: 3800 -7482.341 0.018 0.022
Chain 1: 3900 -7444.603 0.018 0.022
Chain 1: 4000 -7418.511 0.017 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003056 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.56 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87243.828 1.000 1.000
Chain 1: 200 -14045.946 3.106 5.211
Chain 1: 300 -10295.951 2.192 1.000
Chain 1: 400 -11842.892 1.677 1.000
Chain 1: 500 -9179.878 1.399 0.364
Chain 1: 600 -9538.643 1.172 0.364
Chain 1: 700 -9102.918 1.012 0.290
Chain 1: 800 -9123.713 0.886 0.290
Chain 1: 900 -9140.854 0.787 0.131
Chain 1: 1000 -8761.959 0.713 0.131
Chain 1: 1100 -8926.279 0.615 0.048 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8609.075 0.097 0.043
Chain 1: 1300 -9000.883 0.065 0.043
Chain 1: 1400 -8705.091 0.056 0.038
Chain 1: 1500 -8783.530 0.027 0.037
Chain 1: 1600 -8895.117 0.025 0.034
Chain 1: 1700 -8953.678 0.021 0.018
Chain 1: 1800 -8510.847 0.026 0.034
Chain 1: 1900 -8613.969 0.027 0.034
Chain 1: 2000 -8596.445 0.023 0.018
Chain 1: 2100 -8730.138 0.022 0.015
Chain 1: 2200 -8510.455 0.021 0.015
Chain 1: 2300 -8616.622 0.018 0.013
Chain 1: 2400 -8675.475 0.015 0.012
Chain 1: 2500 -8623.026 0.015 0.012
Chain 1: 2600 -8637.336 0.014 0.012
Chain 1: 2700 -8544.564 0.014 0.012
Chain 1: 2800 -8490.445 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003583 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.83 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8399763.036 1.000 1.000
Chain 1: 200 -1584973.308 2.650 4.300
Chain 1: 300 -891693.262 2.026 1.000
Chain 1: 400 -458477.304 1.756 1.000
Chain 1: 500 -358801.090 1.460 0.945
Chain 1: 600 -233814.851 1.306 0.945
Chain 1: 700 -119930.864 1.255 0.945
Chain 1: 800 -87104.304 1.145 0.945
Chain 1: 900 -67437.040 1.050 0.777
Chain 1: 1000 -52223.130 0.974 0.777
Chain 1: 1100 -39680.673 0.906 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38861.317 0.478 0.377
Chain 1: 1300 -26788.674 0.445 0.377
Chain 1: 1400 -26507.743 0.352 0.316
Chain 1: 1500 -23086.894 0.339 0.316
Chain 1: 1600 -22301.688 0.289 0.292
Chain 1: 1700 -21171.509 0.200 0.291
Chain 1: 1800 -21115.042 0.162 0.148
Chain 1: 1900 -21441.827 0.134 0.053
Chain 1: 2000 -19949.697 0.113 0.053
Chain 1: 2100 -20188.334 0.082 0.035
Chain 1: 2200 -20415.512 0.081 0.035
Chain 1: 2300 -20031.928 0.038 0.019
Chain 1: 2400 -19803.776 0.038 0.019
Chain 1: 2500 -19605.826 0.025 0.015
Chain 1: 2600 -19235.350 0.023 0.015
Chain 1: 2700 -19192.119 0.018 0.012
Chain 1: 2800 -18908.732 0.019 0.015
Chain 1: 2900 -19190.321 0.019 0.015
Chain 1: 3000 -19176.462 0.012 0.012
Chain 1: 3100 -19261.521 0.011 0.012
Chain 1: 3200 -18951.783 0.011 0.015
Chain 1: 3300 -19156.851 0.010 0.012
Chain 1: 3400 -18631.016 0.012 0.015
Chain 1: 3500 -19244.018 0.014 0.015
Chain 1: 3600 -18549.303 0.016 0.015
Chain 1: 3700 -18937.161 0.018 0.016
Chain 1: 3800 -17894.624 0.022 0.020
Chain 1: 3900 -17890.722 0.021 0.020
Chain 1: 4000 -18008.034 0.021 0.020
Chain 1: 4100 -17921.665 0.021 0.020
Chain 1: 4200 -17737.426 0.021 0.020
Chain 1: 4300 -17876.160 0.021 0.020
Chain 1: 4400 -17832.599 0.018 0.010
Chain 1: 4500 -17735.059 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001241 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.41 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48747.354 1.000 1.000
Chain 1: 200 -13467.235 1.810 2.620
Chain 1: 300 -17507.804 1.283 1.000
Chain 1: 400 -21295.413 1.007 1.000
Chain 1: 500 -14921.350 0.891 0.427
Chain 1: 600 -15347.304 0.747 0.427
Chain 1: 700 -12226.163 0.677 0.255
Chain 1: 800 -11222.087 0.604 0.255
Chain 1: 900 -11578.860 0.540 0.231
Chain 1: 1000 -12878.931 0.496 0.231
Chain 1: 1100 -10169.600 0.423 0.231
Chain 1: 1200 -11035.313 0.168 0.178
Chain 1: 1300 -12569.318 0.158 0.122
Chain 1: 1400 -10373.935 0.161 0.122
Chain 1: 1500 -10684.250 0.121 0.101
Chain 1: 1600 -9853.246 0.127 0.101
Chain 1: 1700 -20813.303 0.154 0.101
Chain 1: 1800 -10175.413 0.250 0.122
Chain 1: 1900 -11000.765 0.254 0.122
Chain 1: 2000 -11108.972 0.245 0.122
Chain 1: 2100 -9815.960 0.231 0.122
Chain 1: 2200 -16971.435 0.266 0.132
Chain 1: 2300 -9607.332 0.330 0.212
Chain 1: 2400 -10783.166 0.320 0.132
Chain 1: 2500 -9252.755 0.334 0.165
Chain 1: 2600 -10779.535 0.339 0.165
Chain 1: 2700 -9011.098 0.306 0.165
Chain 1: 2800 -10291.296 0.214 0.142
Chain 1: 2900 -9532.369 0.215 0.142
Chain 1: 3000 -14113.016 0.246 0.165
Chain 1: 3100 -10009.900 0.274 0.196
Chain 1: 3200 -9894.611 0.233 0.165
Chain 1: 3300 -13056.006 0.180 0.165
Chain 1: 3400 -9639.296 0.205 0.196
Chain 1: 3500 -9009.398 0.195 0.196
Chain 1: 3600 -11683.751 0.204 0.229
Chain 1: 3700 -13295.327 0.197 0.229
Chain 1: 3800 -9737.753 0.221 0.242
Chain 1: 3900 -11268.599 0.226 0.242
Chain 1: 4000 -10151.775 0.205 0.229
Chain 1: 4100 -10010.366 0.165 0.136
Chain 1: 4200 -9932.808 0.165 0.136
Chain 1: 4300 -9096.854 0.150 0.121
Chain 1: 4400 -10594.424 0.129 0.121
Chain 1: 4500 -8755.330 0.143 0.136
Chain 1: 4600 -9267.219 0.125 0.121
Chain 1: 4700 -12687.076 0.140 0.136
Chain 1: 4800 -8900.334 0.146 0.136
Chain 1: 4900 -9510.107 0.139 0.110
Chain 1: 5000 -10764.943 0.140 0.117
Chain 1: 5100 -9097.899 0.157 0.141
Chain 1: 5200 -8781.188 0.159 0.141
Chain 1: 5300 -9123.703 0.154 0.141
Chain 1: 5400 -9116.710 0.140 0.117
Chain 1: 5500 -13585.369 0.152 0.117
Chain 1: 5600 -10747.673 0.173 0.183
Chain 1: 5700 -8864.686 0.167 0.183
Chain 1: 5800 -12347.356 0.153 0.183
Chain 1: 5900 -15897.452 0.168 0.212
Chain 1: 6000 -8839.487 0.237 0.223
Chain 1: 6100 -9247.185 0.223 0.223
Chain 1: 6200 -8302.265 0.231 0.223
Chain 1: 6300 -13438.793 0.265 0.264
Chain 1: 6400 -12674.581 0.271 0.264
Chain 1: 6500 -8280.700 0.291 0.264
Chain 1: 6600 -8706.149 0.270 0.223
Chain 1: 6700 -10297.050 0.264 0.223
Chain 1: 6800 -8606.754 0.255 0.196
Chain 1: 6900 -11624.507 0.259 0.196
Chain 1: 7000 -8428.193 0.217 0.196
Chain 1: 7100 -13079.901 0.248 0.260
Chain 1: 7200 -8470.691 0.291 0.356
Chain 1: 7300 -10509.755 0.272 0.260
Chain 1: 7400 -8412.107 0.291 0.260
Chain 1: 7500 -11120.518 0.263 0.249
Chain 1: 7600 -8467.984 0.289 0.260
Chain 1: 7700 -8743.955 0.277 0.260
Chain 1: 7800 -8556.364 0.259 0.260
Chain 1: 7900 -9240.581 0.241 0.249
Chain 1: 8000 -8524.268 0.211 0.244
Chain 1: 8100 -8343.954 0.178 0.194
Chain 1: 8200 -9891.925 0.139 0.156
Chain 1: 8300 -8197.953 0.140 0.156
Chain 1: 8400 -13890.759 0.156 0.156
Chain 1: 8500 -8335.749 0.199 0.156
Chain 1: 8600 -8207.349 0.169 0.084
Chain 1: 8700 -10425.393 0.187 0.156
Chain 1: 8800 -8489.534 0.208 0.207
Chain 1: 8900 -8386.210 0.201 0.207
Chain 1: 9000 -8680.342 0.196 0.207
Chain 1: 9100 -10488.357 0.211 0.207
Chain 1: 9200 -11076.234 0.201 0.207
Chain 1: 9300 -8538.413 0.210 0.213
Chain 1: 9400 -10050.422 0.184 0.172
Chain 1: 9500 -8040.240 0.143 0.172
Chain 1: 9600 -9040.634 0.152 0.172
Chain 1: 9700 -8421.389 0.138 0.150
Chain 1: 9800 -8635.758 0.118 0.111
Chain 1: 9900 -8428.932 0.119 0.111
Chain 1: 10000 -8282.377 0.117 0.111
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001401 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.01 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57648.974 1.000 1.000
Chain 1: 200 -17550.704 1.642 2.285
Chain 1: 300 -8647.219 1.438 1.030
Chain 1: 400 -8202.077 1.092 1.030
Chain 1: 500 -8083.832 0.877 1.000
Chain 1: 600 -8526.244 0.739 1.000
Chain 1: 700 -7778.171 0.647 0.096
Chain 1: 800 -8079.531 0.571 0.096
Chain 1: 900 -8180.640 0.509 0.054
Chain 1: 1000 -7927.849 0.461 0.054
Chain 1: 1100 -7709.653 0.364 0.052
Chain 1: 1200 -7683.296 0.136 0.037
Chain 1: 1300 -7718.237 0.033 0.032
Chain 1: 1400 -7863.476 0.030 0.028
Chain 1: 1500 -7633.228 0.031 0.030
Chain 1: 1600 -7785.209 0.028 0.028
Chain 1: 1700 -7523.672 0.022 0.028
Chain 1: 1800 -7642.619 0.020 0.020
Chain 1: 1900 -7572.445 0.020 0.020
Chain 1: 2000 -7612.477 0.017 0.018
Chain 1: 2100 -7605.405 0.014 0.016
Chain 1: 2200 -7712.812 0.015 0.016
Chain 1: 2300 -7727.169 0.015 0.016
Chain 1: 2400 -7647.867 0.014 0.014
Chain 1: 2500 -7579.975 0.012 0.010
Chain 1: 2600 -7537.610 0.011 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003055 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.55 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86945.603 1.000 1.000
Chain 1: 200 -13429.243 3.237 5.474
Chain 1: 300 -9800.381 2.282 1.000
Chain 1: 400 -10673.298 1.732 1.000
Chain 1: 500 -8689.014 1.431 0.370
Chain 1: 600 -8296.050 1.200 0.370
Chain 1: 700 -8245.650 1.030 0.228
Chain 1: 800 -8886.216 0.910 0.228
Chain 1: 900 -8652.846 0.812 0.082
Chain 1: 1000 -8432.826 0.733 0.082
Chain 1: 1100 -8616.003 0.635 0.072 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8361.703 0.091 0.047
Chain 1: 1300 -8522.922 0.056 0.030
Chain 1: 1400 -8528.056 0.048 0.027
Chain 1: 1500 -8395.783 0.027 0.026
Chain 1: 1600 -8505.868 0.023 0.021
Chain 1: 1700 -8592.500 0.024 0.021
Chain 1: 1800 -8189.908 0.021 0.021
Chain 1: 1900 -8287.830 0.020 0.019
Chain 1: 2000 -8259.441 0.017 0.016
Chain 1: 2100 -8379.277 0.017 0.014
Chain 1: 2200 -8187.526 0.016 0.014
Chain 1: 2300 -8323.226 0.016 0.014
Chain 1: 2400 -8198.285 0.017 0.015
Chain 1: 2500 -8263.021 0.016 0.014
Chain 1: 2600 -8286.728 0.015 0.014
Chain 1: 2700 -8205.021 0.015 0.014
Chain 1: 2800 -8177.589 0.011 0.012
Chain 1: 2900 -8232.991 0.010 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.009248 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 92.48 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8397956.777 1.000 1.000
Chain 1: 200 -1584347.381 2.650 4.301
Chain 1: 300 -890735.317 2.026 1.000
Chain 1: 400 -457561.703 1.756 1.000
Chain 1: 500 -357881.253 1.461 0.947
Chain 1: 600 -232769.761 1.307 0.947
Chain 1: 700 -119055.064 1.257 0.947
Chain 1: 800 -86279.435 1.147 0.947
Chain 1: 900 -66636.420 1.052 0.779
Chain 1: 1000 -51442.010 0.977 0.779
Chain 1: 1100 -38930.873 0.909 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38106.519 0.481 0.380
Chain 1: 1300 -26080.142 0.449 0.380
Chain 1: 1400 -25799.519 0.356 0.321
Chain 1: 1500 -22391.193 0.343 0.321
Chain 1: 1600 -21608.662 0.293 0.295
Chain 1: 1700 -20484.830 0.203 0.295
Chain 1: 1800 -20429.363 0.165 0.152
Chain 1: 1900 -20755.295 0.137 0.055
Chain 1: 2000 -19268.028 0.115 0.055
Chain 1: 2100 -19506.393 0.084 0.036
Chain 1: 2200 -19732.462 0.083 0.036
Chain 1: 2300 -19350.015 0.039 0.020
Chain 1: 2400 -19122.209 0.039 0.020
Chain 1: 2500 -18924.112 0.025 0.016
Chain 1: 2600 -18554.701 0.024 0.016
Chain 1: 2700 -18511.757 0.018 0.012
Chain 1: 2800 -18228.707 0.020 0.016
Chain 1: 2900 -18509.790 0.020 0.015
Chain 1: 3000 -18496.030 0.012 0.012
Chain 1: 3100 -18580.991 0.011 0.012
Chain 1: 3200 -18271.857 0.012 0.015
Chain 1: 3300 -18476.419 0.011 0.012
Chain 1: 3400 -17951.660 0.013 0.015
Chain 1: 3500 -18563.043 0.015 0.016
Chain 1: 3600 -17870.322 0.017 0.016
Chain 1: 3700 -18256.691 0.019 0.017
Chain 1: 3800 -17217.320 0.023 0.021
Chain 1: 3900 -17213.456 0.022 0.021
Chain 1: 4000 -17330.780 0.022 0.021
Chain 1: 4100 -17244.602 0.022 0.021
Chain 1: 4200 -17061.022 0.022 0.021
Chain 1: 4300 -17199.312 0.021 0.021
Chain 1: 4400 -17156.312 0.019 0.011
Chain 1: 4500 -17058.835 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001322 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.22 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48795.831 1.000 1.000
Chain 1: 200 -19296.536 1.264 1.529
Chain 1: 300 -14024.995 0.968 1.000
Chain 1: 400 -17144.878 0.772 1.000
Chain 1: 500 -17598.043 0.622 0.376
Chain 1: 600 -13087.466 0.576 0.376
Chain 1: 700 -12162.233 0.505 0.345
Chain 1: 800 -13534.038 0.454 0.345
Chain 1: 900 -10879.751 0.431 0.244
Chain 1: 1000 -12497.328 0.401 0.244
Chain 1: 1100 -10299.145 0.322 0.213
Chain 1: 1200 -11284.213 0.178 0.182
Chain 1: 1300 -12670.417 0.151 0.129
Chain 1: 1400 -12261.932 0.136 0.109
Chain 1: 1500 -10200.081 0.154 0.129
Chain 1: 1600 -10221.798 0.120 0.109
Chain 1: 1700 -11291.488 0.122 0.109
Chain 1: 1800 -10098.361 0.123 0.118
Chain 1: 1900 -16848.454 0.139 0.118
Chain 1: 2000 -11184.991 0.177 0.118
Chain 1: 2100 -12703.705 0.167 0.118
Chain 1: 2200 -10435.239 0.180 0.120
Chain 1: 2300 -14572.295 0.198 0.202
Chain 1: 2400 -9161.734 0.254 0.217
Chain 1: 2500 -9326.990 0.235 0.217
Chain 1: 2600 -9237.822 0.236 0.217
Chain 1: 2700 -9949.273 0.234 0.217
Chain 1: 2800 -13276.580 0.247 0.251
Chain 1: 2900 -10990.953 0.228 0.217
Chain 1: 3000 -8792.519 0.202 0.217
Chain 1: 3100 -8600.829 0.192 0.217
Chain 1: 3200 -9311.726 0.178 0.208
Chain 1: 3300 -11925.864 0.172 0.208
Chain 1: 3400 -12855.876 0.120 0.076
Chain 1: 3500 -10416.641 0.141 0.208
Chain 1: 3600 -9872.797 0.146 0.208
Chain 1: 3700 -9026.881 0.148 0.208
Chain 1: 3800 -11366.806 0.144 0.206
Chain 1: 3900 -10597.018 0.130 0.094
Chain 1: 4000 -9703.780 0.114 0.092
Chain 1: 4100 -8780.684 0.123 0.094
Chain 1: 4200 -9167.524 0.119 0.094
Chain 1: 4300 -14620.959 0.135 0.094
Chain 1: 4400 -9020.380 0.189 0.105
Chain 1: 4500 -8923.258 0.167 0.094
Chain 1: 4600 -10864.684 0.180 0.105
Chain 1: 4700 -11678.216 0.177 0.105
Chain 1: 4800 -8447.088 0.195 0.105
Chain 1: 4900 -9381.192 0.197 0.105
Chain 1: 5000 -8629.629 0.197 0.105
Chain 1: 5100 -8575.002 0.187 0.100
Chain 1: 5200 -11762.666 0.210 0.179
Chain 1: 5300 -13381.993 0.185 0.121
Chain 1: 5400 -15390.041 0.136 0.121
Chain 1: 5500 -11070.768 0.174 0.130
Chain 1: 5600 -13171.829 0.172 0.130
Chain 1: 5700 -8454.202 0.221 0.160
Chain 1: 5800 -8952.733 0.188 0.130
Chain 1: 5900 -8210.457 0.187 0.130
Chain 1: 6000 -9365.975 0.191 0.130
Chain 1: 6100 -8671.911 0.198 0.130
Chain 1: 6200 -8101.961 0.178 0.123
Chain 1: 6300 -13203.999 0.204 0.130
Chain 1: 6400 -10924.583 0.212 0.160
Chain 1: 6500 -9426.335 0.189 0.159
Chain 1: 6600 -8170.929 0.189 0.154
Chain 1: 6700 -10907.480 0.158 0.154
Chain 1: 6800 -8284.380 0.184 0.159
Chain 1: 6900 -8059.667 0.178 0.159
Chain 1: 7000 -8209.206 0.167 0.159
Chain 1: 7100 -8906.812 0.167 0.159
Chain 1: 7200 -8149.678 0.169 0.159
Chain 1: 7300 -8940.206 0.139 0.154
Chain 1: 7400 -8223.610 0.127 0.093
Chain 1: 7500 -10135.224 0.130 0.093
Chain 1: 7600 -10946.977 0.122 0.088
Chain 1: 7700 -8523.942 0.126 0.088
Chain 1: 7800 -8268.035 0.097 0.087
Chain 1: 7900 -10183.703 0.113 0.088
Chain 1: 8000 -10765.085 0.117 0.088
Chain 1: 8100 -8232.267 0.140 0.093
Chain 1: 8200 -8718.972 0.136 0.088
Chain 1: 8300 -10906.849 0.147 0.188
Chain 1: 8400 -10147.735 0.146 0.188
Chain 1: 8500 -10512.121 0.131 0.075
Chain 1: 8600 -8108.845 0.153 0.188
Chain 1: 8700 -9381.478 0.138 0.136
Chain 1: 8800 -8380.885 0.147 0.136
Chain 1: 8900 -10278.202 0.146 0.136
Chain 1: 9000 -10173.291 0.142 0.136
Chain 1: 9100 -7960.581 0.139 0.136
Chain 1: 9200 -11045.324 0.161 0.185
Chain 1: 9300 -8146.592 0.177 0.185
Chain 1: 9400 -10838.120 0.194 0.248
Chain 1: 9500 -7897.290 0.228 0.278
Chain 1: 9600 -8356.689 0.204 0.248
Chain 1: 9700 -8251.259 0.192 0.248
Chain 1: 9800 -8354.092 0.181 0.248
Chain 1: 9900 -10865.982 0.186 0.248
Chain 1: 10000 -7897.735 0.222 0.278
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001427 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.27 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61289.259 1.000 1.000
Chain 1: 200 -17619.301 1.739 2.479
Chain 1: 300 -8748.144 1.498 1.014
Chain 1: 400 -8282.035 1.137 1.014
Chain 1: 500 -8391.918 0.912 1.000
Chain 1: 600 -8370.246 0.761 1.000
Chain 1: 700 -8183.211 0.655 0.056
Chain 1: 800 -8189.376 0.574 0.056
Chain 1: 900 -7585.904 0.519 0.056
Chain 1: 1000 -7622.936 0.467 0.056
Chain 1: 1100 -7690.342 0.368 0.023
Chain 1: 1200 -7711.019 0.121 0.013
Chain 1: 1300 -7618.373 0.020 0.012
Chain 1: 1400 -7644.697 0.015 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002892 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 28.92 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86280.776 1.000 1.000
Chain 1: 200 -13262.338 3.253 5.506
Chain 1: 300 -9670.716 2.292 1.000
Chain 1: 400 -10574.144 1.741 1.000
Chain 1: 500 -8600.162 1.438 0.371
Chain 1: 600 -8161.907 1.208 0.371
Chain 1: 700 -8209.754 1.036 0.230
Chain 1: 800 -8455.101 0.910 0.230
Chain 1: 900 -8521.022 0.810 0.085
Chain 1: 1000 -8194.521 0.733 0.085
Chain 1: 1100 -8480.667 0.636 0.054 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8239.088 0.089 0.040
Chain 1: 1300 -8381.257 0.053 0.034
Chain 1: 1400 -8376.176 0.045 0.029
Chain 1: 1500 -8254.071 0.023 0.029
Chain 1: 1600 -8363.685 0.019 0.017
Chain 1: 1700 -8449.018 0.020 0.017
Chain 1: 1800 -8048.656 0.022 0.017
Chain 1: 1900 -8147.891 0.022 0.017
Chain 1: 2000 -8119.114 0.018 0.015
Chain 1: 2100 -8239.026 0.016 0.015
Chain 1: 2200 -8030.052 0.016 0.015
Chain 1: 2300 -8179.911 0.016 0.015
Chain 1: 2400 -8060.288 0.018 0.015
Chain 1: 2500 -8123.392 0.017 0.015
Chain 1: 2600 -8145.063 0.016 0.015
Chain 1: 2700 -8064.036 0.016 0.015
Chain 1: 2800 -8037.813 0.011 0.012
Chain 1: 2900 -8093.203 0.011 0.010
Chain 1: 3000 -7977.260 0.012 0.015
Chain 1: 3100 -8115.239 0.012 0.015
Chain 1: 3200 -7995.079 0.011 0.015
Chain 1: 3300 -8016.697 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003374 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.74 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8414347.301 1.000 1.000
Chain 1: 200 -1587258.709 2.651 4.301
Chain 1: 300 -891505.203 2.027 1.000
Chain 1: 400 -457888.579 1.757 1.000
Chain 1: 500 -358147.567 1.461 0.947
Chain 1: 600 -232905.977 1.307 0.947
Chain 1: 700 -119006.716 1.257 0.947
Chain 1: 800 -86203.489 1.148 0.947
Chain 1: 900 -66528.150 1.053 0.780
Chain 1: 1000 -51320.502 0.977 0.780
Chain 1: 1100 -38797.730 0.910 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37969.095 0.482 0.381
Chain 1: 1300 -25932.137 0.450 0.381
Chain 1: 1400 -25649.595 0.357 0.323
Chain 1: 1500 -22239.393 0.344 0.323
Chain 1: 1600 -21455.973 0.294 0.296
Chain 1: 1700 -20331.051 0.204 0.296
Chain 1: 1800 -20275.178 0.166 0.153
Chain 1: 1900 -20601.029 0.138 0.055
Chain 1: 2000 -19113.373 0.116 0.055
Chain 1: 2100 -19351.617 0.085 0.037
Chain 1: 2200 -19577.844 0.084 0.037
Chain 1: 2300 -19195.309 0.040 0.020
Chain 1: 2400 -18967.529 0.040 0.020
Chain 1: 2500 -18769.507 0.025 0.016
Chain 1: 2600 -18400.033 0.024 0.016
Chain 1: 2700 -18357.053 0.019 0.012
Chain 1: 2800 -18074.088 0.020 0.016
Chain 1: 2900 -18355.140 0.020 0.015
Chain 1: 3000 -18341.364 0.012 0.012
Chain 1: 3100 -18426.328 0.011 0.012
Chain 1: 3200 -18117.187 0.012 0.015
Chain 1: 3300 -18321.732 0.011 0.012
Chain 1: 3400 -17797.013 0.013 0.015
Chain 1: 3500 -18408.377 0.015 0.016
Chain 1: 3600 -17715.694 0.017 0.016
Chain 1: 3700 -18102.032 0.019 0.017
Chain 1: 3800 -17062.757 0.023 0.021
Chain 1: 3900 -17058.913 0.022 0.021
Chain 1: 4000 -17176.217 0.022 0.021
Chain 1: 4100 -17090.074 0.022 0.021
Chain 1: 4200 -16906.498 0.022 0.021
Chain 1: 4300 -17044.760 0.022 0.021
Chain 1: 4400 -17001.759 0.019 0.011
Chain 1: 4500 -16904.330 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001286 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.86 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48902.724 1.000 1.000
Chain 1: 200 -17510.200 1.396 1.793
Chain 1: 300 -21253.549 0.990 1.000
Chain 1: 400 -11959.030 0.937 1.000
Chain 1: 500 -14312.649 0.782 0.777
Chain 1: 600 -25880.686 0.726 0.777
Chain 1: 700 -15409.133 0.720 0.680
Chain 1: 800 -18390.487 0.650 0.680
Chain 1: 900 -14586.532 0.607 0.447
Chain 1: 1000 -13531.058 0.554 0.447
Chain 1: 1100 -18665.066 0.481 0.275
Chain 1: 1200 -12728.115 0.349 0.275
Chain 1: 1300 -12861.148 0.332 0.275
Chain 1: 1400 -9767.125 0.286 0.275
Chain 1: 1500 -26790.663 0.333 0.317
Chain 1: 1600 -10381.718 0.447 0.317
Chain 1: 1700 -9367.457 0.389 0.275
Chain 1: 1800 -13818.330 0.405 0.317
Chain 1: 1900 -10169.502 0.415 0.322
Chain 1: 2000 -11047.796 0.415 0.322
Chain 1: 2100 -11974.781 0.396 0.322
Chain 1: 2200 -12261.563 0.351 0.317
Chain 1: 2300 -11641.232 0.356 0.317
Chain 1: 2400 -9060.884 0.352 0.285
Chain 1: 2500 -10224.505 0.300 0.114
Chain 1: 2600 -9418.921 0.151 0.108
Chain 1: 2700 -9327.576 0.141 0.086
Chain 1: 2800 -9080.323 0.111 0.079
Chain 1: 2900 -15814.075 0.118 0.079
Chain 1: 3000 -9942.579 0.169 0.086
Chain 1: 3100 -10329.027 0.165 0.086
Chain 1: 3200 -17026.986 0.202 0.114
Chain 1: 3300 -12445.861 0.234 0.285
Chain 1: 3400 -10683.577 0.222 0.165
Chain 1: 3500 -11973.466 0.221 0.165
Chain 1: 3600 -10436.801 0.227 0.165
Chain 1: 3700 -11801.351 0.238 0.165
Chain 1: 3800 -9116.518 0.265 0.295
Chain 1: 3900 -8972.537 0.224 0.165
Chain 1: 4000 -10328.991 0.178 0.147
Chain 1: 4100 -9545.948 0.182 0.147
Chain 1: 4200 -11161.439 0.157 0.145
Chain 1: 4300 -12633.656 0.132 0.131
Chain 1: 4400 -9510.831 0.148 0.131
Chain 1: 4500 -8645.090 0.148 0.131
Chain 1: 4600 -9723.362 0.144 0.117
Chain 1: 4700 -10488.642 0.140 0.117
Chain 1: 4800 -8979.643 0.127 0.117
Chain 1: 4900 -11345.794 0.146 0.131
Chain 1: 5000 -11970.053 0.138 0.117
Chain 1: 5100 -8475.609 0.171 0.145
Chain 1: 5200 -9315.768 0.166 0.117
Chain 1: 5300 -10505.655 0.166 0.113
Chain 1: 5400 -14524.160 0.161 0.113
Chain 1: 5500 -13734.900 0.156 0.113
Chain 1: 5600 -11374.909 0.166 0.168
Chain 1: 5700 -11829.339 0.162 0.168
Chain 1: 5800 -8852.890 0.179 0.207
Chain 1: 5900 -11225.624 0.180 0.207
Chain 1: 6000 -10477.415 0.181 0.207
Chain 1: 6100 -14388.258 0.167 0.207
Chain 1: 6200 -8641.054 0.225 0.211
Chain 1: 6300 -8653.775 0.214 0.211
Chain 1: 6400 -11159.465 0.209 0.211
Chain 1: 6500 -11595.224 0.207 0.211
Chain 1: 6600 -8862.543 0.217 0.225
Chain 1: 6700 -9551.361 0.220 0.225
Chain 1: 6800 -10207.161 0.193 0.211
Chain 1: 6900 -12614.751 0.191 0.191
Chain 1: 7000 -9070.725 0.223 0.225
Chain 1: 7100 -12678.828 0.224 0.225
Chain 1: 7200 -8281.853 0.211 0.225
Chain 1: 7300 -8486.842 0.213 0.225
Chain 1: 7400 -8346.246 0.192 0.191
Chain 1: 7500 -10451.171 0.208 0.201
Chain 1: 7600 -8449.432 0.201 0.201
Chain 1: 7700 -8322.453 0.196 0.201
Chain 1: 7800 -8308.990 0.189 0.201
Chain 1: 7900 -9830.821 0.186 0.201
Chain 1: 8000 -10077.855 0.149 0.155
Chain 1: 8100 -9765.355 0.124 0.032
Chain 1: 8200 -9709.765 0.071 0.025
Chain 1: 8300 -8586.701 0.082 0.032
Chain 1: 8400 -8552.936 0.081 0.032
Chain 1: 8500 -8988.632 0.065 0.032
Chain 1: 8600 -10050.336 0.052 0.032
Chain 1: 8700 -8272.839 0.072 0.048
Chain 1: 8800 -8337.840 0.073 0.048
Chain 1: 8900 -12648.521 0.091 0.048
Chain 1: 9000 -8513.281 0.138 0.106
Chain 1: 9100 -8745.235 0.137 0.106
Chain 1: 9200 -8911.066 0.138 0.106
Chain 1: 9300 -8198.431 0.134 0.087
Chain 1: 9400 -8170.227 0.134 0.087
Chain 1: 9500 -8078.181 0.130 0.087
Chain 1: 9600 -8424.887 0.124 0.041
Chain 1: 9700 -10767.721 0.124 0.041
Chain 1: 9800 -8479.201 0.150 0.087
Chain 1: 9900 -9279.382 0.125 0.086
Chain 1: 10000 -8221.882 0.089 0.086
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001383 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.83 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61457.720 1.000 1.000
Chain 1: 200 -17758.657 1.730 2.461
Chain 1: 300 -8792.022 1.494 1.020
Chain 1: 400 -9165.988 1.130 1.020
Chain 1: 500 -7960.506 0.935 1.000
Chain 1: 600 -8913.566 0.797 1.000
Chain 1: 700 -7712.249 0.705 0.156
Chain 1: 800 -7998.812 0.621 0.156
Chain 1: 900 -7676.064 0.557 0.151
Chain 1: 1000 -7702.639 0.502 0.151
Chain 1: 1100 -7758.292 0.402 0.107
Chain 1: 1200 -7604.931 0.158 0.042
Chain 1: 1300 -7691.130 0.057 0.041
Chain 1: 1400 -7797.080 0.055 0.036
Chain 1: 1500 -7561.058 0.043 0.031
Chain 1: 1600 -7738.088 0.034 0.023
Chain 1: 1700 -7481.204 0.022 0.023
Chain 1: 1800 -7537.379 0.019 0.020
Chain 1: 1900 -7552.563 0.015 0.014
Chain 1: 2000 -7581.111 0.015 0.014
Chain 1: 2100 -7570.987 0.015 0.014
Chain 1: 2200 -7652.296 0.014 0.011
Chain 1: 2300 -7535.015 0.014 0.014
Chain 1: 2400 -7599.504 0.014 0.011
Chain 1: 2500 -7535.759 0.011 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002984 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.84 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86694.306 1.000 1.000
Chain 1: 200 -13441.166 3.225 5.450
Chain 1: 300 -9870.975 2.271 1.000
Chain 1: 400 -10661.482 1.721 1.000
Chain 1: 500 -8810.258 1.419 0.362
Chain 1: 600 -8576.770 1.187 0.362
Chain 1: 700 -8493.231 1.019 0.210
Chain 1: 800 -8890.129 0.897 0.210
Chain 1: 900 -8690.663 0.800 0.074
Chain 1: 1000 -8525.837 0.722 0.074
Chain 1: 1100 -8745.207 0.624 0.045 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8576.017 0.081 0.027
Chain 1: 1300 -8489.151 0.046 0.025
Chain 1: 1400 -8624.456 0.040 0.023
Chain 1: 1500 -8491.198 0.021 0.020
Chain 1: 1600 -8598.617 0.020 0.019
Chain 1: 1700 -8685.849 0.020 0.019
Chain 1: 1800 -8297.376 0.020 0.019
Chain 1: 1900 -8399.695 0.019 0.016
Chain 1: 2000 -8369.739 0.017 0.016
Chain 1: 2100 -8499.921 0.016 0.015
Chain 1: 2200 -8286.267 0.017 0.015
Chain 1: 2300 -8428.749 0.017 0.016
Chain 1: 2400 -8441.936 0.016 0.015
Chain 1: 2500 -8409.640 0.015 0.012
Chain 1: 2600 -8410.206 0.014 0.012
Chain 1: 2700 -8317.966 0.014 0.012
Chain 1: 2800 -8293.172 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002898 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 28.98 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8422158.361 1.000 1.000
Chain 1: 200 -1588322.462 2.651 4.303
Chain 1: 300 -892355.081 2.027 1.000
Chain 1: 400 -458491.873 1.757 1.000
Chain 1: 500 -358463.371 1.462 0.946
Chain 1: 600 -233110.201 1.308 0.946
Chain 1: 700 -119184.088 1.257 0.946
Chain 1: 800 -86373.495 1.148 0.946
Chain 1: 900 -66694.880 1.053 0.780
Chain 1: 1000 -51480.595 0.977 0.780
Chain 1: 1100 -38957.450 0.909 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38126.448 0.481 0.380
Chain 1: 1300 -26093.025 0.449 0.380
Chain 1: 1400 -25810.204 0.356 0.321
Chain 1: 1500 -22401.016 0.343 0.321
Chain 1: 1600 -21617.653 0.293 0.296
Chain 1: 1700 -20493.282 0.203 0.295
Chain 1: 1800 -20437.375 0.165 0.152
Chain 1: 1900 -20763.043 0.137 0.055
Chain 1: 2000 -19275.971 0.115 0.055
Chain 1: 2100 -19514.220 0.084 0.036
Chain 1: 2200 -19740.295 0.083 0.036
Chain 1: 2300 -19357.928 0.039 0.020
Chain 1: 2400 -19130.211 0.039 0.020
Chain 1: 2500 -18932.215 0.025 0.016
Chain 1: 2600 -18563.000 0.024 0.016
Chain 1: 2700 -18520.031 0.018 0.012
Chain 1: 2800 -18237.188 0.020 0.016
Chain 1: 2900 -18518.108 0.020 0.015
Chain 1: 3000 -18504.349 0.012 0.012
Chain 1: 3100 -18589.319 0.011 0.012
Chain 1: 3200 -18280.313 0.012 0.015
Chain 1: 3300 -18484.717 0.011 0.012
Chain 1: 3400 -17960.292 0.013 0.015
Chain 1: 3500 -18571.260 0.015 0.016
Chain 1: 3600 -17879.024 0.017 0.016
Chain 1: 3700 -18265.064 0.019 0.017
Chain 1: 3800 -17226.543 0.023 0.021
Chain 1: 3900 -17222.696 0.022 0.021
Chain 1: 4000 -17339.996 0.022 0.021
Chain 1: 4100 -17253.940 0.022 0.021
Chain 1: 4200 -17070.467 0.022 0.021
Chain 1: 4300 -17208.644 0.021 0.021
Chain 1: 4400 -17165.788 0.019 0.011
Chain 1: 4500 -17068.355 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001276 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.76 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12672.719 1.000 1.000
Chain 1: 200 -9591.844 0.661 1.000
Chain 1: 300 -8180.107 0.498 0.321
Chain 1: 400 -8402.759 0.380 0.321
Chain 1: 500 -8294.076 0.307 0.173
Chain 1: 600 -8101.755 0.260 0.173
Chain 1: 700 -7971.085 0.225 0.026
Chain 1: 800 -7962.247 0.197 0.026
Chain 1: 900 -7947.876 0.175 0.024
Chain 1: 1000 -8122.927 0.160 0.024
Chain 1: 1100 -8133.090 0.060 0.022
Chain 1: 1200 -8036.129 0.029 0.016
Chain 1: 1300 -7980.570 0.012 0.013
Chain 1: 1400 -7975.661 0.010 0.012 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001388 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.88 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49634.081 1.000 1.000
Chain 1: 200 -16477.691 1.506 2.012
Chain 1: 300 -8926.294 1.286 1.000
Chain 1: 400 -8516.762 0.977 1.000
Chain 1: 500 -8189.074 0.789 0.846
Chain 1: 600 -9365.161 0.679 0.846
Chain 1: 700 -7815.070 0.610 0.198
Chain 1: 800 -7709.410 0.535 0.198
Chain 1: 900 -8120.567 0.482 0.126
Chain 1: 1000 -7779.152 0.438 0.126
Chain 1: 1100 -7719.797 0.339 0.051
Chain 1: 1200 -7671.928 0.138 0.048
Chain 1: 1300 -7779.743 0.055 0.044
Chain 1: 1400 -7616.751 0.052 0.040
Chain 1: 1500 -7531.761 0.049 0.021
Chain 1: 1600 -7677.515 0.039 0.019
Chain 1: 1700 -7539.405 0.021 0.018
Chain 1: 1800 -7620.592 0.020 0.018
Chain 1: 1900 -7720.050 0.017 0.014
Chain 1: 2000 -7645.598 0.013 0.013
Chain 1: 2100 -7586.370 0.013 0.013
Chain 1: 2200 -7826.574 0.016 0.014
Chain 1: 2300 -7558.843 0.018 0.018
Chain 1: 2400 -7613.754 0.016 0.013
Chain 1: 2500 -7627.768 0.015 0.013
Chain 1: 2600 -7519.991 0.015 0.013
Chain 1: 2700 -7482.519 0.014 0.011
Chain 1: 2800 -7514.363 0.013 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002917 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.17 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86520.903 1.000 1.000
Chain 1: 200 -13920.119 3.108 5.216
Chain 1: 300 -10151.959 2.196 1.000
Chain 1: 400 -11920.310 1.684 1.000
Chain 1: 500 -8695.397 1.421 0.371
Chain 1: 600 -8470.919 1.189 0.371
Chain 1: 700 -8541.854 1.020 0.371
Chain 1: 800 -8709.033 0.895 0.371
Chain 1: 900 -8844.064 0.797 0.148
Chain 1: 1000 -9099.525 0.720 0.148
Chain 1: 1100 -8794.631 0.624 0.035 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8459.972 0.106 0.035
Chain 1: 1300 -8773.040 0.073 0.035
Chain 1: 1400 -8531.198 0.061 0.028
Chain 1: 1500 -8626.857 0.025 0.028
Chain 1: 1600 -8734.003 0.023 0.028
Chain 1: 1700 -8784.918 0.023 0.028
Chain 1: 1800 -8331.324 0.027 0.028
Chain 1: 1900 -8441.682 0.026 0.028
Chain 1: 2000 -8442.758 0.024 0.028
Chain 1: 2100 -8579.775 0.022 0.016
Chain 1: 2200 -8340.053 0.021 0.016
Chain 1: 2300 -8490.599 0.019 0.016
Chain 1: 2400 -8341.994 0.018 0.016
Chain 1: 2500 -8415.109 0.017 0.016
Chain 1: 2600 -8326.873 0.017 0.016
Chain 1: 2700 -8358.909 0.017 0.016
Chain 1: 2800 -8311.050 0.012 0.013
Chain 1: 2900 -8422.246 0.012 0.013
Chain 1: 3000 -8360.600 0.013 0.013
Chain 1: 3100 -8303.033 0.012 0.011
Chain 1: 3200 -8276.101 0.010 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003303 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.03 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8396816.202 1.000 1.000
Chain 1: 200 -1581612.257 2.655 4.309
Chain 1: 300 -891573.845 2.028 1.000
Chain 1: 400 -458427.415 1.757 1.000
Chain 1: 500 -359211.306 1.461 0.945
Chain 1: 600 -234098.136 1.306 0.945
Chain 1: 700 -120020.461 1.256 0.945
Chain 1: 800 -87152.935 1.146 0.945
Chain 1: 900 -67427.558 1.051 0.774
Chain 1: 1000 -52182.038 0.975 0.774
Chain 1: 1100 -39612.676 0.907 0.534 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38789.344 0.478 0.377
Chain 1: 1300 -26687.344 0.446 0.377
Chain 1: 1400 -26404.031 0.353 0.317
Chain 1: 1500 -22975.871 0.340 0.317
Chain 1: 1600 -22188.682 0.290 0.293
Chain 1: 1700 -21054.801 0.200 0.292
Chain 1: 1800 -20997.613 0.163 0.149
Chain 1: 1900 -21324.439 0.135 0.054
Chain 1: 2000 -19830.361 0.113 0.054
Chain 1: 2100 -20069.023 0.083 0.035
Chain 1: 2200 -20296.604 0.082 0.035
Chain 1: 2300 -19912.671 0.039 0.019
Chain 1: 2400 -19684.427 0.039 0.019
Chain 1: 2500 -19486.646 0.025 0.015
Chain 1: 2600 -19115.897 0.023 0.015
Chain 1: 2700 -19072.594 0.018 0.012
Chain 1: 2800 -18789.209 0.019 0.015
Chain 1: 2900 -19070.901 0.019 0.015
Chain 1: 3000 -19056.969 0.012 0.012
Chain 1: 3100 -19142.060 0.011 0.012
Chain 1: 3200 -18832.217 0.011 0.015
Chain 1: 3300 -19037.363 0.011 0.012
Chain 1: 3400 -18511.410 0.012 0.015
Chain 1: 3500 -19124.652 0.014 0.015
Chain 1: 3600 -18429.605 0.016 0.015
Chain 1: 3700 -18817.732 0.018 0.016
Chain 1: 3800 -17774.748 0.022 0.021
Chain 1: 3900 -17770.850 0.021 0.021
Chain 1: 4000 -17888.137 0.022 0.021
Chain 1: 4100 -17801.763 0.022 0.021
Chain 1: 4200 -17617.417 0.021 0.021
Chain 1: 4300 -17756.202 0.021 0.021
Chain 1: 4400 -17712.547 0.018 0.010
Chain 1: 4500 -17615.019 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001304 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.04 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12362.416 1.000 1.000
Chain 1: 200 -9328.688 0.663 1.000
Chain 1: 300 -8099.042 0.492 0.325
Chain 1: 400 -8221.769 0.373 0.325
Chain 1: 500 -8127.290 0.301 0.152
Chain 1: 600 -8029.370 0.253 0.152
Chain 1: 700 -7940.900 0.218 0.015
Chain 1: 800 -7950.472 0.191 0.015
Chain 1: 900 -7842.112 0.171 0.014
Chain 1: 1000 -8051.458 0.157 0.015
Chain 1: 1100 -8087.696 0.057 0.014
Chain 1: 1200 -7979.563 0.026 0.014
Chain 1: 1300 -7916.622 0.012 0.012
Chain 1: 1400 -7933.809 0.010 0.012
Chain 1: 1500 -8022.157 0.010 0.011
Chain 1: 1600 -7985.558 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001391 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.91 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61938.380 1.000 1.000
Chain 1: 200 -17928.869 1.727 2.455
Chain 1: 300 -8900.405 1.490 1.014
Chain 1: 400 -9515.728 1.133 1.014
Chain 1: 500 -8511.489 0.930 1.000
Chain 1: 600 -8527.564 0.776 1.000
Chain 1: 700 -7902.881 0.676 0.118
Chain 1: 800 -8072.298 0.594 0.118
Chain 1: 900 -8093.390 0.528 0.079
Chain 1: 1000 -7856.892 0.479 0.079
Chain 1: 1100 -7833.096 0.379 0.065
Chain 1: 1200 -7596.916 0.137 0.031
Chain 1: 1300 -7763.518 0.037 0.030
Chain 1: 1400 -7940.939 0.033 0.022
Chain 1: 1500 -7625.665 0.025 0.022
Chain 1: 1600 -7794.994 0.027 0.022
Chain 1: 1700 -7536.966 0.023 0.022
Chain 1: 1800 -7669.783 0.023 0.022
Chain 1: 1900 -7677.050 0.022 0.022
Chain 1: 2000 -7681.673 0.019 0.022
Chain 1: 2100 -7637.223 0.020 0.022
Chain 1: 2200 -7735.552 0.018 0.021
Chain 1: 2300 -7630.145 0.017 0.017
Chain 1: 2400 -7680.188 0.016 0.014
Chain 1: 2500 -7599.220 0.012 0.013
Chain 1: 2600 -7590.211 0.010 0.011
Chain 1: 2700 -7577.316 0.007 0.007 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003132 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.32 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85726.829 1.000 1.000
Chain 1: 200 -13518.172 3.171 5.342
Chain 1: 300 -9908.385 2.235 1.000
Chain 1: 400 -10671.582 1.694 1.000
Chain 1: 500 -8869.970 1.396 0.364
Chain 1: 600 -8380.463 1.173 0.364
Chain 1: 700 -8727.306 1.011 0.203
Chain 1: 800 -8850.116 0.887 0.203
Chain 1: 900 -8765.444 0.789 0.072
Chain 1: 1000 -8436.407 0.714 0.072
Chain 1: 1100 -8788.967 0.618 0.058 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8406.801 0.089 0.045
Chain 1: 1300 -8606.833 0.054 0.040
Chain 1: 1400 -8612.421 0.047 0.040
Chain 1: 1500 -8474.858 0.029 0.039
Chain 1: 1600 -8587.470 0.024 0.023
Chain 1: 1700 -8673.171 0.021 0.016
Chain 1: 1800 -8264.936 0.025 0.023
Chain 1: 1900 -8360.793 0.025 0.023
Chain 1: 2000 -8333.353 0.021 0.016
Chain 1: 2100 -8454.549 0.019 0.014
Chain 1: 2200 -8295.968 0.016 0.014
Chain 1: 2300 -8360.419 0.015 0.013
Chain 1: 2400 -8425.020 0.015 0.013
Chain 1: 2500 -8370.959 0.014 0.011
Chain 1: 2600 -8369.574 0.013 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003476 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.76 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8393494.648 1.000 1.000
Chain 1: 200 -1582436.747 2.652 4.304
Chain 1: 300 -890938.561 2.027 1.000
Chain 1: 400 -457902.557 1.756 1.000
Chain 1: 500 -358406.754 1.461 0.946
Chain 1: 600 -233351.432 1.307 0.946
Chain 1: 700 -119430.018 1.256 0.946
Chain 1: 800 -86594.078 1.147 0.946
Chain 1: 900 -66900.611 1.052 0.776
Chain 1: 1000 -51669.876 0.976 0.776
Chain 1: 1100 -39118.223 0.908 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38290.385 0.480 0.379
Chain 1: 1300 -26223.090 0.448 0.379
Chain 1: 1400 -25938.617 0.355 0.321
Chain 1: 1500 -22520.618 0.342 0.321
Chain 1: 1600 -21734.922 0.292 0.295
Chain 1: 1700 -20606.358 0.202 0.294
Chain 1: 1800 -20549.924 0.165 0.152
Chain 1: 1900 -20875.896 0.137 0.055
Chain 1: 2000 -19386.095 0.115 0.055
Chain 1: 2100 -19624.391 0.084 0.036
Chain 1: 2200 -19851.034 0.083 0.036
Chain 1: 2300 -19468.189 0.039 0.020
Chain 1: 2400 -19240.343 0.039 0.020
Chain 1: 2500 -19042.413 0.025 0.016
Chain 1: 2600 -18672.629 0.024 0.016
Chain 1: 2700 -18629.592 0.018 0.012
Chain 1: 2800 -18346.532 0.020 0.015
Chain 1: 2900 -18627.803 0.019 0.015
Chain 1: 3000 -18613.948 0.012 0.012
Chain 1: 3100 -18698.917 0.011 0.012
Chain 1: 3200 -18389.629 0.012 0.015
Chain 1: 3300 -18594.334 0.011 0.012
Chain 1: 3400 -18069.364 0.013 0.015
Chain 1: 3500 -18681.038 0.015 0.015
Chain 1: 3600 -17988.060 0.017 0.015
Chain 1: 3700 -18374.659 0.019 0.017
Chain 1: 3800 -17334.785 0.023 0.021
Chain 1: 3900 -17330.973 0.021 0.021
Chain 1: 4000 -17448.278 0.022 0.021
Chain 1: 4100 -17362.047 0.022 0.021
Chain 1: 4200 -17178.405 0.021 0.021
Chain 1: 4300 -17316.704 0.021 0.021
Chain 1: 4400 -17273.610 0.019 0.011
Chain 1: 4500 -17176.203 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001372 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.72 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12712.929 1.000 1.000
Chain 1: 200 -9504.810 0.669 1.000
Chain 1: 300 -8070.785 0.505 0.338
Chain 1: 400 -8311.550 0.386 0.338
Chain 1: 500 -8144.886 0.313 0.178
Chain 1: 600 -8011.530 0.264 0.178
Chain 1: 700 -8137.723 0.228 0.029
Chain 1: 800 -7915.438 0.203 0.029
Chain 1: 900 -7793.164 0.182 0.028
Chain 1: 1000 -7970.484 0.166 0.028
Chain 1: 1100 -7999.060 0.067 0.022
Chain 1: 1200 -7911.357 0.034 0.020
Chain 1: 1300 -7876.578 0.017 0.017
Chain 1: 1400 -7885.093 0.014 0.016
Chain 1: 1500 -7973.952 0.013 0.016
Chain 1: 1600 -7888.262 0.012 0.011
Chain 1: 1700 -7857.655 0.011 0.011
Chain 1: 1800 -7829.738 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001396 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.96 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58507.886 1.000 1.000
Chain 1: 200 -18071.856 1.619 2.238
Chain 1: 300 -8884.148 1.424 1.034
Chain 1: 400 -8120.499 1.091 1.034
Chain 1: 500 -9083.548 0.894 1.000
Chain 1: 600 -8515.844 0.756 1.000
Chain 1: 700 -8187.910 0.654 0.106
Chain 1: 800 -8209.969 0.573 0.106
Chain 1: 900 -7792.765 0.515 0.094
Chain 1: 1000 -7847.350 0.464 0.094
Chain 1: 1100 -7781.479 0.365 0.067
Chain 1: 1200 -7732.079 0.142 0.054
Chain 1: 1300 -7699.562 0.039 0.040
Chain 1: 1400 -7896.454 0.032 0.025
Chain 1: 1500 -7577.670 0.026 0.025
Chain 1: 1600 -7811.463 0.022 0.025
Chain 1: 1700 -7614.372 0.021 0.025
Chain 1: 1800 -7686.245 0.021 0.025
Chain 1: 1900 -7763.379 0.017 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003116 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.16 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87771.322 1.000 1.000
Chain 1: 200 -13866.639 3.165 5.330
Chain 1: 300 -10076.886 2.235 1.000
Chain 1: 400 -11688.728 1.711 1.000
Chain 1: 500 -8685.988 1.438 0.376
Chain 1: 600 -8525.371 1.201 0.376
Chain 1: 700 -8392.648 1.032 0.346
Chain 1: 800 -8924.254 0.910 0.346
Chain 1: 900 -8804.664 0.811 0.138
Chain 1: 1000 -8951.226 0.731 0.138
Chain 1: 1100 -8850.029 0.632 0.060 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8352.808 0.105 0.060
Chain 1: 1300 -8683.394 0.072 0.038
Chain 1: 1400 -8482.908 0.060 0.024
Chain 1: 1500 -8543.215 0.026 0.019
Chain 1: 1600 -8646.663 0.026 0.016
Chain 1: 1700 -8695.514 0.025 0.016
Chain 1: 1800 -8238.962 0.024 0.016
Chain 1: 1900 -8349.570 0.024 0.016
Chain 1: 2000 -8361.715 0.023 0.013
Chain 1: 2100 -8293.210 0.022 0.013
Chain 1: 2200 -8247.954 0.017 0.012
Chain 1: 2300 -8416.646 0.015 0.012
Chain 1: 2400 -8245.759 0.015 0.012
Chain 1: 2500 -8317.077 0.015 0.012
Chain 1: 2600 -8235.020 0.015 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003485 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.85 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8462142.464 1.000 1.000
Chain 1: 200 -1591072.540 2.659 4.319
Chain 1: 300 -890199.786 2.035 1.000
Chain 1: 400 -456919.872 1.764 1.000
Chain 1: 500 -356805.705 1.467 0.948
Chain 1: 600 -232017.664 1.312 0.948
Chain 1: 700 -118939.229 1.260 0.948
Chain 1: 800 -86319.231 1.150 0.948
Chain 1: 900 -66804.139 1.055 0.787
Chain 1: 1000 -51717.499 0.978 0.787
Chain 1: 1100 -39297.656 0.910 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38494.930 0.480 0.378
Chain 1: 1300 -26538.512 0.447 0.378
Chain 1: 1400 -26270.293 0.353 0.316
Chain 1: 1500 -22879.245 0.340 0.316
Chain 1: 1600 -22103.887 0.289 0.292
Chain 1: 1700 -20986.783 0.200 0.292
Chain 1: 1800 -20933.764 0.162 0.148
Chain 1: 1900 -21260.766 0.134 0.053
Chain 1: 2000 -19775.453 0.113 0.053
Chain 1: 2100 -20013.705 0.082 0.035
Chain 1: 2200 -20239.852 0.081 0.035
Chain 1: 2300 -19857.125 0.038 0.019
Chain 1: 2400 -19629.024 0.038 0.019
Chain 1: 2500 -19430.638 0.025 0.015
Chain 1: 2600 -19060.388 0.023 0.015
Chain 1: 2700 -19017.326 0.018 0.012
Chain 1: 2800 -18733.622 0.019 0.015
Chain 1: 2900 -19015.132 0.019 0.015
Chain 1: 3000 -19001.366 0.012 0.012
Chain 1: 3100 -19086.423 0.011 0.012
Chain 1: 3200 -18776.681 0.011 0.015
Chain 1: 3300 -18981.789 0.011 0.012
Chain 1: 3400 -18455.761 0.012 0.015
Chain 1: 3500 -19068.867 0.014 0.015
Chain 1: 3600 -18373.922 0.016 0.015
Chain 1: 3700 -18761.820 0.018 0.016
Chain 1: 3800 -17718.891 0.023 0.021
Chain 1: 3900 -17714.902 0.021 0.021
Chain 1: 4000 -17832.282 0.022 0.021
Chain 1: 4100 -17745.840 0.022 0.021
Chain 1: 4200 -17561.556 0.021 0.021
Chain 1: 4300 -17700.390 0.021 0.021
Chain 1: 4400 -17656.750 0.018 0.010
Chain 1: 4500 -17559.131 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001462 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.62 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12037.893 1.000 1.000
Chain 1: 200 -8952.358 0.672 1.000
Chain 1: 300 -7842.018 0.495 0.345
Chain 1: 400 -7901.951 0.373 0.345
Chain 1: 500 -7831.756 0.301 0.142
Chain 1: 600 -7761.004 0.252 0.142
Chain 1: 700 -7683.505 0.217 0.010
Chain 1: 800 -7907.409 0.194 0.028
Chain 1: 900 -7673.923 0.176 0.028
Chain 1: 1000 -7701.586 0.158 0.028
Chain 1: 1100 -7779.291 0.059 0.010
Chain 1: 1200 -7708.262 0.026 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001384 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.84 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56384.738 1.000 1.000
Chain 1: 200 -17139.041 1.645 2.290
Chain 1: 300 -8497.354 1.436 1.017
Chain 1: 400 -7878.434 1.096 1.017
Chain 1: 500 -8356.261 0.889 1.000
Chain 1: 600 -8034.683 0.747 1.000
Chain 1: 700 -7734.626 0.646 0.079
Chain 1: 800 -7958.815 0.569 0.079
Chain 1: 900 -7919.032 0.506 0.057
Chain 1: 1000 -7707.866 0.458 0.057
Chain 1: 1100 -7637.141 0.359 0.040
Chain 1: 1200 -7577.788 0.131 0.039
Chain 1: 1300 -7688.405 0.031 0.028
Chain 1: 1400 -7650.728 0.023 0.027
Chain 1: 1500 -7579.443 0.019 0.014
Chain 1: 1600 -7542.378 0.015 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00327 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.7 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86617.797 1.000 1.000
Chain 1: 200 -13150.899 3.293 5.586
Chain 1: 300 -9580.955 2.320 1.000
Chain 1: 400 -10404.651 1.760 1.000
Chain 1: 500 -8502.100 1.452 0.373
Chain 1: 600 -8105.700 1.218 0.373
Chain 1: 700 -8266.710 1.047 0.224
Chain 1: 800 -8847.114 0.924 0.224
Chain 1: 900 -8418.612 0.827 0.079
Chain 1: 1000 -8174.914 0.748 0.079
Chain 1: 1100 -8331.578 0.650 0.066 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8106.426 0.094 0.051
Chain 1: 1300 -8309.294 0.059 0.049
Chain 1: 1400 -8293.134 0.051 0.030
Chain 1: 1500 -8193.717 0.030 0.028
Chain 1: 1600 -8294.305 0.026 0.024
Chain 1: 1700 -8381.639 0.025 0.024
Chain 1: 1800 -7989.434 0.024 0.024
Chain 1: 1900 -8091.470 0.020 0.019
Chain 1: 2000 -8061.722 0.017 0.013
Chain 1: 2100 -8187.871 0.017 0.013
Chain 1: 2200 -7973.513 0.017 0.013
Chain 1: 2300 -8120.224 0.016 0.013
Chain 1: 2400 -8135.656 0.016 0.013
Chain 1: 2500 -8102.256 0.015 0.013
Chain 1: 2600 -8104.193 0.014 0.013
Chain 1: 2700 -8011.074 0.014 0.013
Chain 1: 2800 -7984.036 0.010 0.012 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003479 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.79 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8444181.459 1.000 1.000
Chain 1: 200 -1594388.416 2.648 4.296
Chain 1: 300 -892893.587 2.027 1.000
Chain 1: 400 -458057.797 1.758 1.000
Chain 1: 500 -357654.586 1.462 0.949
Chain 1: 600 -232123.074 1.309 0.949
Chain 1: 700 -118534.194 1.259 0.949
Chain 1: 800 -85820.161 1.149 0.949
Chain 1: 900 -66206.238 1.054 0.786
Chain 1: 1000 -51053.747 0.979 0.786
Chain 1: 1100 -38582.962 0.911 0.541 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37760.498 0.483 0.381
Chain 1: 1300 -25777.735 0.451 0.381
Chain 1: 1400 -25500.115 0.357 0.323
Chain 1: 1500 -22104.591 0.345 0.323
Chain 1: 1600 -21325.535 0.294 0.297
Chain 1: 1700 -20207.162 0.204 0.296
Chain 1: 1800 -20152.833 0.166 0.154
Chain 1: 1900 -20478.571 0.138 0.055
Chain 1: 2000 -18994.737 0.116 0.055
Chain 1: 2100 -19232.673 0.085 0.037
Chain 1: 2200 -19458.327 0.084 0.037
Chain 1: 2300 -19076.395 0.040 0.020
Chain 1: 2400 -18848.735 0.040 0.020
Chain 1: 2500 -18650.549 0.026 0.016
Chain 1: 2600 -18281.373 0.024 0.016
Chain 1: 2700 -18238.519 0.019 0.012
Chain 1: 2800 -17955.510 0.020 0.016
Chain 1: 2900 -18236.478 0.020 0.015
Chain 1: 3000 -18222.700 0.012 0.012
Chain 1: 3100 -18307.636 0.011 0.012
Chain 1: 3200 -17998.631 0.012 0.015
Chain 1: 3300 -18203.111 0.011 0.012
Chain 1: 3400 -17678.551 0.013 0.015
Chain 1: 3500 -18289.532 0.015 0.016
Chain 1: 3600 -17597.372 0.017 0.016
Chain 1: 3700 -17983.291 0.019 0.017
Chain 1: 3800 -16944.714 0.023 0.021
Chain 1: 3900 -16940.874 0.022 0.021
Chain 1: 4000 -17058.218 0.023 0.021
Chain 1: 4100 -16972.070 0.023 0.021
Chain 1: 4200 -16788.670 0.022 0.021
Chain 1: 4300 -16926.815 0.022 0.021
Chain 1: 4400 -16883.948 0.019 0.011
Chain 1: 4500 -16786.526 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001405 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.05 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12361.158 1.000 1.000
Chain 1: 200 -9242.721 0.669 1.000
Chain 1: 300 -8005.933 0.497 0.337
Chain 1: 400 -8234.040 0.380 0.337
Chain 1: 500 -7955.695 0.311 0.154
Chain 1: 600 -7970.899 0.259 0.154
Chain 1: 700 -7890.448 0.224 0.035
Chain 1: 800 -7928.598 0.196 0.035
Chain 1: 900 -8070.636 0.177 0.028
Chain 1: 1000 -7926.999 0.161 0.028
Chain 1: 1100 -7904.032 0.061 0.018
Chain 1: 1200 -7907.980 0.027 0.018
Chain 1: 1300 -7854.311 0.013 0.010
Chain 1: 1400 -7877.992 0.010 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001375 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.75 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57118.259 1.000 1.000
Chain 1: 200 -17547.478 1.628 2.255
Chain 1: 300 -8740.011 1.421 1.008
Chain 1: 400 -8243.438 1.081 1.008
Chain 1: 500 -8424.153 0.869 1.000
Chain 1: 600 -8000.963 0.733 1.000
Chain 1: 700 -7855.706 0.631 0.060
Chain 1: 800 -8067.836 0.555 0.060
Chain 1: 900 -7995.684 0.495 0.053
Chain 1: 1000 -7925.127 0.446 0.053
Chain 1: 1100 -7738.094 0.348 0.026
Chain 1: 1200 -7587.116 0.125 0.024
Chain 1: 1300 -7785.511 0.027 0.024
Chain 1: 1400 -7933.294 0.023 0.021
Chain 1: 1500 -7583.860 0.025 0.024
Chain 1: 1600 -7787.147 0.022 0.024
Chain 1: 1700 -7513.953 0.024 0.025
Chain 1: 1800 -7595.910 0.023 0.024
Chain 1: 1900 -7712.715 0.023 0.024
Chain 1: 2000 -7566.073 0.024 0.024
Chain 1: 2100 -7514.494 0.022 0.020
Chain 1: 2200 -7735.029 0.023 0.025
Chain 1: 2300 -7596.342 0.023 0.019
Chain 1: 2400 -7634.532 0.021 0.019
Chain 1: 2500 -7665.291 0.017 0.018
Chain 1: 2600 -7525.790 0.016 0.018
Chain 1: 2700 -7495.391 0.013 0.015
Chain 1: 2800 -7559.934 0.013 0.015
Chain 1: 2900 -7399.607 0.013 0.018
Chain 1: 3000 -7523.959 0.013 0.017
Chain 1: 3100 -7517.802 0.013 0.017
Chain 1: 3200 -7699.397 0.012 0.017
Chain 1: 3300 -7457.889 0.014 0.017
Chain 1: 3400 -7650.662 0.016 0.019
Chain 1: 3500 -7432.223 0.018 0.022
Chain 1: 3600 -7492.814 0.017 0.022
Chain 1: 3700 -7444.306 0.017 0.022
Chain 1: 3800 -7454.472 0.017 0.022
Chain 1: 3900 -7436.421 0.015 0.017
Chain 1: 4000 -7409.736 0.013 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003006 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.06 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85662.069 1.000 1.000
Chain 1: 200 -13524.312 3.167 5.334
Chain 1: 300 -9881.003 2.234 1.000
Chain 1: 400 -10796.507 1.697 1.000
Chain 1: 500 -8779.116 1.403 0.369
Chain 1: 600 -9342.953 1.180 0.369
Chain 1: 700 -8909.126 1.018 0.230
Chain 1: 800 -8434.773 0.898 0.230
Chain 1: 900 -8330.434 0.799 0.085
Chain 1: 1000 -8424.196 0.721 0.085
Chain 1: 1100 -8733.032 0.624 0.060 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8350.941 0.095 0.056
Chain 1: 1300 -8574.998 0.061 0.049
Chain 1: 1400 -8558.970 0.053 0.046
Chain 1: 1500 -8449.010 0.031 0.035
Chain 1: 1600 -8557.247 0.026 0.026
Chain 1: 1700 -8635.430 0.022 0.013
Chain 1: 1800 -8216.568 0.022 0.013
Chain 1: 1900 -8314.857 0.022 0.013
Chain 1: 2000 -8288.896 0.021 0.013
Chain 1: 2100 -8412.934 0.019 0.013
Chain 1: 2200 -8226.032 0.017 0.013
Chain 1: 2300 -8309.554 0.015 0.013
Chain 1: 2400 -8379.053 0.016 0.013
Chain 1: 2500 -8324.994 0.015 0.012
Chain 1: 2600 -8325.368 0.014 0.010
Chain 1: 2700 -8242.514 0.014 0.010
Chain 1: 2800 -8204.020 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003162 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.62 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8365558.671 1.000 1.000
Chain 1: 200 -1580513.903 2.646 4.293
Chain 1: 300 -891332.925 2.022 1.000
Chain 1: 400 -458288.650 1.753 1.000
Chain 1: 500 -359126.079 1.457 0.945
Chain 1: 600 -233902.577 1.304 0.945
Chain 1: 700 -119717.610 1.254 0.945
Chain 1: 800 -86824.862 1.144 0.945
Chain 1: 900 -67077.688 1.050 0.773
Chain 1: 1000 -51807.213 0.974 0.773
Chain 1: 1100 -39219.338 0.907 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38388.845 0.479 0.379
Chain 1: 1300 -26276.680 0.448 0.379
Chain 1: 1400 -25989.523 0.355 0.321
Chain 1: 1500 -22558.889 0.342 0.321
Chain 1: 1600 -21770.327 0.292 0.295
Chain 1: 1700 -20635.740 0.203 0.294
Chain 1: 1800 -20578.093 0.165 0.152
Chain 1: 1900 -20904.280 0.137 0.055
Chain 1: 2000 -19411.129 0.115 0.055
Chain 1: 2100 -19649.634 0.084 0.036
Chain 1: 2200 -19876.857 0.083 0.036
Chain 1: 2300 -19493.401 0.039 0.020
Chain 1: 2400 -19265.383 0.039 0.020
Chain 1: 2500 -19067.743 0.025 0.016
Chain 1: 2600 -18697.496 0.024 0.016
Chain 1: 2700 -18654.406 0.018 0.012
Chain 1: 2800 -18371.329 0.020 0.015
Chain 1: 2900 -18652.764 0.019 0.015
Chain 1: 3000 -18638.835 0.012 0.012
Chain 1: 3100 -18723.826 0.011 0.012
Chain 1: 3200 -18414.423 0.012 0.015
Chain 1: 3300 -18619.261 0.011 0.012
Chain 1: 3400 -18094.081 0.013 0.015
Chain 1: 3500 -18706.217 0.015 0.015
Chain 1: 3600 -18012.670 0.017 0.015
Chain 1: 3700 -18399.646 0.018 0.017
Chain 1: 3800 -17359.049 0.023 0.021
Chain 1: 3900 -17355.270 0.021 0.021
Chain 1: 4000 -17472.505 0.022 0.021
Chain 1: 4100 -17386.226 0.022 0.021
Chain 1: 4200 -17202.465 0.021 0.021
Chain 1: 4300 -17340.835 0.021 0.021
Chain 1: 4400 -17297.596 0.019 0.011
Chain 1: 4500 -17200.186 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001397 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.97 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48500.543 1.000 1.000
Chain 1: 200 -16287.959 1.489 1.978
Chain 1: 300 -17385.226 1.014 1.000
Chain 1: 400 -13506.132 0.832 1.000
Chain 1: 500 -11573.471 0.699 0.287
Chain 1: 600 -14798.255 0.619 0.287
Chain 1: 700 -11826.492 0.566 0.251
Chain 1: 800 -14294.380 0.517 0.251
Chain 1: 900 -14079.050 0.461 0.218
Chain 1: 1000 -13372.393 0.420 0.218
Chain 1: 1100 -15911.041 0.336 0.173
Chain 1: 1200 -15930.172 0.139 0.167
Chain 1: 1300 -10768.001 0.180 0.173
Chain 1: 1400 -13102.808 0.170 0.173
Chain 1: 1500 -10544.728 0.177 0.178
Chain 1: 1600 -10512.022 0.156 0.173
Chain 1: 1700 -9469.309 0.141 0.160
Chain 1: 1800 -15248.638 0.162 0.160
Chain 1: 1900 -13937.758 0.170 0.160
Chain 1: 2000 -9751.678 0.208 0.178
Chain 1: 2100 -9701.255 0.192 0.178
Chain 1: 2200 -10021.776 0.195 0.178
Chain 1: 2300 -11182.905 0.158 0.110
Chain 1: 2400 -8680.849 0.169 0.110
Chain 1: 2500 -10352.649 0.161 0.110
Chain 1: 2600 -9696.108 0.167 0.110
Chain 1: 2700 -8975.511 0.164 0.104
Chain 1: 2800 -13814.890 0.161 0.104
Chain 1: 2900 -9058.265 0.204 0.161
Chain 1: 3000 -9241.822 0.163 0.104
Chain 1: 3100 -8454.217 0.172 0.104
Chain 1: 3200 -8574.852 0.170 0.104
Chain 1: 3300 -9661.109 0.171 0.112
Chain 1: 3400 -8961.732 0.150 0.093
Chain 1: 3500 -9092.235 0.136 0.080
Chain 1: 3600 -10820.764 0.145 0.093
Chain 1: 3700 -8846.940 0.159 0.112
Chain 1: 3800 -14287.375 0.162 0.112
Chain 1: 3900 -11231.588 0.137 0.112
Chain 1: 4000 -8441.817 0.168 0.160
Chain 1: 4100 -15001.768 0.202 0.223
Chain 1: 4200 -8391.652 0.280 0.272
Chain 1: 4300 -10083.525 0.285 0.272
Chain 1: 4400 -12112.691 0.294 0.272
Chain 1: 4500 -8702.841 0.332 0.330
Chain 1: 4600 -8840.186 0.317 0.330
Chain 1: 4700 -8495.832 0.299 0.330
Chain 1: 4800 -8163.151 0.265 0.272
Chain 1: 4900 -9321.720 0.250 0.168
Chain 1: 5000 -9296.779 0.218 0.168
Chain 1: 5100 -8886.078 0.178 0.124
Chain 1: 5200 -11833.433 0.125 0.124
Chain 1: 5300 -11812.585 0.108 0.046
Chain 1: 5400 -12809.871 0.099 0.046
Chain 1: 5500 -8167.960 0.117 0.046
Chain 1: 5600 -8222.897 0.116 0.046
Chain 1: 5700 -14460.598 0.155 0.078
Chain 1: 5800 -8786.310 0.215 0.124
Chain 1: 5900 -8606.967 0.205 0.078
Chain 1: 6000 -11481.335 0.230 0.249
Chain 1: 6100 -11848.229 0.228 0.249
Chain 1: 6200 -8089.942 0.250 0.250
Chain 1: 6300 -8595.040 0.256 0.250
Chain 1: 6400 -12705.858 0.280 0.324
Chain 1: 6500 -8889.377 0.266 0.324
Chain 1: 6600 -10785.560 0.283 0.324
Chain 1: 6700 -12483.044 0.254 0.250
Chain 1: 6800 -8554.502 0.235 0.250
Chain 1: 6900 -8549.023 0.233 0.250
Chain 1: 7000 -10409.326 0.226 0.179
Chain 1: 7100 -7962.647 0.253 0.307
Chain 1: 7200 -10914.457 0.234 0.270
Chain 1: 7300 -10666.820 0.230 0.270
Chain 1: 7400 -11670.400 0.207 0.179
Chain 1: 7500 -8733.063 0.197 0.179
Chain 1: 7600 -8169.551 0.187 0.179
Chain 1: 7700 -8851.417 0.181 0.179
Chain 1: 7800 -8202.784 0.143 0.086
Chain 1: 7900 -8659.119 0.148 0.086
Chain 1: 8000 -8043.941 0.138 0.079
Chain 1: 8100 -10170.090 0.128 0.079
Chain 1: 8200 -10478.941 0.104 0.077
Chain 1: 8300 -11385.569 0.109 0.079
Chain 1: 8400 -9336.193 0.123 0.079
Chain 1: 8500 -8948.613 0.094 0.077
Chain 1: 8600 -8499.920 0.092 0.077
Chain 1: 8700 -8180.468 0.088 0.076
Chain 1: 8800 -8304.889 0.082 0.053
Chain 1: 8900 -9178.830 0.086 0.076
Chain 1: 9000 -11557.133 0.099 0.080
Chain 1: 9100 -8676.615 0.111 0.080
Chain 1: 9200 -8399.646 0.112 0.080
Chain 1: 9300 -8172.467 0.106 0.053
Chain 1: 9400 -11642.734 0.114 0.053
Chain 1: 9500 -8416.162 0.148 0.095
Chain 1: 9600 -8193.838 0.146 0.095
Chain 1: 9700 -8717.729 0.148 0.095
Chain 1: 9800 -7977.812 0.156 0.095
Chain 1: 9900 -10586.032 0.171 0.206
Chain 1: 10000 -8116.044 0.180 0.246
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001417 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.17 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61389.655 1.000 1.000
Chain 1: 200 -17430.841 1.761 2.522
Chain 1: 300 -8629.416 1.514 1.020
Chain 1: 400 -8837.705 1.141 1.020
Chain 1: 500 -7948.258 0.935 1.000
Chain 1: 600 -8649.917 0.793 1.000
Chain 1: 700 -8206.927 0.687 0.112
Chain 1: 800 -8072.596 0.604 0.112
Chain 1: 900 -7541.495 0.544 0.081
Chain 1: 1000 -7667.334 0.492 0.081
Chain 1: 1100 -7577.761 0.393 0.070
Chain 1: 1200 -7621.041 0.141 0.054
Chain 1: 1300 -7522.168 0.040 0.024
Chain 1: 1400 -7629.270 0.040 0.017
Chain 1: 1500 -7564.611 0.029 0.016
Chain 1: 1600 -7472.315 0.022 0.014
Chain 1: 1700 -7449.626 0.017 0.013
Chain 1: 1800 -7473.025 0.016 0.012
Chain 1: 1900 -7557.854 0.010 0.012 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003255 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.55 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85854.794 1.000 1.000
Chain 1: 200 -13073.083 3.284 5.567
Chain 1: 300 -9580.132 2.311 1.000
Chain 1: 400 -10482.953 1.755 1.000
Chain 1: 500 -8421.751 1.453 0.365
Chain 1: 600 -8197.982 1.215 0.365
Chain 1: 700 -8470.192 1.046 0.245
Chain 1: 800 -8686.624 0.918 0.245
Chain 1: 900 -8477.555 0.819 0.086
Chain 1: 1000 -8232.372 0.740 0.086
Chain 1: 1100 -8477.309 0.643 0.032 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8153.285 0.090 0.032
Chain 1: 1300 -8202.699 0.054 0.030
Chain 1: 1400 -8198.766 0.046 0.029
Chain 1: 1500 -8230.357 0.022 0.027
Chain 1: 1600 -8235.930 0.019 0.025
Chain 1: 1700 -8177.404 0.017 0.025
Chain 1: 1800 -8056.180 0.016 0.015
Chain 1: 1900 -8169.380 0.015 0.014
Chain 1: 2000 -8131.515 0.012 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004164 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 41.64 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8424870.899 1.000 1.000
Chain 1: 200 -1589506.294 2.650 4.300
Chain 1: 300 -891168.171 2.028 1.000
Chain 1: 400 -456985.353 1.759 1.000
Chain 1: 500 -356797.149 1.463 0.950
Chain 1: 600 -231837.867 1.309 0.950
Chain 1: 700 -118394.020 1.259 0.950
Chain 1: 800 -85688.191 1.149 0.950
Chain 1: 900 -66104.062 1.054 0.784
Chain 1: 1000 -50952.675 0.979 0.784
Chain 1: 1100 -38482.222 0.911 0.539 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37659.143 0.483 0.382
Chain 1: 1300 -25682.284 0.452 0.382
Chain 1: 1400 -25403.937 0.358 0.324
Chain 1: 1500 -22009.294 0.345 0.324
Chain 1: 1600 -21230.080 0.295 0.297
Chain 1: 1700 -20112.673 0.205 0.296
Chain 1: 1800 -20058.365 0.167 0.154
Chain 1: 1900 -20383.694 0.139 0.056
Chain 1: 2000 -18901.072 0.117 0.056
Chain 1: 2100 -19139.019 0.086 0.037
Chain 1: 2200 -19364.206 0.084 0.037
Chain 1: 2300 -18982.775 0.040 0.020
Chain 1: 2400 -18755.271 0.040 0.020
Chain 1: 2500 -18557.034 0.026 0.016
Chain 1: 2600 -18188.339 0.024 0.016
Chain 1: 2700 -18145.665 0.019 0.012
Chain 1: 2800 -17862.788 0.020 0.016
Chain 1: 2900 -18143.568 0.020 0.015
Chain 1: 3000 -18129.873 0.012 0.012
Chain 1: 3100 -18214.711 0.011 0.012
Chain 1: 3200 -17906.014 0.012 0.015
Chain 1: 3300 -18110.271 0.011 0.012
Chain 1: 3400 -17586.211 0.013 0.015
Chain 1: 3500 -18196.430 0.015 0.016
Chain 1: 3600 -17505.295 0.017 0.016
Chain 1: 3700 -17890.419 0.019 0.017
Chain 1: 3800 -16853.432 0.024 0.022
Chain 1: 3900 -16849.638 0.022 0.022
Chain 1: 4000 -16966.970 0.023 0.022
Chain 1: 4100 -16880.862 0.023 0.022
Chain 1: 4200 -16697.865 0.022 0.022
Chain 1: 4300 -16835.752 0.022 0.022
Chain 1: 4400 -16793.167 0.019 0.011
Chain 1: 4500 -16695.796 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001548 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.48 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12512.833 1.000 1.000
Chain 1: 200 -9351.396 0.669 1.000
Chain 1: 300 -8012.993 0.502 0.338
Chain 1: 400 -8103.986 0.379 0.338
Chain 1: 500 -8088.168 0.304 0.167
Chain 1: 600 -7946.932 0.256 0.167
Chain 1: 700 -7840.311 0.221 0.018
Chain 1: 800 -7860.817 0.194 0.018
Chain 1: 900 -7882.151 0.173 0.014
Chain 1: 1000 -8025.292 0.157 0.018
Chain 1: 1100 -7946.300 0.058 0.014
Chain 1: 1200 -7849.586 0.026 0.012
Chain 1: 1300 -7830.834 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00177 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 17.7 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -59908.869 1.000 1.000
Chain 1: 200 -17926.301 1.671 2.342
Chain 1: 300 -8721.899 1.466 1.055
Chain 1: 400 -8467.811 1.107 1.055
Chain 1: 500 -8305.546 0.889 1.000
Chain 1: 600 -9113.662 0.756 1.000
Chain 1: 700 -8327.310 0.661 0.094
Chain 1: 800 -8038.962 0.583 0.094
Chain 1: 900 -8085.658 0.519 0.089
Chain 1: 1000 -7631.362 0.473 0.089
Chain 1: 1100 -7823.454 0.376 0.060
Chain 1: 1200 -7935.663 0.143 0.036
Chain 1: 1300 -7724.784 0.040 0.030
Chain 1: 1400 -7925.678 0.040 0.027
Chain 1: 1500 -7565.029 0.042 0.036
Chain 1: 1600 -7649.738 0.035 0.027
Chain 1: 1700 -7498.442 0.027 0.025
Chain 1: 1800 -7589.509 0.025 0.025
Chain 1: 1900 -7551.800 0.025 0.025
Chain 1: 2000 -7610.331 0.019 0.020
Chain 1: 2100 -7557.260 0.018 0.014
Chain 1: 2200 -7675.438 0.018 0.015
Chain 1: 2300 -7584.072 0.016 0.012
Chain 1: 2400 -7629.757 0.014 0.012
Chain 1: 2500 -7577.922 0.010 0.011
Chain 1: 2600 -7499.228 0.010 0.010
Chain 1: 2700 -7515.333 0.008 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003948 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 39.48 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86839.198 1.000 1.000
Chain 1: 200 -13528.054 3.210 5.419
Chain 1: 300 -9856.618 2.264 1.000
Chain 1: 400 -10907.096 1.722 1.000
Chain 1: 500 -8829.847 1.425 0.372
Chain 1: 600 -8848.213 1.188 0.372
Chain 1: 700 -8264.984 1.028 0.235
Chain 1: 800 -9218.645 0.912 0.235
Chain 1: 900 -8563.011 0.820 0.103
Chain 1: 1000 -8306.720 0.741 0.103
Chain 1: 1100 -8679.628 0.645 0.096 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8202.241 0.109 0.077
Chain 1: 1300 -8538.529 0.076 0.071
Chain 1: 1400 -8520.069 0.066 0.058
Chain 1: 1500 -8416.404 0.044 0.043
Chain 1: 1600 -8520.119 0.045 0.043
Chain 1: 1700 -8595.774 0.039 0.039
Chain 1: 1800 -8172.568 0.034 0.039
Chain 1: 1900 -8272.976 0.027 0.031
Chain 1: 2000 -8247.466 0.024 0.012
Chain 1: 2100 -8373.032 0.022 0.012
Chain 1: 2200 -8176.097 0.018 0.012
Chain 1: 2300 -8267.877 0.015 0.012
Chain 1: 2400 -8336.648 0.016 0.012
Chain 1: 2500 -8282.866 0.015 0.012
Chain 1: 2600 -8284.232 0.014 0.011
Chain 1: 2700 -8200.976 0.014 0.011
Chain 1: 2800 -8160.826 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003955 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 39.55 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8407762.333 1.000 1.000
Chain 1: 200 -1583517.103 2.655 4.310
Chain 1: 300 -891597.187 2.029 1.000
Chain 1: 400 -458399.871 1.758 1.000
Chain 1: 500 -358745.031 1.462 0.945
Chain 1: 600 -233383.625 1.308 0.945
Chain 1: 700 -119422.635 1.257 0.945
Chain 1: 800 -86595.155 1.147 0.945
Chain 1: 900 -66893.000 1.053 0.776
Chain 1: 1000 -51661.508 0.977 0.776
Chain 1: 1100 -39120.280 0.909 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38294.607 0.480 0.379
Chain 1: 1300 -26231.930 0.448 0.379
Chain 1: 1400 -25949.376 0.355 0.321
Chain 1: 1500 -22532.058 0.342 0.321
Chain 1: 1600 -21747.606 0.292 0.295
Chain 1: 1700 -20618.942 0.202 0.295
Chain 1: 1800 -20562.719 0.165 0.152
Chain 1: 1900 -20888.917 0.137 0.055
Chain 1: 2000 -19398.823 0.115 0.055
Chain 1: 2100 -19637.248 0.084 0.036
Chain 1: 2200 -19863.959 0.083 0.036
Chain 1: 2300 -19480.904 0.039 0.020
Chain 1: 2400 -19252.917 0.039 0.020
Chain 1: 2500 -19055.042 0.025 0.016
Chain 1: 2600 -18684.977 0.024 0.016
Chain 1: 2700 -18641.912 0.018 0.012
Chain 1: 2800 -18358.741 0.020 0.015
Chain 1: 2900 -18640.102 0.019 0.015
Chain 1: 3000 -18626.173 0.012 0.012
Chain 1: 3100 -18711.221 0.011 0.012
Chain 1: 3200 -18401.794 0.012 0.015
Chain 1: 3300 -18606.625 0.011 0.012
Chain 1: 3400 -18081.410 0.013 0.015
Chain 1: 3500 -18693.495 0.015 0.015
Chain 1: 3600 -17999.880 0.017 0.015
Chain 1: 3700 -18386.926 0.019 0.017
Chain 1: 3800 -17346.192 0.023 0.021
Chain 1: 3900 -17342.334 0.021 0.021
Chain 1: 4000 -17459.622 0.022 0.021
Chain 1: 4100 -17373.346 0.022 0.021
Chain 1: 4200 -17189.505 0.021 0.021
Chain 1: 4300 -17327.950 0.021 0.021
Chain 1: 4400 -17284.709 0.019 0.011
Chain 1: 4500 -17187.209 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001375 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.75 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13054.691 1.000 1.000
Chain 1: 200 -9729.779 0.671 1.000
Chain 1: 300 -8192.984 0.510 0.342
Chain 1: 400 -8241.082 0.384 0.342
Chain 1: 500 -8201.466 0.308 0.188
Chain 1: 600 -8036.818 0.260 0.188
Chain 1: 700 -7955.216 0.224 0.020
Chain 1: 800 -7960.379 0.196 0.020
Chain 1: 900 -7854.300 0.176 0.014
Chain 1: 1000 -8066.966 0.161 0.020
Chain 1: 1100 -7999.663 0.062 0.014
Chain 1: 1200 -7965.973 0.028 0.010
Chain 1: 1300 -7931.729 0.010 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002918 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.18 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58103.318 1.000 1.000
Chain 1: 200 -17716.302 1.640 2.280
Chain 1: 300 -8713.229 1.438 1.033
Chain 1: 400 -8224.635 1.093 1.033
Chain 1: 500 -8440.514 0.880 1.000
Chain 1: 600 -8360.243 0.735 1.000
Chain 1: 700 -8458.500 0.631 0.059
Chain 1: 800 -8066.665 0.558 0.059
Chain 1: 900 -7955.274 0.498 0.049
Chain 1: 1000 -7713.699 0.451 0.049
Chain 1: 1100 -7687.706 0.352 0.031
Chain 1: 1200 -7648.176 0.124 0.026
Chain 1: 1300 -7751.957 0.022 0.014
Chain 1: 1400 -7982.175 0.019 0.014
Chain 1: 1500 -7643.525 0.021 0.014
Chain 1: 1600 -7619.943 0.020 0.014
Chain 1: 1700 -7561.779 0.020 0.014
Chain 1: 1800 -7638.890 0.016 0.013
Chain 1: 1900 -7645.700 0.015 0.010
Chain 1: 2000 -7675.391 0.012 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003905 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 39.05 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85912.028 1.000 1.000
Chain 1: 200 -13532.400 3.174 5.349
Chain 1: 300 -9899.135 2.239 1.000
Chain 1: 400 -10814.843 1.700 1.000
Chain 1: 500 -8825.401 1.405 0.367
Chain 1: 600 -8457.009 1.178 0.367
Chain 1: 700 -8403.694 1.011 0.225
Chain 1: 800 -9210.789 0.895 0.225
Chain 1: 900 -8700.725 0.802 0.088
Chain 1: 1000 -8475.197 0.725 0.088
Chain 1: 1100 -8710.141 0.628 0.085 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8421.406 0.096 0.059
Chain 1: 1300 -8633.666 0.062 0.044
Chain 1: 1400 -8625.154 0.054 0.034
Chain 1: 1500 -8487.871 0.033 0.027
Chain 1: 1600 -8598.927 0.030 0.027
Chain 1: 1700 -8685.666 0.030 0.027
Chain 1: 1800 -8278.844 0.026 0.027
Chain 1: 1900 -8375.213 0.021 0.025
Chain 1: 2000 -8347.460 0.019 0.016
Chain 1: 2100 -8468.182 0.018 0.014
Chain 1: 2200 -8284.345 0.017 0.014
Chain 1: 2300 -8414.997 0.016 0.014
Chain 1: 2400 -8424.778 0.016 0.014
Chain 1: 2500 -8387.661 0.014 0.013
Chain 1: 2600 -8386.092 0.013 0.012
Chain 1: 2700 -8300.798 0.013 0.012
Chain 1: 2800 -8265.669 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003695 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 36.95 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8380392.527 1.000 1.000
Chain 1: 200 -1581442.553 2.650 4.299
Chain 1: 300 -891326.931 2.024 1.000
Chain 1: 400 -458749.338 1.754 1.000
Chain 1: 500 -359338.316 1.459 0.943
Chain 1: 600 -233956.145 1.305 0.943
Chain 1: 700 -119708.697 1.255 0.943
Chain 1: 800 -86797.236 1.145 0.943
Chain 1: 900 -67043.895 1.051 0.774
Chain 1: 1000 -51770.503 0.975 0.774
Chain 1: 1100 -39185.720 0.907 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38350.824 0.480 0.379
Chain 1: 1300 -26248.160 0.448 0.379
Chain 1: 1400 -25960.264 0.355 0.321
Chain 1: 1500 -22532.704 0.343 0.321
Chain 1: 1600 -21744.558 0.293 0.295
Chain 1: 1700 -20611.373 0.203 0.295
Chain 1: 1800 -20553.740 0.165 0.152
Chain 1: 1900 -20879.748 0.137 0.055
Chain 1: 2000 -19387.615 0.115 0.055
Chain 1: 2100 -19626.034 0.084 0.036
Chain 1: 2200 -19853.095 0.083 0.036
Chain 1: 2300 -19469.797 0.039 0.020
Chain 1: 2400 -19241.865 0.039 0.020
Chain 1: 2500 -19044.208 0.025 0.016
Chain 1: 2600 -18674.262 0.024 0.016
Chain 1: 2700 -18631.143 0.018 0.012
Chain 1: 2800 -18348.237 0.020 0.015
Chain 1: 2900 -18629.467 0.020 0.015
Chain 1: 3000 -18615.593 0.012 0.012
Chain 1: 3100 -18700.607 0.011 0.012
Chain 1: 3200 -18391.328 0.012 0.015
Chain 1: 3300 -18595.992 0.011 0.012
Chain 1: 3400 -18071.118 0.013 0.015
Chain 1: 3500 -18682.851 0.015 0.015
Chain 1: 3600 -17989.710 0.017 0.015
Chain 1: 3700 -18376.448 0.018 0.017
Chain 1: 3800 -17336.556 0.023 0.021
Chain 1: 3900 -17332.759 0.021 0.021
Chain 1: 4000 -17450.003 0.022 0.021
Chain 1: 4100 -17363.851 0.022 0.021
Chain 1: 4200 -17180.138 0.021 0.021
Chain 1: 4300 -17318.465 0.021 0.021
Chain 1: 4400 -17275.361 0.019 0.011
Chain 1: 4500 -17177.946 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001428 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.28 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48787.797 1.000 1.000
Chain 1: 200 -14473.094 1.685 2.371
Chain 1: 300 -20621.568 1.223 1.000
Chain 1: 400 -15097.656 1.009 1.000
Chain 1: 500 -18086.603 0.840 0.366
Chain 1: 600 -11733.230 0.790 0.541
Chain 1: 700 -13192.384 0.693 0.366
Chain 1: 800 -14304.544 0.616 0.366
Chain 1: 900 -10977.817 0.581 0.303
Chain 1: 1000 -25166.807 0.580 0.366
Chain 1: 1100 -10441.121 0.621 0.366 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -14366.209 0.411 0.303
Chain 1: 1300 -12481.592 0.396 0.303
Chain 1: 1400 -10632.441 0.377 0.273
Chain 1: 1500 -11843.590 0.371 0.273
Chain 1: 1600 -10738.627 0.327 0.174
Chain 1: 1700 -20047.469 0.362 0.273
Chain 1: 1800 -13981.922 0.398 0.303
Chain 1: 1900 -10657.565 0.399 0.312
Chain 1: 2000 -11869.104 0.353 0.273
Chain 1: 2100 -9481.804 0.237 0.252
Chain 1: 2200 -11195.985 0.225 0.174
Chain 1: 2300 -9138.173 0.232 0.225
Chain 1: 2400 -9333.339 0.217 0.225
Chain 1: 2500 -9690.037 0.210 0.225
Chain 1: 2600 -9518.263 0.202 0.225
Chain 1: 2700 -9147.770 0.159 0.153
Chain 1: 2800 -10446.856 0.128 0.124
Chain 1: 2900 -9997.619 0.102 0.102
Chain 1: 3000 -10858.813 0.099 0.079
Chain 1: 3100 -10704.723 0.076 0.045
Chain 1: 3200 -9037.262 0.079 0.045
Chain 1: 3300 -10633.424 0.071 0.045
Chain 1: 3400 -9090.313 0.086 0.079
Chain 1: 3500 -9475.613 0.087 0.079
Chain 1: 3600 -11607.010 0.103 0.124
Chain 1: 3700 -15123.844 0.122 0.150
Chain 1: 3800 -8657.479 0.185 0.170
Chain 1: 3900 -9764.439 0.192 0.170
Chain 1: 4000 -9312.200 0.188 0.170
Chain 1: 4100 -8833.869 0.192 0.170
Chain 1: 4200 -8649.924 0.176 0.150
Chain 1: 4300 -15557.802 0.205 0.170
Chain 1: 4400 -9181.057 0.258 0.184
Chain 1: 4500 -8871.601 0.257 0.184
Chain 1: 4600 -13844.039 0.275 0.233
Chain 1: 4700 -12298.625 0.264 0.126
Chain 1: 4800 -12690.184 0.193 0.113
Chain 1: 4900 -8583.296 0.229 0.126
Chain 1: 5000 -10477.278 0.242 0.181
Chain 1: 5100 -8494.140 0.260 0.233
Chain 1: 5200 -10689.479 0.279 0.233
Chain 1: 5300 -12472.362 0.249 0.205
Chain 1: 5400 -13246.496 0.185 0.181
Chain 1: 5500 -8574.151 0.236 0.205
Chain 1: 5600 -9907.340 0.214 0.181
Chain 1: 5700 -9065.251 0.210 0.181
Chain 1: 5800 -9389.023 0.211 0.181
Chain 1: 5900 -8853.439 0.169 0.143
Chain 1: 6000 -8721.017 0.152 0.135
Chain 1: 6100 -8457.832 0.132 0.093
Chain 1: 6200 -8481.587 0.112 0.060
Chain 1: 6300 -12904.965 0.132 0.060
Chain 1: 6400 -9910.696 0.156 0.093
Chain 1: 6500 -12136.995 0.120 0.093
Chain 1: 6600 -8397.535 0.151 0.093
Chain 1: 6700 -8385.905 0.142 0.060
Chain 1: 6800 -8947.699 0.145 0.063
Chain 1: 6900 -8814.580 0.140 0.063
Chain 1: 7000 -10899.610 0.158 0.183
Chain 1: 7100 -8275.540 0.186 0.191
Chain 1: 7200 -8738.343 0.191 0.191
Chain 1: 7300 -10306.210 0.172 0.183
Chain 1: 7400 -9106.294 0.155 0.152
Chain 1: 7500 -11541.950 0.158 0.152
Chain 1: 7600 -10335.201 0.125 0.132
Chain 1: 7700 -8183.054 0.151 0.152
Chain 1: 7800 -8244.456 0.146 0.152
Chain 1: 7900 -8168.211 0.145 0.152
Chain 1: 8000 -9675.320 0.142 0.152
Chain 1: 8100 -8615.699 0.122 0.132
Chain 1: 8200 -8645.378 0.117 0.132
Chain 1: 8300 -12271.571 0.132 0.132
Chain 1: 8400 -8930.265 0.156 0.156
Chain 1: 8500 -11638.452 0.158 0.156
Chain 1: 8600 -9245.473 0.172 0.233
Chain 1: 8700 -8115.239 0.160 0.156
Chain 1: 8800 -8641.120 0.165 0.156
Chain 1: 8900 -11457.086 0.189 0.233
Chain 1: 9000 -10505.756 0.182 0.233
Chain 1: 9100 -8339.490 0.196 0.246
Chain 1: 9200 -8295.404 0.196 0.246
Chain 1: 9300 -11754.218 0.196 0.246
Chain 1: 9400 -8243.979 0.201 0.246
Chain 1: 9500 -8200.687 0.179 0.246
Chain 1: 9600 -8341.205 0.154 0.139
Chain 1: 9700 -9955.560 0.157 0.162
Chain 1: 9800 -8311.643 0.170 0.198
Chain 1: 9900 -9284.753 0.156 0.162
Chain 1: 10000 -8082.817 0.162 0.162
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001468 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.68 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58347.421 1.000 1.000
Chain 1: 200 -17665.248 1.651 2.303
Chain 1: 300 -8665.457 1.447 1.039
Chain 1: 400 -8242.172 1.098 1.039
Chain 1: 500 -8194.042 0.880 1.000
Chain 1: 600 -8322.476 0.736 1.000
Chain 1: 700 -7809.529 0.640 0.066
Chain 1: 800 -8181.914 0.566 0.066
Chain 1: 900 -7960.680 0.506 0.051
Chain 1: 1000 -7644.678 0.459 0.051
Chain 1: 1100 -7663.845 0.360 0.046
Chain 1: 1200 -7610.208 0.130 0.041
Chain 1: 1300 -7728.413 0.028 0.028
Chain 1: 1400 -7810.353 0.024 0.015
Chain 1: 1500 -7583.341 0.026 0.028
Chain 1: 1600 -7743.646 0.027 0.028
Chain 1: 1700 -7515.422 0.023 0.028
Chain 1: 1800 -7583.599 0.019 0.021
Chain 1: 1900 -7571.859 0.017 0.015
Chain 1: 2000 -7606.319 0.013 0.010
Chain 1: 2100 -7586.861 0.013 0.010
Chain 1: 2200 -7708.887 0.014 0.015
Chain 1: 2300 -7603.133 0.014 0.014
Chain 1: 2400 -7605.672 0.013 0.014
Chain 1: 2500 -7637.387 0.010 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003347 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.47 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86085.692 1.000 1.000
Chain 1: 200 -13422.061 3.207 5.414
Chain 1: 300 -9822.179 2.260 1.000
Chain 1: 400 -10639.675 1.714 1.000
Chain 1: 500 -8798.163 1.413 0.367
Chain 1: 600 -8300.556 1.188 0.367
Chain 1: 700 -8405.468 1.020 0.209
Chain 1: 800 -8732.776 0.897 0.209
Chain 1: 900 -8643.418 0.799 0.077
Chain 1: 1000 -8452.522 0.721 0.077
Chain 1: 1100 -8682.038 0.624 0.060 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8218.712 0.088 0.056
Chain 1: 1300 -8532.297 0.055 0.037
Chain 1: 1400 -8531.539 0.047 0.037
Chain 1: 1500 -8405.466 0.028 0.026
Chain 1: 1600 -8512.962 0.023 0.023
Chain 1: 1700 -8599.228 0.023 0.023
Chain 1: 1800 -8192.585 0.024 0.023
Chain 1: 1900 -8289.463 0.024 0.023
Chain 1: 2000 -8261.522 0.022 0.015
Chain 1: 2100 -8382.067 0.021 0.014
Chain 1: 2200 -8193.175 0.018 0.014
Chain 1: 2300 -8329.236 0.016 0.014
Chain 1: 2400 -8336.559 0.016 0.014
Chain 1: 2500 -8302.650 0.015 0.013
Chain 1: 2600 -8300.615 0.013 0.012
Chain 1: 2700 -8214.746 0.013 0.012
Chain 1: 2800 -8179.907 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003314 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.14 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8429437.085 1.000 1.000
Chain 1: 200 -1587763.115 2.655 4.309
Chain 1: 300 -889737.820 2.031 1.000
Chain 1: 400 -457043.605 1.760 1.000
Chain 1: 500 -356876.377 1.464 0.947
Chain 1: 600 -232045.087 1.310 0.947
Chain 1: 700 -118706.704 1.259 0.947
Chain 1: 800 -86028.060 1.149 0.947
Chain 1: 900 -66455.231 1.054 0.785
Chain 1: 1000 -51320.991 0.978 0.785
Chain 1: 1100 -38863.333 0.910 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38049.026 0.482 0.380
Chain 1: 1300 -26073.594 0.449 0.380
Chain 1: 1400 -25799.312 0.355 0.321
Chain 1: 1500 -22404.370 0.343 0.321
Chain 1: 1600 -21626.442 0.292 0.295
Chain 1: 1700 -20508.084 0.202 0.295
Chain 1: 1800 -20454.278 0.165 0.152
Chain 1: 1900 -20780.268 0.137 0.055
Chain 1: 2000 -19296.148 0.115 0.055
Chain 1: 2100 -19534.214 0.084 0.036
Chain 1: 2200 -19759.878 0.083 0.036
Chain 1: 2300 -19377.861 0.039 0.020
Chain 1: 2400 -19150.087 0.039 0.020
Chain 1: 2500 -18951.884 0.025 0.016
Chain 1: 2600 -18582.422 0.024 0.016
Chain 1: 2700 -18539.580 0.018 0.012
Chain 1: 2800 -18256.340 0.020 0.016
Chain 1: 2900 -18537.508 0.020 0.015
Chain 1: 3000 -18523.736 0.012 0.012
Chain 1: 3100 -18608.662 0.011 0.012
Chain 1: 3200 -18299.515 0.012 0.015
Chain 1: 3300 -18504.172 0.011 0.012
Chain 1: 3400 -17979.265 0.013 0.015
Chain 1: 3500 -18590.702 0.015 0.016
Chain 1: 3600 -17897.996 0.017 0.016
Chain 1: 3700 -18284.265 0.019 0.017
Chain 1: 3800 -17244.802 0.023 0.021
Chain 1: 3900 -17240.957 0.022 0.021
Chain 1: 4000 -17358.305 0.022 0.021
Chain 1: 4100 -17272.013 0.022 0.021
Chain 1: 4200 -17088.525 0.022 0.021
Chain 1: 4300 -17226.772 0.021 0.021
Chain 1: 4400 -17183.743 0.019 0.011
Chain 1: 4500 -17086.290 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001242 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.42 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48905.209 1.000 1.000
Chain 1: 200 -15251.819 1.603 2.207
Chain 1: 300 -13695.223 1.107 1.000
Chain 1: 400 -20044.462 0.909 1.000
Chain 1: 500 -21872.997 0.744 0.317
Chain 1: 600 -11758.314 0.763 0.860
Chain 1: 700 -16784.762 0.697 0.317
Chain 1: 800 -13490.542 0.641 0.317
Chain 1: 900 -11113.006 0.593 0.299
Chain 1: 1000 -12160.251 0.542 0.299
Chain 1: 1100 -12507.007 0.445 0.244
Chain 1: 1200 -10793.269 0.240 0.214
Chain 1: 1300 -9903.354 0.238 0.214
Chain 1: 1400 -13001.276 0.230 0.214
Chain 1: 1500 -10390.809 0.247 0.238
Chain 1: 1600 -12259.378 0.176 0.214
Chain 1: 1700 -24917.389 0.197 0.214
Chain 1: 1800 -12884.552 0.266 0.214
Chain 1: 1900 -10770.773 0.264 0.196
Chain 1: 2000 -12763.336 0.271 0.196
Chain 1: 2100 -9643.616 0.301 0.238
Chain 1: 2200 -10295.403 0.291 0.238
Chain 1: 2300 -14933.182 0.313 0.251
Chain 1: 2400 -11130.937 0.324 0.311
Chain 1: 2500 -13443.522 0.316 0.311
Chain 1: 2600 -9833.968 0.337 0.324
Chain 1: 2700 -9167.655 0.294 0.311
Chain 1: 2800 -9410.883 0.203 0.196
Chain 1: 2900 -9118.537 0.186 0.172
Chain 1: 3000 -17260.114 0.218 0.311
Chain 1: 3100 -12513.515 0.224 0.311
Chain 1: 3200 -16027.112 0.239 0.311
Chain 1: 3300 -9222.240 0.282 0.342
Chain 1: 3400 -15645.049 0.289 0.367
Chain 1: 3500 -9083.200 0.344 0.379
Chain 1: 3600 -9022.498 0.308 0.379
Chain 1: 3700 -8797.162 0.303 0.379
Chain 1: 3800 -9238.949 0.305 0.379
Chain 1: 3900 -9327.524 0.303 0.379
Chain 1: 4000 -15811.199 0.297 0.379
Chain 1: 4100 -9703.814 0.322 0.410
Chain 1: 4200 -9072.586 0.307 0.410
Chain 1: 4300 -10062.505 0.243 0.098
Chain 1: 4400 -9424.017 0.209 0.070
Chain 1: 4500 -8453.949 0.148 0.070
Chain 1: 4600 -11846.661 0.176 0.098
Chain 1: 4700 -12831.240 0.181 0.098
Chain 1: 4800 -10637.867 0.197 0.115
Chain 1: 4900 -9627.936 0.206 0.115
Chain 1: 5000 -14702.954 0.200 0.115
Chain 1: 5100 -8639.767 0.207 0.115
Chain 1: 5200 -9185.753 0.206 0.115
Chain 1: 5300 -13298.069 0.227 0.206
Chain 1: 5400 -9941.901 0.254 0.286
Chain 1: 5500 -11883.914 0.259 0.286
Chain 1: 5600 -12839.098 0.238 0.206
Chain 1: 5700 -13016.601 0.232 0.206
Chain 1: 5800 -8762.274 0.260 0.309
Chain 1: 5900 -13052.481 0.282 0.329
Chain 1: 6000 -10727.543 0.269 0.309
Chain 1: 6100 -8606.648 0.224 0.246
Chain 1: 6200 -10451.588 0.235 0.246
Chain 1: 6300 -8718.146 0.224 0.217
Chain 1: 6400 -9061.404 0.194 0.199
Chain 1: 6500 -10504.890 0.192 0.199
Chain 1: 6600 -8425.855 0.209 0.217
Chain 1: 6700 -9331.920 0.217 0.217
Chain 1: 6800 -11804.163 0.190 0.209
Chain 1: 6900 -10599.826 0.168 0.199
Chain 1: 7000 -8566.970 0.170 0.199
Chain 1: 7100 -8339.867 0.148 0.177
Chain 1: 7200 -10943.804 0.154 0.199
Chain 1: 7300 -9361.182 0.151 0.169
Chain 1: 7400 -8335.708 0.160 0.169
Chain 1: 7500 -9456.961 0.158 0.169
Chain 1: 7600 -12041.959 0.155 0.169
Chain 1: 7700 -9720.824 0.169 0.209
Chain 1: 7800 -10281.377 0.153 0.169
Chain 1: 7900 -8526.415 0.163 0.206
Chain 1: 8000 -10108.710 0.155 0.169
Chain 1: 8100 -8455.739 0.171 0.195
Chain 1: 8200 -8954.304 0.153 0.169
Chain 1: 8300 -8307.536 0.144 0.157
Chain 1: 8400 -8206.239 0.133 0.157
Chain 1: 8500 -8545.430 0.125 0.157
Chain 1: 8600 -9925.059 0.118 0.139
Chain 1: 8700 -8532.153 0.110 0.139
Chain 1: 8800 -8748.000 0.107 0.139
Chain 1: 8900 -9149.865 0.091 0.078
Chain 1: 9000 -8674.842 0.081 0.056
Chain 1: 9100 -9658.942 0.071 0.056
Chain 1: 9200 -10027.001 0.069 0.055
Chain 1: 9300 -12728.336 0.083 0.055
Chain 1: 9400 -9651.685 0.113 0.102
Chain 1: 9500 -8142.707 0.128 0.139
Chain 1: 9600 -9488.077 0.128 0.142
Chain 1: 9700 -8375.685 0.125 0.133
Chain 1: 9800 -8618.977 0.126 0.133
Chain 1: 9900 -8648.042 0.122 0.133
Chain 1: 10000 -8593.488 0.117 0.133
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001406 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.06 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57894.773 1.000 1.000
Chain 1: 200 -17573.124 1.647 2.295
Chain 1: 300 -8686.376 1.439 1.023
Chain 1: 400 -8210.893 1.094 1.023
Chain 1: 500 -8350.715 0.878 1.000
Chain 1: 600 -8545.791 0.736 1.000
Chain 1: 700 -7800.881 0.644 0.095
Chain 1: 800 -8194.162 0.570 0.095
Chain 1: 900 -8009.499 0.509 0.058
Chain 1: 1000 -7885.516 0.460 0.058
Chain 1: 1100 -7859.807 0.360 0.048
Chain 1: 1200 -7654.060 0.133 0.027
Chain 1: 1300 -7834.897 0.033 0.023
Chain 1: 1400 -7756.780 0.029 0.023
Chain 1: 1500 -7676.144 0.028 0.023
Chain 1: 1600 -7640.613 0.026 0.023
Chain 1: 1700 -7607.548 0.017 0.016
Chain 1: 1800 -7670.056 0.013 0.011
Chain 1: 1900 -7640.965 0.011 0.010
Chain 1: 2000 -7661.744 0.010 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003134 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.34 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86703.560 1.000 1.000
Chain 1: 200 -13465.346 3.220 5.439
Chain 1: 300 -9912.601 2.266 1.000
Chain 1: 400 -10832.298 1.721 1.000
Chain 1: 500 -8828.039 1.422 0.358
Chain 1: 600 -8510.129 1.191 0.358
Chain 1: 700 -8818.236 1.026 0.227
Chain 1: 800 -9328.134 0.905 0.227
Chain 1: 900 -8778.073 0.811 0.085
Chain 1: 1000 -8527.494 0.733 0.085
Chain 1: 1100 -8801.861 0.636 0.063 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8354.463 0.097 0.055
Chain 1: 1300 -8617.367 0.065 0.054
Chain 1: 1400 -8644.291 0.056 0.037
Chain 1: 1500 -8532.752 0.035 0.035
Chain 1: 1600 -8635.323 0.032 0.031
Chain 1: 1700 -8721.908 0.030 0.031
Chain 1: 1800 -8333.970 0.029 0.031
Chain 1: 1900 -8436.367 0.024 0.029
Chain 1: 2000 -8406.651 0.022 0.013
Chain 1: 2100 -8537.573 0.020 0.013
Chain 1: 2200 -8323.659 0.017 0.013
Chain 1: 2300 -8465.778 0.016 0.013
Chain 1: 2400 -8478.544 0.016 0.013
Chain 1: 2500 -8446.461 0.015 0.012
Chain 1: 2600 -8446.655 0.014 0.012
Chain 1: 2700 -8354.629 0.014 0.012
Chain 1: 2800 -8330.249 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003372 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.72 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8417441.879 1.000 1.000
Chain 1: 200 -1587975.050 2.650 4.301
Chain 1: 300 -890448.686 2.028 1.000
Chain 1: 400 -457152.449 1.758 1.000
Chain 1: 500 -357239.705 1.462 0.948
Chain 1: 600 -232205.806 1.308 0.948
Chain 1: 700 -118777.622 1.258 0.948
Chain 1: 800 -86098.730 1.148 0.948
Chain 1: 900 -66507.504 1.053 0.783
Chain 1: 1000 -51356.646 0.977 0.783
Chain 1: 1100 -38890.286 0.909 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38068.635 0.482 0.380
Chain 1: 1300 -26089.318 0.449 0.380
Chain 1: 1400 -25812.110 0.355 0.321
Chain 1: 1500 -22416.962 0.343 0.321
Chain 1: 1600 -21638.161 0.292 0.295
Chain 1: 1700 -20520.033 0.202 0.295
Chain 1: 1800 -20465.833 0.165 0.151
Chain 1: 1900 -20791.530 0.137 0.054
Chain 1: 2000 -19307.944 0.115 0.054
Chain 1: 2100 -19545.944 0.084 0.036
Chain 1: 2200 -19771.456 0.083 0.036
Chain 1: 2300 -19389.599 0.039 0.020
Chain 1: 2400 -19161.957 0.039 0.020
Chain 1: 2500 -18963.762 0.025 0.016
Chain 1: 2600 -18594.703 0.024 0.016
Chain 1: 2700 -18551.888 0.018 0.012
Chain 1: 2800 -18268.911 0.020 0.015
Chain 1: 2900 -18549.820 0.020 0.015
Chain 1: 3000 -18536.056 0.012 0.012
Chain 1: 3100 -18620.977 0.011 0.012
Chain 1: 3200 -18312.060 0.012 0.015
Chain 1: 3300 -18516.470 0.011 0.012
Chain 1: 3400 -17992.038 0.013 0.015
Chain 1: 3500 -18602.882 0.015 0.015
Chain 1: 3600 -17910.852 0.017 0.015
Chain 1: 3700 -18296.665 0.019 0.017
Chain 1: 3800 -17258.361 0.023 0.021
Chain 1: 3900 -17254.517 0.022 0.021
Chain 1: 4000 -17371.843 0.022 0.021
Chain 1: 4100 -17285.722 0.022 0.021
Chain 1: 4200 -17102.379 0.022 0.021
Chain 1: 4300 -17240.496 0.021 0.021
Chain 1: 4400 -17197.674 0.019 0.011
Chain 1: 4500 -17100.244 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001381 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.81 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12418.644 1.000 1.000
Chain 1: 200 -9347.060 0.664 1.000
Chain 1: 300 -7943.122 0.502 0.329
Chain 1: 400 -8198.292 0.384 0.329
Chain 1: 500 -8093.079 0.310 0.177
Chain 1: 600 -7923.321 0.262 0.177
Chain 1: 700 -7823.776 0.226 0.031
Chain 1: 800 -7830.607 0.198 0.031
Chain 1: 900 -7749.992 0.177 0.021
Chain 1: 1000 -7945.160 0.162 0.025
Chain 1: 1100 -7974.289 0.062 0.021
Chain 1: 1200 -7858.451 0.031 0.015
Chain 1: 1300 -7788.183 0.014 0.013
Chain 1: 1400 -7813.896 0.011 0.013
Chain 1: 1500 -7903.315 0.011 0.011
Chain 1: 1600 -7849.144 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001383 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.83 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57226.864 1.000 1.000
Chain 1: 200 -17583.818 1.627 2.255
Chain 1: 300 -8784.219 1.419 1.002
Chain 1: 400 -8293.067 1.079 1.002
Chain 1: 500 -8303.297 0.863 1.000
Chain 1: 600 -8985.057 0.732 1.000
Chain 1: 700 -7709.843 0.651 0.165
Chain 1: 800 -8346.006 0.579 0.165
Chain 1: 900 -7900.002 0.521 0.076
Chain 1: 1000 -8005.563 0.470 0.076
Chain 1: 1100 -7688.950 0.375 0.076
Chain 1: 1200 -7715.756 0.149 0.059
Chain 1: 1300 -7770.806 0.050 0.056
Chain 1: 1400 -7917.719 0.046 0.041
Chain 1: 1500 -7576.846 0.050 0.045
Chain 1: 1600 -7777.179 0.045 0.041
Chain 1: 1700 -7544.858 0.032 0.031
Chain 1: 1800 -7657.124 0.026 0.026
Chain 1: 1900 -7648.312 0.020 0.019
Chain 1: 2000 -7661.654 0.019 0.019
Chain 1: 2100 -7588.275 0.016 0.015
Chain 1: 2200 -7725.175 0.017 0.018
Chain 1: 2300 -7614.124 0.018 0.018
Chain 1: 2400 -7663.781 0.017 0.015
Chain 1: 2500 -7606.791 0.013 0.015
Chain 1: 2600 -7535.118 0.011 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002508 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.08 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86368.835 1.000 1.000
Chain 1: 200 -13545.349 3.188 5.376
Chain 1: 300 -9851.892 2.250 1.000
Chain 1: 400 -11206.108 1.718 1.000
Chain 1: 500 -8830.625 1.428 0.375
Chain 1: 600 -8789.312 1.191 0.375
Chain 1: 700 -8721.409 1.022 0.269
Chain 1: 800 -8920.034 0.897 0.269
Chain 1: 900 -8619.968 0.801 0.121
Chain 1: 1000 -8662.354 0.722 0.121
Chain 1: 1100 -8545.347 0.623 0.035 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8273.931 0.089 0.033
Chain 1: 1300 -8520.809 0.054 0.029
Chain 1: 1400 -8552.799 0.042 0.022
Chain 1: 1500 -8394.399 0.017 0.019
Chain 1: 1600 -8509.584 0.018 0.019
Chain 1: 1700 -8579.896 0.018 0.019
Chain 1: 1800 -8150.301 0.021 0.019
Chain 1: 1900 -8254.081 0.019 0.014
Chain 1: 2000 -8229.169 0.019 0.014
Chain 1: 2100 -8360.601 0.019 0.016
Chain 1: 2200 -8156.870 0.018 0.016
Chain 1: 2300 -8252.004 0.016 0.014
Chain 1: 2400 -8317.428 0.017 0.014
Chain 1: 2500 -8262.742 0.016 0.013
Chain 1: 2600 -8266.592 0.014 0.012
Chain 1: 2700 -8182.084 0.015 0.012
Chain 1: 2800 -8139.212 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003228 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.28 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8405314.175 1.000 1.000
Chain 1: 200 -1585471.930 2.651 4.301
Chain 1: 300 -892271.123 2.026 1.000
Chain 1: 400 -458753.759 1.756 1.000
Chain 1: 500 -358913.364 1.460 0.945
Chain 1: 600 -233712.986 1.306 0.945
Chain 1: 700 -119590.028 1.256 0.945
Chain 1: 800 -86694.132 1.146 0.945
Chain 1: 900 -66985.336 1.052 0.777
Chain 1: 1000 -51740.076 0.976 0.777
Chain 1: 1100 -39179.073 0.908 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38351.289 0.480 0.379
Chain 1: 1300 -26273.674 0.448 0.379
Chain 1: 1400 -25990.103 0.355 0.321
Chain 1: 1500 -22567.923 0.342 0.321
Chain 1: 1600 -21781.477 0.292 0.295
Chain 1: 1700 -20651.408 0.202 0.294
Chain 1: 1800 -20594.624 0.165 0.152
Chain 1: 1900 -20920.921 0.137 0.055
Chain 1: 2000 -19429.313 0.115 0.055
Chain 1: 2100 -19668.040 0.084 0.036
Chain 1: 2200 -19894.905 0.083 0.036
Chain 1: 2300 -19511.652 0.039 0.020
Chain 1: 2400 -19283.608 0.039 0.020
Chain 1: 2500 -19085.641 0.025 0.016
Chain 1: 2600 -18715.712 0.023 0.016
Chain 1: 2700 -18672.485 0.018 0.012
Chain 1: 2800 -18389.303 0.020 0.015
Chain 1: 2900 -18670.662 0.019 0.015
Chain 1: 3000 -18656.873 0.012 0.012
Chain 1: 3100 -18741.922 0.011 0.012
Chain 1: 3200 -18432.439 0.012 0.015
Chain 1: 3300 -18637.222 0.011 0.012
Chain 1: 3400 -18111.914 0.012 0.015
Chain 1: 3500 -18724.202 0.015 0.015
Chain 1: 3600 -18030.286 0.017 0.015
Chain 1: 3700 -18417.600 0.018 0.017
Chain 1: 3800 -17376.430 0.023 0.021
Chain 1: 3900 -17372.510 0.021 0.021
Chain 1: 4000 -17489.837 0.022 0.021
Chain 1: 4100 -17403.609 0.022 0.021
Chain 1: 4200 -17219.564 0.021 0.021
Chain 1: 4300 -17358.162 0.021 0.021
Chain 1: 4400 -17314.846 0.019 0.011
Chain 1: 4500 -17217.312 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001384 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.84 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12828.148 1.000 1.000
Chain 1: 200 -9659.069 0.664 1.000
Chain 1: 300 -8117.517 0.506 0.328
Chain 1: 400 -8378.461 0.387 0.328
Chain 1: 500 -8192.507 0.314 0.190
Chain 1: 600 -8089.270 0.264 0.190
Chain 1: 700 -7978.941 0.228 0.031
Chain 1: 800 -7978.855 0.200 0.031
Chain 1: 900 -7910.346 0.179 0.023
Chain 1: 1000 -8110.343 0.163 0.025
Chain 1: 1100 -8126.854 0.063 0.023
Chain 1: 1200 -7998.220 0.032 0.016
Chain 1: 1300 -7969.682 0.014 0.014
Chain 1: 1400 -7972.790 0.010 0.013
Chain 1: 1500 -8065.737 0.009 0.012 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001558 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.58 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -47795.882 1.000 1.000
Chain 1: 200 -16155.810 1.479 1.958
Chain 1: 300 -8670.191 1.274 1.000
Chain 1: 400 -8482.408 0.961 1.000
Chain 1: 500 -8527.893 0.770 0.863
Chain 1: 600 -8957.221 0.650 0.863
Chain 1: 700 -8287.872 0.568 0.081
Chain 1: 800 -8107.573 0.500 0.081
Chain 1: 900 -7972.226 0.446 0.048
Chain 1: 1000 -7839.252 0.403 0.048
Chain 1: 1100 -7613.852 0.306 0.030
Chain 1: 1200 -7812.562 0.113 0.025
Chain 1: 1300 -7672.680 0.029 0.022
Chain 1: 1400 -7675.887 0.026 0.022
Chain 1: 1500 -7519.470 0.028 0.022
Chain 1: 1600 -7704.747 0.026 0.022
Chain 1: 1700 -7513.637 0.020 0.022
Chain 1: 1800 -7604.712 0.019 0.021
Chain 1: 1900 -7582.915 0.018 0.021
Chain 1: 2000 -7551.693 0.016 0.021
Chain 1: 2100 -7489.815 0.014 0.018
Chain 1: 2200 -7677.049 0.014 0.018
Chain 1: 2300 -7505.142 0.015 0.021
Chain 1: 2400 -7479.462 0.015 0.021
Chain 1: 2500 -7476.549 0.013 0.012
Chain 1: 2600 -7459.057 0.011 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00296 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.6 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87294.571 1.000 1.000
Chain 1: 200 -13814.017 3.160 5.319
Chain 1: 300 -10086.570 2.230 1.000
Chain 1: 400 -11379.356 1.701 1.000
Chain 1: 500 -9034.503 1.412 0.370
Chain 1: 600 -8986.447 1.178 0.370
Chain 1: 700 -9153.955 1.012 0.260
Chain 1: 800 -8379.642 0.897 0.260
Chain 1: 900 -8422.186 0.798 0.114
Chain 1: 1000 -8672.639 0.721 0.114
Chain 1: 1100 -8680.533 0.621 0.092 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8484.562 0.092 0.029
Chain 1: 1300 -8752.246 0.058 0.029
Chain 1: 1400 -8724.741 0.047 0.023
Chain 1: 1500 -8590.801 0.022 0.018
Chain 1: 1600 -8701.737 0.023 0.018
Chain 1: 1700 -8762.963 0.022 0.016
Chain 1: 1800 -8324.721 0.018 0.016
Chain 1: 1900 -8428.810 0.019 0.016
Chain 1: 2000 -8407.450 0.016 0.013
Chain 1: 2100 -8391.940 0.016 0.013
Chain 1: 2200 -8346.826 0.014 0.012
Chain 1: 2300 -8481.916 0.013 0.012
Chain 1: 2400 -8327.147 0.014 0.013
Chain 1: 2500 -8398.251 0.014 0.012
Chain 1: 2600 -8310.864 0.014 0.011
Chain 1: 2700 -8348.369 0.013 0.011
Chain 1: 2800 -8306.315 0.009 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003497 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.97 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8397846.750 1.000 1.000
Chain 1: 200 -1583709.544 2.651 4.303
Chain 1: 300 -890240.884 2.027 1.000
Chain 1: 400 -456884.884 1.758 1.000
Chain 1: 500 -357433.125 1.462 0.949
Chain 1: 600 -232666.921 1.307 0.949
Chain 1: 700 -119297.712 1.256 0.949
Chain 1: 800 -86559.856 1.147 0.949
Chain 1: 900 -66975.775 1.052 0.779
Chain 1: 1000 -51825.197 0.976 0.779
Chain 1: 1100 -39336.810 0.908 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38525.497 0.479 0.378
Chain 1: 1300 -26509.974 0.447 0.378
Chain 1: 1400 -26234.238 0.353 0.317
Chain 1: 1500 -22827.185 0.340 0.317
Chain 1: 1600 -22045.994 0.290 0.292
Chain 1: 1700 -20922.656 0.200 0.292
Chain 1: 1800 -20867.865 0.163 0.149
Chain 1: 1900 -21194.569 0.135 0.054
Chain 1: 2000 -19706.071 0.113 0.054
Chain 1: 2100 -19944.647 0.083 0.035
Chain 1: 2200 -20171.034 0.082 0.035
Chain 1: 2300 -19788.152 0.038 0.019
Chain 1: 2400 -19560.095 0.039 0.019
Chain 1: 2500 -19361.797 0.025 0.015
Chain 1: 2600 -18991.822 0.023 0.015
Chain 1: 2700 -18948.771 0.018 0.012
Chain 1: 2800 -18665.241 0.019 0.015
Chain 1: 2900 -18946.717 0.019 0.015
Chain 1: 3000 -18932.969 0.012 0.012
Chain 1: 3100 -19017.965 0.011 0.012
Chain 1: 3200 -18708.420 0.011 0.015
Chain 1: 3300 -18913.354 0.011 0.012
Chain 1: 3400 -18387.684 0.012 0.015
Chain 1: 3500 -19000.338 0.015 0.015
Chain 1: 3600 -18306.058 0.016 0.015
Chain 1: 3700 -18693.531 0.018 0.017
Chain 1: 3800 -17651.612 0.023 0.021
Chain 1: 3900 -17647.674 0.021 0.021
Chain 1: 4000 -17765.038 0.022 0.021
Chain 1: 4100 -17678.653 0.022 0.021
Chain 1: 4200 -17494.581 0.021 0.021
Chain 1: 4300 -17633.261 0.021 0.021
Chain 1: 4400 -17589.819 0.018 0.011
Chain 1: 4500 -17492.246 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001376 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.76 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12194.225 1.000 1.000
Chain 1: 200 -9118.634 0.669 1.000
Chain 1: 300 -8118.910 0.487 0.337
Chain 1: 400 -8232.782 0.369 0.337
Chain 1: 500 -8045.094 0.300 0.123
Chain 1: 600 -7938.203 0.252 0.123
Chain 1: 700 -7892.897 0.217 0.023
Chain 1: 800 -7867.970 0.190 0.023
Chain 1: 900 -7965.606 0.170 0.014
Chain 1: 1000 -7947.597 0.153 0.014
Chain 1: 1100 -8002.704 0.054 0.013
Chain 1: 1200 -7906.030 0.022 0.012
Chain 1: 1300 -7929.898 0.010 0.012 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001411 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.11 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56688.795 1.000 1.000
Chain 1: 200 -17085.273 1.659 2.318
Chain 1: 300 -8614.036 1.434 1.000
Chain 1: 400 -7894.703 1.098 1.000
Chain 1: 500 -8396.422 0.890 0.983
Chain 1: 600 -8989.224 0.753 0.983
Chain 1: 700 -7781.379 0.668 0.155
Chain 1: 800 -8051.458 0.588 0.155
Chain 1: 900 -7949.574 0.524 0.091
Chain 1: 1000 -7879.502 0.473 0.091
Chain 1: 1100 -7694.943 0.375 0.066
Chain 1: 1200 -7593.929 0.145 0.060
Chain 1: 1300 -7588.999 0.047 0.034
Chain 1: 1400 -7896.988 0.041 0.034
Chain 1: 1500 -7598.338 0.039 0.034
Chain 1: 1600 -7565.224 0.033 0.024
Chain 1: 1700 -7499.653 0.018 0.013
Chain 1: 1800 -7572.927 0.016 0.013
Chain 1: 1900 -7608.864 0.015 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002575 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.75 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86209.556 1.000 1.000
Chain 1: 200 -13245.703 3.254 5.508
Chain 1: 300 -9726.833 2.290 1.000
Chain 1: 400 -10614.093 1.738 1.000
Chain 1: 500 -8597.093 1.438 0.362
Chain 1: 600 -8333.839 1.203 0.362
Chain 1: 700 -8386.922 1.032 0.235
Chain 1: 800 -8800.773 0.909 0.235
Chain 1: 900 -8606.098 0.811 0.084
Chain 1: 1000 -8379.563 0.732 0.084
Chain 1: 1100 -8623.677 0.635 0.047 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8286.266 0.088 0.041
Chain 1: 1300 -8339.295 0.053 0.032
Chain 1: 1400 -8335.132 0.045 0.028
Chain 1: 1500 -8366.439 0.021 0.027
Chain 1: 1600 -8371.427 0.018 0.023
Chain 1: 1700 -8310.240 0.018 0.023
Chain 1: 1800 -8189.956 0.015 0.015
Chain 1: 1900 -8304.625 0.014 0.014
Chain 1: 2000 -8265.482 0.012 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003193 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.93 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8410058.862 1.000 1.000
Chain 1: 200 -1585922.608 2.651 4.303
Chain 1: 300 -891127.728 2.028 1.000
Chain 1: 400 -457688.106 1.757 1.000
Chain 1: 500 -357731.525 1.462 0.947
Chain 1: 600 -232502.713 1.308 0.947
Chain 1: 700 -118808.994 1.258 0.947
Chain 1: 800 -86052.768 1.148 0.947
Chain 1: 900 -66408.390 1.053 0.780
Chain 1: 1000 -51218.297 0.978 0.780
Chain 1: 1100 -38716.780 0.910 0.539 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37889.063 0.482 0.381
Chain 1: 1300 -25878.339 0.450 0.381
Chain 1: 1400 -25596.748 0.357 0.323
Chain 1: 1500 -22193.614 0.344 0.323
Chain 1: 1600 -21412.250 0.294 0.297
Chain 1: 1700 -20290.574 0.204 0.296
Chain 1: 1800 -20235.445 0.166 0.153
Chain 1: 1900 -20560.827 0.138 0.055
Chain 1: 2000 -19076.010 0.116 0.055
Chain 1: 2100 -19313.952 0.085 0.036
Chain 1: 2200 -19539.632 0.084 0.036
Chain 1: 2300 -19157.733 0.040 0.020
Chain 1: 2400 -18930.170 0.040 0.020
Chain 1: 2500 -18732.137 0.025 0.016
Chain 1: 2600 -18363.108 0.024 0.016
Chain 1: 2700 -18320.347 0.019 0.012
Chain 1: 2800 -18037.546 0.020 0.016
Chain 1: 2900 -18318.384 0.020 0.015
Chain 1: 3000 -18304.597 0.012 0.012
Chain 1: 3100 -18389.492 0.011 0.012
Chain 1: 3200 -18080.683 0.012 0.015
Chain 1: 3300 -18285.022 0.011 0.012
Chain 1: 3400 -17760.861 0.013 0.015
Chain 1: 3500 -18371.339 0.015 0.016
Chain 1: 3600 -17679.842 0.017 0.016
Chain 1: 3700 -18065.264 0.019 0.017
Chain 1: 3800 -17027.818 0.023 0.021
Chain 1: 3900 -17024.052 0.022 0.021
Chain 1: 4000 -17141.333 0.022 0.021
Chain 1: 4100 -17055.241 0.022 0.021
Chain 1: 4200 -16872.138 0.022 0.021
Chain 1: 4300 -17010.074 0.022 0.021
Chain 1: 4400 -16967.400 0.019 0.011
Chain 1: 4500 -16870.048 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001357 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.57 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49704.473 1.000 1.000
Chain 1: 200 -14742.122 1.686 2.372
Chain 1: 300 -15526.521 1.141 1.000
Chain 1: 400 -14130.779 0.880 1.000
Chain 1: 500 -18825.427 0.754 0.249
Chain 1: 600 -17282.726 0.643 0.249
Chain 1: 700 -14807.669 0.575 0.167
Chain 1: 800 -15102.556 0.506 0.167
Chain 1: 900 -12793.688 0.470 0.167
Chain 1: 1000 -13294.777 0.426 0.167
Chain 1: 1100 -12005.367 0.337 0.107
Chain 1: 1200 -13890.583 0.114 0.107
Chain 1: 1300 -11478.030 0.130 0.136
Chain 1: 1400 -27963.984 0.179 0.167
Chain 1: 1500 -12643.716 0.275 0.167
Chain 1: 1600 -12766.647 0.267 0.167
Chain 1: 1700 -12101.050 0.256 0.136
Chain 1: 1800 -17730.430 0.285 0.180
Chain 1: 1900 -11861.725 0.317 0.210
Chain 1: 2000 -24080.342 0.364 0.317
Chain 1: 2100 -20058.368 0.373 0.317
Chain 1: 2200 -11639.232 0.432 0.495
Chain 1: 2300 -10418.862 0.423 0.495
Chain 1: 2400 -11437.016 0.373 0.317
Chain 1: 2500 -13756.833 0.268 0.201
Chain 1: 2600 -9718.469 0.309 0.317
Chain 1: 2700 -9687.110 0.304 0.317
Chain 1: 2800 -12117.331 0.292 0.201
Chain 1: 2900 -9762.643 0.267 0.201
Chain 1: 3000 -13318.640 0.243 0.201
Chain 1: 3100 -10589.728 0.248 0.241
Chain 1: 3200 -15707.548 0.209 0.241
Chain 1: 3300 -14510.728 0.205 0.241
Chain 1: 3400 -17293.265 0.212 0.241
Chain 1: 3500 -9950.284 0.269 0.258
Chain 1: 3600 -10854.042 0.236 0.241
Chain 1: 3700 -12551.547 0.249 0.241
Chain 1: 3800 -11252.391 0.241 0.241
Chain 1: 3900 -9429.971 0.236 0.193
Chain 1: 4000 -9175.981 0.212 0.161
Chain 1: 4100 -10178.882 0.196 0.135
Chain 1: 4200 -13927.627 0.190 0.135
Chain 1: 4300 -10393.139 0.216 0.161
Chain 1: 4400 -13547.826 0.223 0.193
Chain 1: 4500 -9201.214 0.197 0.193
Chain 1: 4600 -15164.964 0.228 0.233
Chain 1: 4700 -9323.300 0.277 0.269
Chain 1: 4800 -9562.950 0.268 0.269
Chain 1: 4900 -10380.650 0.256 0.269
Chain 1: 5000 -16405.063 0.290 0.340
Chain 1: 5100 -9214.487 0.359 0.367
Chain 1: 5200 -10048.130 0.340 0.367
Chain 1: 5300 -10223.622 0.308 0.367
Chain 1: 5400 -9804.038 0.289 0.367
Chain 1: 5500 -9924.073 0.243 0.083
Chain 1: 5600 -14330.387 0.234 0.083
Chain 1: 5700 -11550.788 0.195 0.083
Chain 1: 5800 -9180.190 0.219 0.241
Chain 1: 5900 -10121.496 0.220 0.241
Chain 1: 6000 -13352.376 0.208 0.241
Chain 1: 6100 -9493.014 0.170 0.241
Chain 1: 6200 -10040.090 0.167 0.241
Chain 1: 6300 -11718.720 0.180 0.241
Chain 1: 6400 -9344.581 0.201 0.242
Chain 1: 6500 -9449.438 0.201 0.242
Chain 1: 6600 -9364.256 0.171 0.241
Chain 1: 6700 -10079.730 0.154 0.143
Chain 1: 6800 -9582.796 0.134 0.093
Chain 1: 6900 -12642.319 0.149 0.143
Chain 1: 7000 -9100.926 0.163 0.143
Chain 1: 7100 -9664.299 0.128 0.071
Chain 1: 7200 -12303.318 0.144 0.143
Chain 1: 7300 -11673.026 0.136 0.071
Chain 1: 7400 -8689.426 0.144 0.071
Chain 1: 7500 -10734.502 0.162 0.191
Chain 1: 7600 -9318.533 0.177 0.191
Chain 1: 7700 -9580.067 0.172 0.191
Chain 1: 7800 -8713.305 0.177 0.191
Chain 1: 7900 -10964.553 0.173 0.191
Chain 1: 8000 -9328.875 0.152 0.175
Chain 1: 8100 -10935.562 0.161 0.175
Chain 1: 8200 -9970.558 0.149 0.152
Chain 1: 8300 -8807.925 0.157 0.152
Chain 1: 8400 -8578.484 0.125 0.147
Chain 1: 8500 -8844.040 0.109 0.132
Chain 1: 8600 -8928.839 0.095 0.099
Chain 1: 8700 -9164.717 0.095 0.099
Chain 1: 8800 -9478.421 0.088 0.097
Chain 1: 8900 -10884.214 0.081 0.097
Chain 1: 9000 -10451.648 0.067 0.041
Chain 1: 9100 -8818.889 0.071 0.041
Chain 1: 9200 -8664.150 0.063 0.033
Chain 1: 9300 -8616.959 0.050 0.030
Chain 1: 9400 -8634.484 0.048 0.030
Chain 1: 9500 -8708.918 0.046 0.026
Chain 1: 9600 -8789.136 0.046 0.026
Chain 1: 9700 -8576.538 0.046 0.025
Chain 1: 9800 -10163.067 0.058 0.025
Chain 1: 9900 -10088.226 0.046 0.018
Chain 1: 10000 -8758.457 0.057 0.018
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001416 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.16 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58850.017 1.000 1.000
Chain 1: 200 -18442.432 1.596 2.191
Chain 1: 300 -9045.401 1.410 1.039
Chain 1: 400 -8187.494 1.084 1.039
Chain 1: 500 -8481.405 0.874 1.000
Chain 1: 600 -8517.133 0.729 1.000
Chain 1: 700 -8514.012 0.625 0.105
Chain 1: 800 -8496.654 0.547 0.105
Chain 1: 900 -7816.699 0.496 0.087
Chain 1: 1000 -8207.247 0.451 0.087
Chain 1: 1100 -7728.540 0.357 0.062
Chain 1: 1200 -7753.988 0.138 0.048
Chain 1: 1300 -7773.790 0.035 0.035
Chain 1: 1400 -7998.697 0.027 0.028
Chain 1: 1500 -7724.825 0.027 0.028
Chain 1: 1600 -7942.976 0.030 0.028
Chain 1: 1700 -7742.813 0.032 0.028
Chain 1: 1800 -7728.320 0.032 0.028
Chain 1: 1900 -7704.300 0.024 0.027
Chain 1: 2000 -7775.921 0.020 0.026
Chain 1: 2100 -7662.274 0.015 0.015
Chain 1: 2200 -7874.421 0.018 0.026
Chain 1: 2300 -7645.690 0.020 0.027
Chain 1: 2400 -7801.707 0.019 0.026
Chain 1: 2500 -7714.803 0.017 0.020
Chain 1: 2600 -7621.249 0.016 0.015
Chain 1: 2700 -7602.365 0.013 0.012
Chain 1: 2800 -7626.312 0.013 0.012
Chain 1: 2900 -7471.076 0.015 0.015
Chain 1: 3000 -7631.789 0.016 0.020
Chain 1: 3100 -7622.892 0.015 0.020
Chain 1: 3200 -7845.406 0.015 0.020
Chain 1: 3300 -7544.331 0.016 0.020
Chain 1: 3400 -7803.321 0.017 0.021
Chain 1: 3500 -7540.750 0.020 0.021
Chain 1: 3600 -7599.173 0.019 0.021
Chain 1: 3700 -7554.485 0.020 0.021
Chain 1: 3800 -7555.903 0.019 0.021
Chain 1: 3900 -7520.919 0.018 0.021
Chain 1: 4000 -7499.976 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002534 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.34 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87607.577 1.000 1.000
Chain 1: 200 -14148.127 3.096 5.192
Chain 1: 300 -10434.603 2.183 1.000
Chain 1: 400 -11744.826 1.665 1.000
Chain 1: 500 -9410.786 1.382 0.356
Chain 1: 600 -8825.417 1.162 0.356
Chain 1: 700 -8980.300 0.999 0.248
Chain 1: 800 -9571.662 0.882 0.248
Chain 1: 900 -9264.477 0.787 0.112
Chain 1: 1000 -9367.393 0.710 0.112
Chain 1: 1100 -9024.509 0.614 0.066 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8809.790 0.097 0.062
Chain 1: 1300 -9113.989 0.064 0.038
Chain 1: 1400 -9105.089 0.053 0.033
Chain 1: 1500 -8940.764 0.030 0.033
Chain 1: 1600 -9052.308 0.025 0.024
Chain 1: 1700 -9122.082 0.024 0.024
Chain 1: 1800 -8685.733 0.023 0.024
Chain 1: 1900 -8790.501 0.021 0.018
Chain 1: 2000 -8766.690 0.020 0.018
Chain 1: 2100 -8909.211 0.018 0.016
Chain 1: 2200 -8697.438 0.018 0.016
Chain 1: 2300 -8855.242 0.016 0.016
Chain 1: 2400 -8692.875 0.018 0.018
Chain 1: 2500 -8764.296 0.017 0.016
Chain 1: 2600 -8676.624 0.017 0.016
Chain 1: 2700 -8710.759 0.016 0.016
Chain 1: 2800 -8670.757 0.012 0.012
Chain 1: 2900 -8764.132 0.012 0.011
Chain 1: 3000 -8596.903 0.013 0.016
Chain 1: 3100 -8753.251 0.014 0.018
Chain 1: 3200 -8625.378 0.013 0.015
Chain 1: 3300 -8633.011 0.011 0.011
Chain 1: 3400 -8792.724 0.011 0.011
Chain 1: 3500 -8800.484 0.010 0.011
Chain 1: 3600 -8581.913 0.012 0.015
Chain 1: 3700 -8727.766 0.013 0.017
Chain 1: 3800 -8588.441 0.014 0.017
Chain 1: 3900 -8523.002 0.014 0.017
Chain 1: 4000 -8598.620 0.013 0.016
Chain 1: 4100 -8587.657 0.011 0.015
Chain 1: 4200 -8578.775 0.010 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002574 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.74 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8435163.616 1.000 1.000
Chain 1: 200 -1592761.025 2.648 4.296
Chain 1: 300 -893306.451 2.026 1.000
Chain 1: 400 -459101.852 1.756 1.000
Chain 1: 500 -358773.730 1.461 0.946
Chain 1: 600 -233446.722 1.307 0.946
Chain 1: 700 -119769.177 1.256 0.946
Chain 1: 800 -86970.440 1.146 0.946
Chain 1: 900 -67345.542 1.051 0.783
Chain 1: 1000 -52174.048 0.975 0.783
Chain 1: 1100 -39675.780 0.906 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38859.765 0.479 0.377
Chain 1: 1300 -26842.615 0.445 0.377
Chain 1: 1400 -26564.970 0.352 0.315
Chain 1: 1500 -23158.642 0.339 0.315
Chain 1: 1600 -22377.188 0.288 0.291
Chain 1: 1700 -21254.296 0.199 0.291
Chain 1: 1800 -21199.382 0.161 0.147
Chain 1: 1900 -21525.909 0.134 0.053
Chain 1: 2000 -20038.060 0.112 0.053
Chain 1: 2100 -20276.396 0.082 0.035
Chain 1: 2200 -20502.735 0.081 0.035
Chain 1: 2300 -20119.999 0.038 0.019
Chain 1: 2400 -19892.028 0.038 0.019
Chain 1: 2500 -19693.744 0.024 0.015
Chain 1: 2600 -19323.746 0.023 0.015
Chain 1: 2700 -19280.739 0.018 0.012
Chain 1: 2800 -18997.233 0.019 0.015
Chain 1: 2900 -19278.720 0.019 0.015
Chain 1: 3000 -19264.920 0.011 0.012
Chain 1: 3100 -19349.895 0.011 0.011
Chain 1: 3200 -19040.393 0.011 0.015
Chain 1: 3300 -19245.311 0.010 0.011
Chain 1: 3400 -18719.698 0.012 0.015
Chain 1: 3500 -19332.189 0.014 0.015
Chain 1: 3600 -18638.209 0.016 0.015
Chain 1: 3700 -19025.431 0.018 0.016
Chain 1: 3800 -17983.876 0.022 0.020
Chain 1: 3900 -17979.980 0.021 0.020
Chain 1: 4000 -18097.346 0.021 0.020
Chain 1: 4100 -18010.940 0.021 0.020
Chain 1: 4200 -17826.984 0.021 0.020
Chain 1: 4300 -17965.556 0.020 0.020
Chain 1: 4400 -17922.182 0.018 0.010
Chain 1: 4500 -17824.668 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001286 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.86 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48484.287 1.000 1.000
Chain 1: 200 -22357.807 1.084 1.169
Chain 1: 300 -11697.929 1.027 1.000
Chain 1: 400 -31566.176 0.927 1.000
Chain 1: 500 -13374.711 1.014 1.000
Chain 1: 600 -13281.695 0.846 1.000
Chain 1: 700 -13754.270 0.730 0.911
Chain 1: 800 -13720.234 0.639 0.911
Chain 1: 900 -14667.430 0.575 0.629
Chain 1: 1000 -16425.569 0.528 0.629
Chain 1: 1100 -12117.620 0.464 0.356
Chain 1: 1200 -13303.118 0.356 0.107
Chain 1: 1300 -25330.693 0.312 0.107
Chain 1: 1400 -9318.008 0.421 0.107
Chain 1: 1500 -10053.623 0.293 0.089
Chain 1: 1600 -9622.076 0.296 0.089
Chain 1: 1700 -24800.748 0.354 0.107
Chain 1: 1800 -10053.942 0.501 0.356 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1900 -11050.392 0.503 0.356 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2000 -11741.091 0.498 0.356
Chain 1: 2100 -9965.613 0.481 0.178
Chain 1: 2200 -9067.603 0.482 0.178
Chain 1: 2300 -9120.462 0.435 0.099
Chain 1: 2400 -10569.438 0.277 0.099
Chain 1: 2500 -9040.447 0.286 0.137
Chain 1: 2600 -8912.512 0.283 0.137
Chain 1: 2700 -10281.557 0.235 0.133
Chain 1: 2800 -9181.471 0.101 0.120
Chain 1: 2900 -9236.722 0.092 0.120
Chain 1: 3000 -8634.627 0.093 0.120
Chain 1: 3100 -9556.258 0.085 0.099
Chain 1: 3200 -10630.325 0.085 0.101
Chain 1: 3300 -8815.630 0.105 0.120
Chain 1: 3400 -9452.251 0.098 0.101
Chain 1: 3500 -8921.560 0.087 0.096
Chain 1: 3600 -12818.405 0.116 0.101
Chain 1: 3700 -9137.346 0.143 0.101
Chain 1: 3800 -13013.253 0.161 0.101
Chain 1: 3900 -8489.509 0.214 0.206
Chain 1: 4000 -8605.430 0.208 0.206
Chain 1: 4100 -8879.798 0.202 0.206
Chain 1: 4200 -12869.270 0.222 0.298
Chain 1: 4300 -9669.332 0.235 0.304
Chain 1: 4400 -9691.802 0.228 0.304
Chain 1: 4500 -9584.886 0.224 0.304
Chain 1: 4600 -8647.939 0.204 0.298
Chain 1: 4700 -13009.756 0.197 0.298
Chain 1: 4800 -8687.696 0.217 0.310
Chain 1: 4900 -9344.215 0.171 0.108
Chain 1: 5000 -10302.599 0.179 0.108
Chain 1: 5100 -9296.887 0.187 0.108
Chain 1: 5200 -8540.458 0.165 0.108
Chain 1: 5300 -8661.529 0.133 0.093
Chain 1: 5400 -9358.804 0.140 0.093
Chain 1: 5500 -8355.900 0.151 0.108
Chain 1: 5600 -12390.367 0.173 0.108
Chain 1: 5700 -13947.590 0.150 0.108
Chain 1: 5800 -8508.669 0.165 0.108
Chain 1: 5900 -14081.354 0.197 0.112
Chain 1: 6000 -9804.481 0.231 0.120
Chain 1: 6100 -10693.829 0.229 0.120
Chain 1: 6200 -8607.503 0.244 0.242
Chain 1: 6300 -12650.653 0.275 0.320
Chain 1: 6400 -11910.999 0.274 0.320
Chain 1: 6500 -8520.086 0.301 0.326
Chain 1: 6600 -8877.072 0.273 0.320
Chain 1: 6700 -10216.399 0.275 0.320
Chain 1: 6800 -9247.165 0.221 0.242
Chain 1: 6900 -11425.962 0.201 0.191
Chain 1: 7000 -8572.588 0.190 0.191
Chain 1: 7100 -7950.199 0.190 0.191
Chain 1: 7200 -8768.912 0.175 0.131
Chain 1: 7300 -9600.082 0.152 0.105
Chain 1: 7400 -8139.159 0.164 0.131
Chain 1: 7500 -8600.970 0.129 0.105
Chain 1: 7600 -9865.639 0.138 0.128
Chain 1: 7700 -9347.082 0.130 0.105
Chain 1: 7800 -8564.115 0.129 0.093
Chain 1: 7900 -8517.105 0.110 0.091
Chain 1: 8000 -8841.001 0.081 0.087
Chain 1: 8100 -7985.462 0.084 0.091
Chain 1: 8200 -11127.128 0.103 0.091
Chain 1: 8300 -8089.657 0.132 0.107
Chain 1: 8400 -8462.334 0.118 0.091
Chain 1: 8500 -10029.689 0.128 0.107
Chain 1: 8600 -8726.692 0.130 0.107
Chain 1: 8700 -8381.903 0.129 0.107
Chain 1: 8800 -8425.799 0.120 0.107
Chain 1: 8900 -8616.978 0.122 0.107
Chain 1: 9000 -10834.538 0.139 0.149
Chain 1: 9100 -8616.285 0.154 0.156
Chain 1: 9200 -8471.543 0.127 0.149
Chain 1: 9300 -10488.386 0.109 0.149
Chain 1: 9400 -8696.457 0.125 0.156
Chain 1: 9500 -8233.987 0.115 0.149
Chain 1: 9600 -9076.264 0.110 0.093
Chain 1: 9700 -8291.295 0.115 0.095
Chain 1: 9800 -9052.295 0.123 0.095
Chain 1: 9900 -8574.943 0.126 0.095
Chain 1: 10000 -8002.747 0.113 0.093
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001393 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.93 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57681.788 1.000 1.000
Chain 1: 200 -17376.127 1.660 2.320
Chain 1: 300 -8570.778 1.449 1.027
Chain 1: 400 -8174.746 1.099 1.027
Chain 1: 500 -8158.522 0.879 1.000
Chain 1: 600 -8441.244 0.738 1.000
Chain 1: 700 -7853.624 0.644 0.075
Chain 1: 800 -8019.214 0.566 0.075
Chain 1: 900 -7921.981 0.504 0.048
Chain 1: 1000 -7749.018 0.456 0.048
Chain 1: 1100 -7778.619 0.356 0.033
Chain 1: 1200 -7833.546 0.125 0.022
Chain 1: 1300 -7615.187 0.025 0.022
Chain 1: 1400 -7851.294 0.024 0.022
Chain 1: 1500 -7621.637 0.026 0.029
Chain 1: 1600 -7535.304 0.024 0.022
Chain 1: 1700 -7524.020 0.017 0.021
Chain 1: 1800 -7560.517 0.015 0.012
Chain 1: 1900 -7591.390 0.014 0.011
Chain 1: 2000 -7597.397 0.012 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003169 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.69 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86769.588 1.000 1.000
Chain 1: 200 -13198.980 3.287 5.574
Chain 1: 300 -9626.260 2.315 1.000
Chain 1: 400 -10382.873 1.754 1.000
Chain 1: 500 -8561.332 1.446 0.371
Chain 1: 600 -8484.438 1.207 0.371
Chain 1: 700 -8481.021 1.034 0.213
Chain 1: 800 -9020.849 0.913 0.213
Chain 1: 900 -8387.259 0.820 0.076
Chain 1: 1000 -8293.988 0.739 0.076
Chain 1: 1100 -8494.736 0.641 0.073 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8225.407 0.087 0.060
Chain 1: 1300 -8371.998 0.052 0.033
Chain 1: 1400 -8368.809 0.044 0.024
Chain 1: 1500 -8241.725 0.025 0.018
Chain 1: 1600 -8347.875 0.025 0.018
Chain 1: 1700 -8433.780 0.026 0.018
Chain 1: 1800 -8043.377 0.025 0.018
Chain 1: 1900 -8145.215 0.018 0.015
Chain 1: 2000 -8115.487 0.018 0.015
Chain 1: 2100 -8243.037 0.017 0.015
Chain 1: 2200 -8029.352 0.016 0.015
Chain 1: 2300 -8174.140 0.016 0.015
Chain 1: 2400 -8189.161 0.016 0.015
Chain 1: 2500 -8155.742 0.015 0.013
Chain 1: 2600 -8157.362 0.014 0.013
Chain 1: 2700 -8064.481 0.014 0.013
Chain 1: 2800 -8038.176 0.010 0.012 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003401 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.01 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8426209.599 1.000 1.000
Chain 1: 200 -1588055.106 2.653 4.306
Chain 1: 300 -891360.895 2.029 1.000
Chain 1: 400 -457630.331 1.759 1.000
Chain 1: 500 -357572.139 1.463 0.948
Chain 1: 600 -232370.043 1.309 0.948
Chain 1: 700 -118743.665 1.259 0.948
Chain 1: 800 -85965.726 1.149 0.948
Chain 1: 900 -66341.460 1.054 0.782
Chain 1: 1000 -51160.340 0.978 0.782
Chain 1: 1100 -38664.399 0.911 0.539 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37839.027 0.482 0.381
Chain 1: 1300 -25836.489 0.451 0.381
Chain 1: 1400 -25556.285 0.357 0.323
Chain 1: 1500 -22154.469 0.344 0.323
Chain 1: 1600 -21373.175 0.294 0.297
Chain 1: 1700 -20252.853 0.204 0.296
Chain 1: 1800 -20197.848 0.166 0.154
Chain 1: 1900 -20523.591 0.138 0.055
Chain 1: 2000 -19038.395 0.116 0.055
Chain 1: 2100 -19276.611 0.085 0.037
Chain 1: 2200 -19502.256 0.084 0.037
Chain 1: 2300 -19120.272 0.040 0.020
Chain 1: 2400 -18892.602 0.040 0.020
Chain 1: 2500 -18694.319 0.026 0.016
Chain 1: 2600 -18325.290 0.024 0.016
Chain 1: 2700 -18282.435 0.019 0.012
Chain 1: 2800 -17999.424 0.020 0.016
Chain 1: 2900 -18280.370 0.020 0.015
Chain 1: 3000 -18266.668 0.012 0.012
Chain 1: 3100 -18351.586 0.011 0.012
Chain 1: 3200 -18042.623 0.012 0.015
Chain 1: 3300 -18247.036 0.011 0.012
Chain 1: 3400 -17722.522 0.013 0.015
Chain 1: 3500 -18333.461 0.015 0.016
Chain 1: 3600 -17641.337 0.017 0.016
Chain 1: 3700 -18027.257 0.019 0.017
Chain 1: 3800 -16988.723 0.023 0.021
Chain 1: 3900 -16984.861 0.022 0.021
Chain 1: 4000 -17102.214 0.022 0.021
Chain 1: 4100 -17016.076 0.023 0.021
Chain 1: 4200 -16832.668 0.022 0.021
Chain 1: 4300 -16970.849 0.022 0.021
Chain 1: 4400 -16928.014 0.019 0.011
Chain 1: 4500 -16830.553 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001273 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.73 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49406.315 1.000 1.000
Chain 1: 200 -21137.399 1.169 1.337
Chain 1: 300 -16863.758 0.864 1.000
Chain 1: 400 -21865.857 0.705 1.000
Chain 1: 500 -12402.315 0.717 0.763
Chain 1: 600 -14769.840 0.624 0.763
Chain 1: 700 -16795.956 0.552 0.253
Chain 1: 800 -11053.405 0.548 0.520
Chain 1: 900 -14627.634 0.514 0.253
Chain 1: 1000 -15488.244 0.468 0.253
Chain 1: 1100 -11216.840 0.406 0.253
Chain 1: 1200 -10333.363 0.281 0.244
Chain 1: 1300 -12162.000 0.271 0.229
Chain 1: 1400 -11033.539 0.258 0.160
Chain 1: 1500 -12422.380 0.193 0.150
Chain 1: 1600 -13402.626 0.184 0.121
Chain 1: 1700 -10256.877 0.203 0.150
Chain 1: 1800 -10095.510 0.153 0.112
Chain 1: 1900 -14197.563 0.157 0.112
Chain 1: 2000 -13558.589 0.156 0.112
Chain 1: 2100 -9772.634 0.157 0.112
Chain 1: 2200 -18088.936 0.194 0.150
Chain 1: 2300 -9730.411 0.265 0.289
Chain 1: 2400 -13672.054 0.284 0.289
Chain 1: 2500 -12666.158 0.281 0.289
Chain 1: 2600 -9911.900 0.301 0.289
Chain 1: 2700 -9387.674 0.276 0.288
Chain 1: 2800 -10544.348 0.285 0.288
Chain 1: 2900 -9439.353 0.268 0.278
Chain 1: 3000 -13668.093 0.294 0.288
Chain 1: 3100 -10098.322 0.291 0.288
Chain 1: 3200 -15016.720 0.278 0.288
Chain 1: 3300 -13300.728 0.205 0.278
Chain 1: 3400 -10490.822 0.203 0.268
Chain 1: 3500 -13461.222 0.217 0.268
Chain 1: 3600 -9593.570 0.229 0.268
Chain 1: 3700 -11377.691 0.239 0.268
Chain 1: 3800 -8933.642 0.256 0.274
Chain 1: 3900 -9034.560 0.245 0.274
Chain 1: 4000 -9915.126 0.223 0.268
Chain 1: 4100 -11094.898 0.198 0.221
Chain 1: 4200 -10739.130 0.169 0.157
Chain 1: 4300 -10233.770 0.161 0.157
Chain 1: 4400 -10277.481 0.135 0.106
Chain 1: 4500 -11044.070 0.120 0.089
Chain 1: 4600 -8915.109 0.103 0.089
Chain 1: 4700 -12046.057 0.113 0.089
Chain 1: 4800 -9170.262 0.117 0.089
Chain 1: 4900 -11873.127 0.139 0.106
Chain 1: 5000 -10424.372 0.144 0.139
Chain 1: 5100 -9429.965 0.144 0.139
Chain 1: 5200 -8885.374 0.147 0.139
Chain 1: 5300 -10576.495 0.158 0.160
Chain 1: 5400 -10921.436 0.161 0.160
Chain 1: 5500 -8681.775 0.180 0.228
Chain 1: 5600 -10345.152 0.172 0.161
Chain 1: 5700 -10376.650 0.146 0.160
Chain 1: 5800 -8789.622 0.133 0.160
Chain 1: 5900 -12533.368 0.140 0.160
Chain 1: 6000 -8935.708 0.166 0.161
Chain 1: 6100 -8432.836 0.162 0.161
Chain 1: 6200 -8742.516 0.159 0.161
Chain 1: 6300 -10576.877 0.160 0.173
Chain 1: 6400 -8701.292 0.179 0.181
Chain 1: 6500 -9393.027 0.160 0.173
Chain 1: 6600 -9200.586 0.146 0.173
Chain 1: 6700 -8622.938 0.153 0.173
Chain 1: 6800 -9045.910 0.139 0.074
Chain 1: 6900 -12619.895 0.138 0.074
Chain 1: 7000 -8612.253 0.144 0.074
Chain 1: 7100 -10605.014 0.157 0.173
Chain 1: 7200 -14034.771 0.178 0.188
Chain 1: 7300 -12533.592 0.172 0.188
Chain 1: 7400 -8864.953 0.192 0.188
Chain 1: 7500 -9250.598 0.189 0.188
Chain 1: 7600 -9631.203 0.191 0.188
Chain 1: 7700 -8479.889 0.198 0.188
Chain 1: 7800 -8528.249 0.194 0.188
Chain 1: 7900 -9095.102 0.172 0.136
Chain 1: 8000 -10268.929 0.137 0.120
Chain 1: 8100 -10060.604 0.120 0.114
Chain 1: 8200 -9517.336 0.101 0.062
Chain 1: 8300 -9900.825 0.093 0.057
Chain 1: 8400 -9658.791 0.054 0.042
Chain 1: 8500 -9667.201 0.050 0.040
Chain 1: 8600 -8612.266 0.058 0.057
Chain 1: 8700 -8633.287 0.045 0.039
Chain 1: 8800 -8461.024 0.046 0.039
Chain 1: 8900 -10300.193 0.058 0.039
Chain 1: 9000 -9950.635 0.050 0.035
Chain 1: 9100 -9028.321 0.058 0.039
Chain 1: 9200 -9250.651 0.055 0.035
Chain 1: 9300 -8749.671 0.057 0.035
Chain 1: 9400 -9221.520 0.059 0.051
Chain 1: 9500 -9420.880 0.061 0.051
Chain 1: 9600 -10177.205 0.057 0.051
Chain 1: 9700 -8256.885 0.080 0.057
Chain 1: 9800 -11802.638 0.108 0.074
Chain 1: 9900 -9762.571 0.111 0.074
Chain 1: 10000 -8642.546 0.120 0.102
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001649 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 16.49 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56702.456 1.000 1.000
Chain 1: 200 -17631.181 1.608 2.216
Chain 1: 300 -8804.143 1.406 1.003
Chain 1: 400 -8159.076 1.074 1.003
Chain 1: 500 -8753.763 0.873 1.000
Chain 1: 600 -8074.360 0.742 1.000
Chain 1: 700 -7914.704 0.639 0.084
Chain 1: 800 -8116.222 0.562 0.084
Chain 1: 900 -7922.047 0.502 0.079
Chain 1: 1000 -7977.788 0.453 0.079
Chain 1: 1100 -7835.928 0.354 0.068
Chain 1: 1200 -7911.427 0.134 0.025
Chain 1: 1300 -7693.124 0.036 0.025
Chain 1: 1400 -8055.640 0.033 0.025
Chain 1: 1500 -7680.931 0.031 0.025
Chain 1: 1600 -7878.275 0.025 0.025
Chain 1: 1700 -7629.009 0.026 0.025
Chain 1: 1800 -7685.718 0.025 0.025
Chain 1: 1900 -7705.299 0.022 0.025
Chain 1: 2000 -7732.471 0.022 0.025
Chain 1: 2100 -7733.534 0.020 0.025
Chain 1: 2200 -7831.414 0.021 0.025
Chain 1: 2300 -7711.327 0.019 0.016
Chain 1: 2400 -7770.636 0.016 0.012
Chain 1: 2500 -7690.615 0.012 0.010
Chain 1: 2600 -7639.795 0.010 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002476 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 24.76 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86502.948 1.000 1.000
Chain 1: 200 -13713.308 3.154 5.308
Chain 1: 300 -10041.415 2.225 1.000
Chain 1: 400 -11343.204 1.697 1.000
Chain 1: 500 -8712.214 1.418 0.366
Chain 1: 600 -8578.283 1.184 0.366
Chain 1: 700 -8377.223 1.019 0.302
Chain 1: 800 -9546.404 0.907 0.302
Chain 1: 900 -8799.467 0.815 0.122
Chain 1: 1000 -8672.274 0.735 0.122
Chain 1: 1100 -8802.688 0.637 0.115 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8331.223 0.112 0.085
Chain 1: 1300 -8557.056 0.078 0.057
Chain 1: 1400 -8746.984 0.068 0.026
Chain 1: 1500 -8567.945 0.040 0.024
Chain 1: 1600 -8681.474 0.040 0.024
Chain 1: 1700 -8748.307 0.038 0.022
Chain 1: 1800 -8317.630 0.031 0.022
Chain 1: 1900 -8421.568 0.024 0.021
Chain 1: 2000 -8396.885 0.023 0.021
Chain 1: 2100 -8533.765 0.023 0.021
Chain 1: 2200 -8327.039 0.020 0.021
Chain 1: 2300 -8415.401 0.018 0.016
Chain 1: 2400 -8488.360 0.017 0.013
Chain 1: 2500 -8430.460 0.015 0.012
Chain 1: 2600 -8436.091 0.014 0.010
Chain 1: 2700 -8350.507 0.014 0.010
Chain 1: 2800 -8305.848 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003172 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.72 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8443180.638 1.000 1.000
Chain 1: 200 -1588931.544 2.657 4.314
Chain 1: 300 -890388.678 2.033 1.000
Chain 1: 400 -457647.435 1.761 1.000
Chain 1: 500 -357769.972 1.465 0.946
Chain 1: 600 -232679.648 1.310 0.946
Chain 1: 700 -119130.454 1.259 0.946
Chain 1: 800 -86454.227 1.149 0.946
Chain 1: 900 -66844.979 1.054 0.785
Chain 1: 1000 -51695.812 0.978 0.785
Chain 1: 1100 -39226.969 0.910 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38409.565 0.480 0.378
Chain 1: 1300 -26409.012 0.447 0.378
Chain 1: 1400 -26133.274 0.354 0.318
Chain 1: 1500 -22732.919 0.341 0.318
Chain 1: 1600 -21954.118 0.291 0.293
Chain 1: 1700 -20832.186 0.201 0.293
Chain 1: 1800 -20777.655 0.163 0.150
Chain 1: 1900 -21104.154 0.135 0.054
Chain 1: 2000 -19617.598 0.114 0.054
Chain 1: 2100 -19855.542 0.083 0.035
Chain 1: 2200 -20082.036 0.082 0.035
Chain 1: 2300 -19699.176 0.039 0.019
Chain 1: 2400 -19471.257 0.039 0.019
Chain 1: 2500 -19273.287 0.025 0.015
Chain 1: 2600 -18903.128 0.023 0.015
Chain 1: 2700 -18860.075 0.018 0.012
Chain 1: 2800 -18576.859 0.019 0.015
Chain 1: 2900 -18858.143 0.019 0.015
Chain 1: 3000 -18844.251 0.012 0.012
Chain 1: 3100 -18929.303 0.011 0.012
Chain 1: 3200 -18619.817 0.012 0.015
Chain 1: 3300 -18824.708 0.011 0.012
Chain 1: 3400 -18299.320 0.012 0.015
Chain 1: 3500 -18911.621 0.015 0.015
Chain 1: 3600 -18217.728 0.016 0.015
Chain 1: 3700 -18604.926 0.018 0.017
Chain 1: 3800 -17563.758 0.023 0.021
Chain 1: 3900 -17559.902 0.021 0.021
Chain 1: 4000 -17677.186 0.022 0.021
Chain 1: 4100 -17590.926 0.022 0.021
Chain 1: 4200 -17406.993 0.021 0.021
Chain 1: 4300 -17545.496 0.021 0.021
Chain 1: 4400 -17502.146 0.018 0.011
Chain 1: 4500 -17404.680 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00125 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.5 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49088.664 1.000 1.000
Chain 1: 200 -19520.048 1.257 1.515
Chain 1: 300 -19922.139 0.845 1.000
Chain 1: 400 -42818.117 0.767 1.000
Chain 1: 500 -11679.153 1.147 1.000
Chain 1: 600 -12201.615 0.963 1.000
Chain 1: 700 -13391.867 0.838 0.535
Chain 1: 800 -11819.772 0.750 0.535
Chain 1: 900 -14995.207 0.690 0.212
Chain 1: 1000 -16537.213 0.631 0.212
Chain 1: 1100 -11535.780 0.574 0.212 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -12865.529 0.433 0.133
Chain 1: 1300 -18587.135 0.462 0.212
Chain 1: 1400 -14632.698 0.435 0.212
Chain 1: 1500 -10437.458 0.209 0.212
Chain 1: 1600 -11826.720 0.216 0.212
Chain 1: 1700 -12641.932 0.214 0.212
Chain 1: 1800 -10678.633 0.219 0.212
Chain 1: 1900 -10563.655 0.199 0.184
Chain 1: 2000 -9614.235 0.199 0.184
Chain 1: 2100 -10486.040 0.164 0.117
Chain 1: 2200 -11018.062 0.159 0.117
Chain 1: 2300 -9516.278 0.144 0.117
Chain 1: 2400 -11339.079 0.133 0.117
Chain 1: 2500 -13027.028 0.106 0.117
Chain 1: 2600 -9310.660 0.134 0.130
Chain 1: 2700 -10165.635 0.136 0.130
Chain 1: 2800 -9972.970 0.119 0.099
Chain 1: 2900 -9956.117 0.118 0.099
Chain 1: 3000 -9467.513 0.114 0.084
Chain 1: 3100 -8908.082 0.112 0.084
Chain 1: 3200 -15798.065 0.150 0.130
Chain 1: 3300 -9714.101 0.197 0.130
Chain 1: 3400 -9274.441 0.186 0.084
Chain 1: 3500 -9411.669 0.174 0.063
Chain 1: 3600 -9995.079 0.140 0.058
Chain 1: 3700 -10072.093 0.133 0.052
Chain 1: 3800 -11885.339 0.146 0.058
Chain 1: 3900 -11846.602 0.146 0.058
Chain 1: 4000 -10782.021 0.151 0.063
Chain 1: 4100 -8767.018 0.167 0.099
Chain 1: 4200 -8826.010 0.125 0.058
Chain 1: 4300 -9685.694 0.071 0.058
Chain 1: 4400 -12389.092 0.088 0.089
Chain 1: 4500 -8835.033 0.127 0.099
Chain 1: 4600 -8919.756 0.122 0.099
Chain 1: 4700 -13458.118 0.155 0.153
Chain 1: 4800 -9050.730 0.188 0.218
Chain 1: 4900 -8986.825 0.189 0.218
Chain 1: 5000 -9577.227 0.185 0.218
Chain 1: 5100 -8812.867 0.171 0.089
Chain 1: 5200 -15928.013 0.215 0.218
Chain 1: 5300 -14341.046 0.217 0.218
Chain 1: 5400 -14646.628 0.197 0.111
Chain 1: 5500 -10741.054 0.193 0.111
Chain 1: 5600 -8494.307 0.219 0.265
Chain 1: 5700 -14988.148 0.228 0.265
Chain 1: 5800 -9022.368 0.246 0.265
Chain 1: 5900 -15360.620 0.286 0.364
Chain 1: 6000 -9185.655 0.347 0.413
Chain 1: 6100 -14131.026 0.374 0.413
Chain 1: 6200 -9211.201 0.382 0.413
Chain 1: 6300 -9007.558 0.374 0.413
Chain 1: 6400 -9539.077 0.377 0.413
Chain 1: 6500 -9351.944 0.343 0.413
Chain 1: 6600 -8862.092 0.322 0.413
Chain 1: 6700 -8835.036 0.279 0.350
Chain 1: 6800 -8517.902 0.216 0.056
Chain 1: 6900 -8789.221 0.178 0.055
Chain 1: 7000 -16268.581 0.157 0.055
Chain 1: 7100 -8309.748 0.218 0.055
Chain 1: 7200 -8928.883 0.171 0.055
Chain 1: 7300 -11186.355 0.189 0.056
Chain 1: 7400 -9918.824 0.196 0.069
Chain 1: 7500 -10882.612 0.203 0.089
Chain 1: 7600 -8546.645 0.225 0.128
Chain 1: 7700 -9815.509 0.238 0.129
Chain 1: 7800 -11537.635 0.249 0.149
Chain 1: 7900 -8487.850 0.282 0.202
Chain 1: 8000 -9793.582 0.249 0.149
Chain 1: 8100 -8962.839 0.162 0.133
Chain 1: 8200 -9915.138 0.165 0.133
Chain 1: 8300 -8313.018 0.164 0.133
Chain 1: 8400 -8254.695 0.152 0.133
Chain 1: 8500 -8788.755 0.149 0.133
Chain 1: 8600 -8794.437 0.122 0.129
Chain 1: 8700 -8087.667 0.118 0.096
Chain 1: 8800 -11440.707 0.132 0.096
Chain 1: 8900 -8448.768 0.132 0.096
Chain 1: 9000 -11103.512 0.142 0.096
Chain 1: 9100 -8221.961 0.168 0.193
Chain 1: 9200 -10856.100 0.183 0.239
Chain 1: 9300 -8174.474 0.196 0.243
Chain 1: 9400 -8633.169 0.201 0.243
Chain 1: 9500 -8093.100 0.202 0.243
Chain 1: 9600 -10640.125 0.225 0.243
Chain 1: 9700 -10170.777 0.221 0.243
Chain 1: 9800 -8462.413 0.212 0.239
Chain 1: 9900 -8289.375 0.179 0.239
Chain 1: 10000 -8346.097 0.156 0.202
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00142 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.2 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58101.853 1.000 1.000
Chain 1: 200 -17848.399 1.628 2.255
Chain 1: 300 -8765.317 1.431 1.036
Chain 1: 400 -8119.198 1.093 1.036
Chain 1: 500 -8972.202 0.893 1.000
Chain 1: 600 -8672.723 0.750 1.000
Chain 1: 700 -8118.394 0.653 0.095
Chain 1: 800 -8156.318 0.572 0.095
Chain 1: 900 -7931.251 0.511 0.080
Chain 1: 1000 -7898.422 0.461 0.080
Chain 1: 1100 -7762.383 0.362 0.068
Chain 1: 1200 -7791.173 0.137 0.035
Chain 1: 1300 -7688.658 0.035 0.028
Chain 1: 1400 -7831.388 0.029 0.018
Chain 1: 1500 -7620.394 0.022 0.018
Chain 1: 1600 -7798.955 0.021 0.018
Chain 1: 1700 -7565.900 0.017 0.018
Chain 1: 1800 -7672.018 0.018 0.018
Chain 1: 1900 -7624.383 0.016 0.018
Chain 1: 2000 -7685.380 0.016 0.018
Chain 1: 2100 -7614.603 0.015 0.014
Chain 1: 2200 -7742.606 0.017 0.017
Chain 1: 2300 -7632.594 0.017 0.017
Chain 1: 2400 -7673.097 0.015 0.014
Chain 1: 2500 -7582.401 0.014 0.014
Chain 1: 2600 -7539.256 0.012 0.012
Chain 1: 2700 -7532.759 0.009 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003252 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.52 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86621.081 1.000 1.000
Chain 1: 200 -13694.186 3.163 5.325
Chain 1: 300 -9968.964 2.233 1.000
Chain 1: 400 -11112.374 1.700 1.000
Chain 1: 500 -8991.231 1.408 0.374
Chain 1: 600 -8372.725 1.185 0.374
Chain 1: 700 -8862.790 1.024 0.236
Chain 1: 800 -8703.453 0.898 0.236
Chain 1: 900 -8805.454 0.800 0.103
Chain 1: 1000 -8776.275 0.720 0.103
Chain 1: 1100 -8544.316 0.623 0.074 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8334.021 0.093 0.055
Chain 1: 1300 -8631.880 0.059 0.035
Chain 1: 1400 -8621.011 0.049 0.027
Chain 1: 1500 -8476.730 0.027 0.025
Chain 1: 1600 -8592.673 0.021 0.018
Chain 1: 1700 -8655.052 0.016 0.017
Chain 1: 1800 -8217.651 0.019 0.017
Chain 1: 1900 -8322.135 0.019 0.017
Chain 1: 2000 -8299.676 0.019 0.017
Chain 1: 2100 -8275.389 0.017 0.013
Chain 1: 2200 -8243.487 0.015 0.013
Chain 1: 2300 -8377.793 0.013 0.013
Chain 1: 2400 -8219.365 0.015 0.013
Chain 1: 2500 -8291.814 0.014 0.013
Chain 1: 2600 -8205.116 0.014 0.011
Chain 1: 2700 -8241.702 0.013 0.011
Chain 1: 2800 -8200.187 0.009 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003464 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.64 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8380325.708 1.000 1.000
Chain 1: 200 -1584491.550 2.644 4.289
Chain 1: 300 -890812.201 2.023 1.000
Chain 1: 400 -457173.969 1.754 1.000
Chain 1: 500 -357639.764 1.459 0.949
Chain 1: 600 -232772.955 1.305 0.949
Chain 1: 700 -119274.092 1.255 0.949
Chain 1: 800 -86502.168 1.145 0.949
Chain 1: 900 -66904.703 1.050 0.779
Chain 1: 1000 -51739.998 0.975 0.779
Chain 1: 1100 -39238.857 0.907 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38425.569 0.480 0.379
Chain 1: 1300 -26400.437 0.447 0.379
Chain 1: 1400 -26123.885 0.354 0.319
Chain 1: 1500 -22713.842 0.341 0.319
Chain 1: 1600 -21931.485 0.291 0.293
Chain 1: 1700 -20807.314 0.201 0.293
Chain 1: 1800 -20752.219 0.163 0.150
Chain 1: 1900 -21078.853 0.136 0.054
Chain 1: 2000 -19589.672 0.114 0.054
Chain 1: 2100 -19828.484 0.083 0.036
Chain 1: 2200 -20054.840 0.082 0.036
Chain 1: 2300 -19671.935 0.039 0.019
Chain 1: 2400 -19443.858 0.039 0.019
Chain 1: 2500 -19245.556 0.025 0.015
Chain 1: 2600 -18875.736 0.023 0.015
Chain 1: 2700 -18832.636 0.018 0.012
Chain 1: 2800 -18549.137 0.019 0.015
Chain 1: 2900 -18830.566 0.019 0.015
Chain 1: 3000 -18816.882 0.012 0.012
Chain 1: 3100 -18901.888 0.011 0.012
Chain 1: 3200 -18592.374 0.012 0.015
Chain 1: 3300 -18797.230 0.011 0.012
Chain 1: 3400 -18271.651 0.012 0.015
Chain 1: 3500 -18884.223 0.015 0.015
Chain 1: 3600 -18189.985 0.016 0.015
Chain 1: 3700 -18577.452 0.018 0.017
Chain 1: 3800 -17535.651 0.023 0.021
Chain 1: 3900 -17531.682 0.021 0.021
Chain 1: 4000 -17649.061 0.022 0.021
Chain 1: 4100 -17562.717 0.022 0.021
Chain 1: 4200 -17378.618 0.021 0.021
Chain 1: 4300 -17517.317 0.021 0.021
Chain 1: 4400 -17473.897 0.018 0.011
Chain 1: 4500 -17376.301 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001472 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.72 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48403.705 1.000 1.000
Chain 1: 200 -14405.190 1.680 2.360
Chain 1: 300 -14347.304 1.121 1.000
Chain 1: 400 -12379.250 0.881 1.000
Chain 1: 500 -11911.071 0.712 0.159
Chain 1: 600 -10887.649 0.609 0.159
Chain 1: 700 -20899.236 0.591 0.159
Chain 1: 800 -18391.269 0.534 0.159
Chain 1: 900 -11552.161 0.540 0.159
Chain 1: 1000 -18247.700 0.523 0.367
Chain 1: 1100 -9878.820 0.508 0.367 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -12003.250 0.289 0.177
Chain 1: 1300 -11328.624 0.295 0.177
Chain 1: 1400 -23383.285 0.331 0.367
Chain 1: 1500 -54989.578 0.384 0.479
Chain 1: 1600 -9166.916 0.875 0.516 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1700 -21084.631 0.883 0.565 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1800 -10441.406 0.972 0.575 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1900 -22872.200 0.967 0.565 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2000 -12190.953 1.018 0.575 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2100 -9119.525 0.967 0.565 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2200 -9179.299 0.950 0.565 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2300 -10926.000 0.960 0.565 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2400 -8886.474 0.931 0.565 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2500 -14718.818 0.913 0.543 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2600 -9663.530 0.466 0.523 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2700 -11629.398 0.426 0.396
Chain 1: 2800 -8905.046 0.355 0.337
Chain 1: 2900 -9670.402 0.308 0.306
Chain 1: 3000 -8574.124 0.233 0.230
Chain 1: 3100 -8515.115 0.200 0.169
Chain 1: 3200 -8697.993 0.202 0.169
Chain 1: 3300 -10389.606 0.202 0.169
Chain 1: 3400 -9814.565 0.185 0.163
Chain 1: 3500 -8867.405 0.156 0.128
Chain 1: 3600 -13645.848 0.139 0.128
Chain 1: 3700 -8412.625 0.184 0.128
Chain 1: 3800 -9148.129 0.162 0.107
Chain 1: 3900 -9128.744 0.154 0.107
Chain 1: 4000 -9096.703 0.141 0.080
Chain 1: 4100 -8631.970 0.146 0.080
Chain 1: 4200 -12738.569 0.176 0.107
Chain 1: 4300 -9013.239 0.201 0.107
Chain 1: 4400 -11885.792 0.220 0.242
Chain 1: 4500 -13583.389 0.221 0.242
Chain 1: 4600 -8950.045 0.238 0.242
Chain 1: 4700 -9990.172 0.186 0.125
Chain 1: 4800 -8420.672 0.197 0.186
Chain 1: 4900 -8608.939 0.199 0.186
Chain 1: 5000 -11649.032 0.225 0.242
Chain 1: 5100 -8431.345 0.258 0.261
Chain 1: 5200 -8527.793 0.226 0.242
Chain 1: 5300 -9472.557 0.195 0.186
Chain 1: 5400 -8494.395 0.182 0.125
Chain 1: 5500 -11905.676 0.199 0.186
Chain 1: 5600 -9404.900 0.173 0.186
Chain 1: 5700 -10681.354 0.175 0.186
Chain 1: 5800 -10501.927 0.158 0.120
Chain 1: 5900 -9604.902 0.165 0.120
Chain 1: 6000 -8168.407 0.157 0.120
Chain 1: 6100 -8699.143 0.125 0.115
Chain 1: 6200 -8083.352 0.131 0.115
Chain 1: 6300 -12015.490 0.154 0.120
Chain 1: 6400 -13225.002 0.151 0.120
Chain 1: 6500 -9010.279 0.170 0.120
Chain 1: 6600 -8659.500 0.147 0.093
Chain 1: 6700 -9959.772 0.148 0.093
Chain 1: 6800 -12774.760 0.168 0.131
Chain 1: 6900 -9728.619 0.190 0.176
Chain 1: 7000 -8429.735 0.188 0.154
Chain 1: 7100 -12165.979 0.213 0.220
Chain 1: 7200 -10679.614 0.219 0.220
Chain 1: 7300 -8420.140 0.213 0.220
Chain 1: 7400 -8955.721 0.210 0.220
Chain 1: 7500 -9721.251 0.171 0.154
Chain 1: 7600 -8745.744 0.178 0.154
Chain 1: 7700 -8079.925 0.173 0.154
Chain 1: 7800 -11775.667 0.183 0.154
Chain 1: 7900 -10225.095 0.167 0.152
Chain 1: 8000 -9388.477 0.160 0.139
Chain 1: 8100 -8116.128 0.145 0.139
Chain 1: 8200 -10378.946 0.153 0.152
Chain 1: 8300 -8265.294 0.152 0.152
Chain 1: 8400 -8017.698 0.149 0.152
Chain 1: 8500 -8385.889 0.145 0.152
Chain 1: 8600 -10571.830 0.155 0.157
Chain 1: 8700 -9787.617 0.155 0.157
Chain 1: 8800 -10187.448 0.127 0.152
Chain 1: 8900 -10414.644 0.114 0.089
Chain 1: 9000 -11538.250 0.115 0.097
Chain 1: 9100 -8563.470 0.134 0.097
Chain 1: 9200 -8003.251 0.119 0.080
Chain 1: 9300 -9012.676 0.105 0.080
Chain 1: 9400 -9077.441 0.103 0.080
Chain 1: 9500 -8719.949 0.102 0.080
Chain 1: 9600 -8073.657 0.090 0.080
Chain 1: 9700 -11322.765 0.110 0.080
Chain 1: 9800 -9515.451 0.125 0.097
Chain 1: 9900 -9745.120 0.126 0.097
Chain 1: 10000 -8057.149 0.137 0.112
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001387 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.87 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56545.748 1.000 1.000
Chain 1: 200 -17030.491 1.660 2.320
Chain 1: 300 -8563.823 1.436 1.000
Chain 1: 400 -8908.649 1.087 1.000
Chain 1: 500 -8500.142 0.879 0.989
Chain 1: 600 -9091.076 0.743 0.989
Chain 1: 700 -7765.868 0.662 0.171
Chain 1: 800 -8083.371 0.584 0.171
Chain 1: 900 -7899.933 0.522 0.065
Chain 1: 1000 -7735.623 0.472 0.065
Chain 1: 1100 -7663.197 0.372 0.048
Chain 1: 1200 -7576.592 0.142 0.039
Chain 1: 1300 -7687.956 0.044 0.039
Chain 1: 1400 -7874.840 0.043 0.024
Chain 1: 1500 -7619.851 0.041 0.024
Chain 1: 1600 -7521.936 0.036 0.023
Chain 1: 1700 -7528.582 0.019 0.021
Chain 1: 1800 -7537.275 0.015 0.014
Chain 1: 1900 -7502.992 0.013 0.013
Chain 1: 2000 -7602.445 0.013 0.013
Chain 1: 2100 -7669.404 0.012 0.013
Chain 1: 2200 -7679.588 0.011 0.013
Chain 1: 2300 -7576.864 0.011 0.013
Chain 1: 2400 -7614.770 0.009 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002551 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.51 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86766.212 1.000 1.000
Chain 1: 200 -13109.245 3.309 5.619
Chain 1: 300 -9571.763 2.329 1.000
Chain 1: 400 -10422.068 1.767 1.000
Chain 1: 500 -8497.687 1.459 0.370
Chain 1: 600 -8372.586 1.219 0.370
Chain 1: 700 -8387.579 1.045 0.226
Chain 1: 800 -8861.660 0.921 0.226
Chain 1: 900 -8401.875 0.825 0.082
Chain 1: 1000 -8216.707 0.744 0.082
Chain 1: 1100 -8491.943 0.648 0.055 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8229.104 0.089 0.053
Chain 1: 1300 -8165.860 0.053 0.032
Chain 1: 1400 -8165.770 0.045 0.032
Chain 1: 1500 -8198.630 0.022 0.023
Chain 1: 1600 -8204.085 0.021 0.023
Chain 1: 1700 -8139.463 0.022 0.023
Chain 1: 1800 -8021.257 0.018 0.015
Chain 1: 1900 -8136.727 0.014 0.014
Chain 1: 2000 -8097.145 0.012 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003117 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.17 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8412724.227 1.000 1.000
Chain 1: 200 -1588140.314 2.649 4.297
Chain 1: 300 -890863.088 2.027 1.000
Chain 1: 400 -456574.611 1.758 1.000
Chain 1: 500 -356541.338 1.462 0.951
Chain 1: 600 -231649.184 1.308 0.951
Chain 1: 700 -118355.931 1.258 0.951
Chain 1: 800 -85654.785 1.149 0.951
Chain 1: 900 -66097.800 1.054 0.783
Chain 1: 1000 -50958.541 0.978 0.783
Chain 1: 1100 -38496.084 0.911 0.539 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37678.127 0.483 0.382
Chain 1: 1300 -25707.602 0.451 0.382
Chain 1: 1400 -25431.052 0.357 0.324
Chain 1: 1500 -22036.788 0.345 0.324
Chain 1: 1600 -21257.824 0.294 0.297
Chain 1: 1700 -20141.052 0.204 0.296
Chain 1: 1800 -20087.029 0.166 0.154
Chain 1: 1900 -20412.569 0.138 0.055
Chain 1: 2000 -18929.653 0.117 0.055
Chain 1: 2100 -19167.867 0.085 0.037
Chain 1: 2200 -19392.951 0.084 0.037
Chain 1: 2300 -19011.505 0.040 0.020
Chain 1: 2400 -18783.918 0.040 0.020
Chain 1: 2500 -18585.541 0.026 0.016
Chain 1: 2600 -18216.844 0.024 0.016
Chain 1: 2700 -18174.176 0.019 0.012
Chain 1: 2800 -17891.121 0.020 0.016
Chain 1: 2900 -18172.000 0.020 0.015
Chain 1: 3000 -18158.353 0.012 0.012
Chain 1: 3100 -18243.184 0.011 0.012
Chain 1: 3200 -17934.437 0.012 0.015
Chain 1: 3300 -18138.743 0.011 0.012
Chain 1: 3400 -17614.487 0.013 0.015
Chain 1: 3500 -18224.977 0.015 0.016
Chain 1: 3600 -17533.497 0.017 0.016
Chain 1: 3700 -17918.853 0.019 0.017
Chain 1: 3800 -16881.291 0.024 0.022
Chain 1: 3900 -16877.449 0.022 0.022
Chain 1: 4000 -16994.815 0.023 0.022
Chain 1: 4100 -16908.637 0.023 0.022
Chain 1: 4200 -16725.523 0.022 0.022
Chain 1: 4300 -16863.527 0.022 0.022
Chain 1: 4400 -16820.858 0.019 0.011
Chain 1: 4500 -16723.417 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001281 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.81 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13774.740 1.000 1.000
Chain 1: 200 -10465.930 0.658 1.000
Chain 1: 300 -8447.654 0.518 0.316
Chain 1: 400 -8651.284 0.395 0.316
Chain 1: 500 -8634.424 0.316 0.239
Chain 1: 600 -8415.595 0.268 0.239
Chain 1: 700 -8397.783 0.230 0.026
Chain 1: 800 -8364.469 0.202 0.026
Chain 1: 900 -8402.036 0.180 0.024
Chain 1: 1000 -8365.810 0.162 0.024
Chain 1: 1100 -8426.557 0.063 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002412 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 24.12 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58869.814 1.000 1.000
Chain 1: 200 -18516.469 1.590 2.179
Chain 1: 300 -9128.855 1.403 1.028
Chain 1: 400 -8235.446 1.079 1.028
Chain 1: 500 -8625.132 0.872 1.000
Chain 1: 600 -8530.333 0.729 1.000
Chain 1: 700 -7945.224 0.635 0.108
Chain 1: 800 -8605.215 0.565 0.108
Chain 1: 900 -8294.506 0.507 0.077
Chain 1: 1000 -7868.874 0.461 0.077
Chain 1: 1100 -7882.854 0.362 0.074
Chain 1: 1200 -8188.589 0.147 0.054
Chain 1: 1300 -8254.595 0.045 0.045
Chain 1: 1400 -7840.883 0.040 0.045
Chain 1: 1500 -7677.137 0.037 0.037
Chain 1: 1600 -7949.453 0.040 0.037
Chain 1: 1700 -7798.346 0.034 0.037
Chain 1: 1800 -7719.087 0.028 0.034
Chain 1: 1900 -7853.564 0.026 0.021
Chain 1: 2000 -7834.742 0.020 0.019
Chain 1: 2100 -7726.505 0.022 0.019
Chain 1: 2200 -7975.275 0.021 0.019
Chain 1: 2300 -7760.884 0.023 0.021
Chain 1: 2400 -7767.058 0.018 0.019
Chain 1: 2500 -7682.541 0.017 0.017
Chain 1: 2600 -7683.446 0.013 0.014
Chain 1: 2700 -7664.276 0.012 0.011
Chain 1: 2800 -7676.469 0.011 0.011
Chain 1: 2900 -7581.864 0.010 0.011
Chain 1: 3000 -7703.296 0.012 0.012
Chain 1: 3100 -7683.702 0.011 0.011
Chain 1: 3200 -7891.485 0.010 0.011
Chain 1: 3300 -7581.234 0.011 0.011
Chain 1: 3400 -7801.428 0.014 0.012
Chain 1: 3500 -7591.695 0.016 0.016
Chain 1: 3600 -7653.690 0.017 0.016
Chain 1: 3700 -7605.778 0.017 0.016
Chain 1: 3800 -7578.838 0.017 0.016
Chain 1: 3900 -7553.294 0.016 0.016
Chain 1: 4000 -7548.527 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003418 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.18 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87197.932 1.000 1.000
Chain 1: 200 -14276.496 3.054 5.108
Chain 1: 300 -10503.823 2.156 1.000
Chain 1: 400 -11978.332 1.648 1.000
Chain 1: 500 -9410.612 1.373 0.359
Chain 1: 600 -8948.453 1.152 0.359
Chain 1: 700 -9192.497 0.992 0.273
Chain 1: 800 -9688.544 0.874 0.273
Chain 1: 900 -9123.770 0.784 0.123
Chain 1: 1000 -9219.139 0.706 0.123
Chain 1: 1100 -9318.583 0.608 0.062 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8782.595 0.103 0.061
Chain 1: 1300 -9125.730 0.071 0.052
Chain 1: 1400 -8872.475 0.061 0.051
Chain 1: 1500 -8971.527 0.035 0.038
Chain 1: 1600 -9080.533 0.031 0.029
Chain 1: 1700 -9129.229 0.029 0.029
Chain 1: 1800 -8673.435 0.029 0.029
Chain 1: 1900 -8784.127 0.024 0.013
Chain 1: 2000 -8793.286 0.023 0.013
Chain 1: 2100 -8732.709 0.023 0.013
Chain 1: 2200 -8714.284 0.017 0.012
Chain 1: 2300 -8893.713 0.015 0.012
Chain 1: 2400 -8682.079 0.015 0.012
Chain 1: 2500 -8753.322 0.015 0.012
Chain 1: 2600 -8668.590 0.014 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003519 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.19 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8413991.602 1.000 1.000
Chain 1: 200 -1588555.891 2.648 4.297
Chain 1: 300 -890958.125 2.027 1.000
Chain 1: 400 -457909.989 1.756 1.000
Chain 1: 500 -358197.731 1.461 0.946
Chain 1: 600 -233099.306 1.307 0.946
Chain 1: 700 -119693.551 1.255 0.946
Chain 1: 800 -86983.662 1.145 0.946
Chain 1: 900 -67402.325 1.050 0.783
Chain 1: 1000 -52271.759 0.974 0.783
Chain 1: 1100 -39803.642 0.906 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38994.311 0.478 0.376
Chain 1: 1300 -26993.504 0.444 0.376
Chain 1: 1400 -26719.982 0.351 0.313
Chain 1: 1500 -23317.515 0.337 0.313
Chain 1: 1600 -22538.193 0.287 0.291
Chain 1: 1700 -21416.296 0.198 0.289
Chain 1: 1800 -21362.015 0.160 0.146
Chain 1: 1900 -21688.991 0.133 0.052
Chain 1: 2000 -20200.843 0.111 0.052
Chain 1: 2100 -20439.337 0.081 0.035
Chain 1: 2200 -20665.904 0.080 0.035
Chain 1: 2300 -20282.763 0.038 0.019
Chain 1: 2400 -20054.610 0.038 0.019
Chain 1: 2500 -19856.358 0.024 0.015
Chain 1: 2600 -19486.101 0.023 0.015
Chain 1: 2700 -19442.912 0.018 0.012
Chain 1: 2800 -19159.299 0.019 0.015
Chain 1: 2900 -19440.851 0.019 0.014
Chain 1: 3000 -19427.039 0.011 0.012
Chain 1: 3100 -19512.142 0.011 0.011
Chain 1: 3200 -19202.403 0.011 0.014
Chain 1: 3300 -19407.453 0.010 0.011
Chain 1: 3400 -18881.480 0.012 0.014
Chain 1: 3500 -19494.647 0.014 0.015
Chain 1: 3600 -18799.597 0.016 0.015
Chain 1: 3700 -19187.648 0.018 0.016
Chain 1: 3800 -18144.644 0.022 0.020
Chain 1: 3900 -18140.665 0.021 0.020
Chain 1: 4000 -18258.020 0.021 0.020
Chain 1: 4100 -18171.644 0.021 0.020
Chain 1: 4200 -17987.273 0.021 0.020
Chain 1: 4300 -18126.146 0.020 0.020
Chain 1: 4400 -18082.489 0.018 0.010
Chain 1: 4500 -17984.874 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001312 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.12 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48900.071 1.000 1.000
Chain 1: 200 -20845.615 1.173 1.346
Chain 1: 300 -19412.835 0.807 1.000
Chain 1: 400 -12949.649 0.730 1.000
Chain 1: 500 -15195.677 0.613 0.499
Chain 1: 600 -13152.602 0.537 0.499
Chain 1: 700 -12323.560 0.470 0.155
Chain 1: 800 -14084.099 0.427 0.155
Chain 1: 900 -10701.725 0.414 0.155
Chain 1: 1000 -13428.652 0.393 0.203
Chain 1: 1100 -23297.801 0.336 0.203
Chain 1: 1200 -11624.311 0.302 0.203
Chain 1: 1300 -10268.717 0.307 0.203
Chain 1: 1400 -20684.283 0.308 0.203
Chain 1: 1500 -10270.135 0.394 0.316
Chain 1: 1600 -9895.385 0.383 0.316
Chain 1: 1700 -10387.802 0.381 0.316
Chain 1: 1800 -23983.037 0.425 0.424
Chain 1: 1900 -10721.721 0.517 0.504 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2000 -9952.586 0.504 0.504 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2100 -9667.191 0.465 0.504 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2200 -11816.955 0.383 0.182
Chain 1: 2300 -9918.970 0.389 0.191
Chain 1: 2400 -8927.597 0.349 0.182
Chain 1: 2500 -15732.722 0.291 0.182
Chain 1: 2600 -16503.661 0.292 0.182
Chain 1: 2700 -9453.214 0.362 0.191
Chain 1: 2800 -16507.558 0.348 0.191
Chain 1: 2900 -9321.724 0.301 0.191
Chain 1: 3000 -11400.872 0.312 0.191
Chain 1: 3100 -15666.913 0.336 0.272
Chain 1: 3200 -9157.292 0.389 0.427
Chain 1: 3300 -9255.456 0.371 0.427
Chain 1: 3400 -13626.271 0.392 0.427
Chain 1: 3500 -9187.117 0.397 0.427
Chain 1: 3600 -8990.268 0.395 0.427
Chain 1: 3700 -9145.478 0.322 0.321
Chain 1: 3800 -9677.725 0.284 0.272
Chain 1: 3900 -15032.656 0.243 0.272
Chain 1: 4000 -9497.894 0.283 0.321
Chain 1: 4100 -8866.940 0.263 0.321
Chain 1: 4200 -12902.399 0.223 0.313
Chain 1: 4300 -11842.922 0.231 0.313
Chain 1: 4400 -14446.590 0.217 0.180
Chain 1: 4500 -8940.152 0.230 0.180
Chain 1: 4600 -13224.126 0.260 0.313
Chain 1: 4700 -14381.224 0.267 0.313
Chain 1: 4800 -8595.480 0.329 0.324
Chain 1: 4900 -10590.105 0.312 0.313
Chain 1: 5000 -9825.769 0.261 0.188
Chain 1: 5100 -11900.012 0.272 0.188
Chain 1: 5200 -10610.110 0.253 0.180
Chain 1: 5300 -12276.357 0.257 0.180
Chain 1: 5400 -9975.448 0.262 0.188
Chain 1: 5500 -8800.925 0.214 0.174
Chain 1: 5600 -8422.997 0.186 0.136
Chain 1: 5700 -8887.889 0.183 0.136
Chain 1: 5800 -8610.093 0.119 0.133
Chain 1: 5900 -10519.795 0.118 0.133
Chain 1: 6000 -9279.406 0.124 0.134
Chain 1: 6100 -10494.639 0.118 0.133
Chain 1: 6200 -11413.799 0.114 0.133
Chain 1: 6300 -11524.867 0.101 0.116
Chain 1: 6400 -10735.910 0.086 0.081
Chain 1: 6500 -10371.052 0.076 0.073
Chain 1: 6600 -9591.721 0.080 0.081
Chain 1: 6700 -8963.983 0.081 0.081
Chain 1: 6800 -8542.903 0.083 0.081
Chain 1: 6900 -8696.641 0.067 0.073
Chain 1: 7000 -8516.391 0.055 0.070
Chain 1: 7100 -8504.228 0.044 0.049
Chain 1: 7200 -8241.397 0.039 0.035
Chain 1: 7300 -8272.604 0.039 0.035
Chain 1: 7400 -8400.547 0.033 0.032
Chain 1: 7500 -9448.765 0.040 0.032
Chain 1: 7600 -9612.991 0.034 0.021
Chain 1: 7700 -8476.201 0.040 0.021
Chain 1: 7800 -11826.120 0.064 0.021
Chain 1: 7900 -8391.891 0.103 0.032
Chain 1: 8000 -12862.509 0.135 0.111
Chain 1: 8100 -11862.161 0.144 0.111
Chain 1: 8200 -9351.463 0.167 0.134
Chain 1: 8300 -9176.966 0.169 0.134
Chain 1: 8400 -9230.700 0.168 0.134
Chain 1: 8500 -8870.815 0.161 0.134
Chain 1: 8600 -8792.554 0.160 0.134
Chain 1: 8700 -10189.548 0.160 0.137
Chain 1: 8800 -8747.631 0.149 0.137
Chain 1: 8900 -9303.126 0.114 0.084
Chain 1: 9000 -8246.658 0.092 0.084
Chain 1: 9100 -8592.436 0.087 0.060
Chain 1: 9200 -9722.293 0.072 0.060
Chain 1: 9300 -8155.604 0.089 0.116
Chain 1: 9400 -8792.959 0.096 0.116
Chain 1: 9500 -8696.235 0.093 0.116
Chain 1: 9600 -8359.235 0.096 0.116
Chain 1: 9700 -8422.773 0.083 0.072
Chain 1: 9800 -11297.341 0.092 0.072
Chain 1: 9900 -11337.813 0.087 0.072
Chain 1: 10000 -8317.428 0.110 0.072
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001547 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.47 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58001.095 1.000 1.000
Chain 1: 200 -17736.949 1.635 2.270
Chain 1: 300 -8694.672 1.437 1.040
Chain 1: 400 -8215.994 1.092 1.040
Chain 1: 500 -8098.782 0.877 1.000
Chain 1: 600 -8659.451 0.741 1.000
Chain 1: 700 -8245.900 0.643 0.065
Chain 1: 800 -8268.727 0.563 0.065
Chain 1: 900 -8001.988 0.504 0.058
Chain 1: 1000 -7810.167 0.456 0.058
Chain 1: 1100 -7712.109 0.357 0.050
Chain 1: 1200 -7549.386 0.132 0.033
Chain 1: 1300 -7746.359 0.031 0.025
Chain 1: 1400 -7896.547 0.027 0.025
Chain 1: 1500 -7590.766 0.029 0.025
Chain 1: 1600 -7777.264 0.025 0.025
Chain 1: 1700 -7565.643 0.023 0.025
Chain 1: 1800 -7576.398 0.023 0.025
Chain 1: 1900 -7598.964 0.020 0.024
Chain 1: 2000 -7611.364 0.018 0.022
Chain 1: 2100 -7594.914 0.017 0.022
Chain 1: 2200 -7701.672 0.016 0.019
Chain 1: 2300 -7577.334 0.015 0.016
Chain 1: 2400 -7591.118 0.013 0.014
Chain 1: 2500 -7657.957 0.010 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003348 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.48 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86998.860 1.000 1.000
Chain 1: 200 -13555.335 3.209 5.418
Chain 1: 300 -9940.338 2.261 1.000
Chain 1: 400 -10902.250 1.717 1.000
Chain 1: 500 -8803.177 1.422 0.364
Chain 1: 600 -8440.047 1.192 0.364
Chain 1: 700 -8492.905 1.023 0.238
Chain 1: 800 -8900.110 0.900 0.238
Chain 1: 900 -8712.532 0.803 0.088
Chain 1: 1000 -8454.430 0.726 0.088
Chain 1: 1100 -8825.348 0.630 0.046 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8421.369 0.093 0.046
Chain 1: 1300 -8638.896 0.059 0.043
Chain 1: 1400 -8678.302 0.051 0.042
Chain 1: 1500 -8511.753 0.029 0.031
Chain 1: 1600 -8632.598 0.026 0.025
Chain 1: 1700 -8718.194 0.026 0.025
Chain 1: 1800 -8311.310 0.026 0.025
Chain 1: 1900 -8407.597 0.025 0.025
Chain 1: 2000 -8379.960 0.023 0.020
Chain 1: 2100 -8500.636 0.020 0.014
Chain 1: 2200 -8315.920 0.017 0.014
Chain 1: 2300 -8447.566 0.016 0.014
Chain 1: 2400 -8457.075 0.016 0.014
Chain 1: 2500 -8420.394 0.014 0.014
Chain 1: 2600 -8418.697 0.013 0.011
Chain 1: 2700 -8333.258 0.013 0.011
Chain 1: 2800 -8298.168 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003415 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.15 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8428255.347 1.000 1.000
Chain 1: 200 -1589974.427 2.650 4.301
Chain 1: 300 -891336.624 2.028 1.000
Chain 1: 400 -457630.004 1.758 1.000
Chain 1: 500 -357545.376 1.462 0.948
Chain 1: 600 -232436.880 1.308 0.948
Chain 1: 700 -118975.818 1.258 0.948
Chain 1: 800 -86229.801 1.148 0.948
Chain 1: 900 -66639.784 1.053 0.784
Chain 1: 1000 -51484.797 0.977 0.784
Chain 1: 1100 -39009.806 0.909 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38192.708 0.481 0.380
Chain 1: 1300 -26204.985 0.449 0.380
Chain 1: 1400 -25927.966 0.355 0.320
Chain 1: 1500 -22529.736 0.342 0.320
Chain 1: 1600 -21750.259 0.292 0.294
Chain 1: 1700 -20631.119 0.202 0.294
Chain 1: 1800 -20576.821 0.164 0.151
Chain 1: 1900 -20902.878 0.136 0.054
Chain 1: 2000 -19418.069 0.114 0.054
Chain 1: 2100 -19656.205 0.084 0.036
Chain 1: 2200 -19881.889 0.083 0.036
Chain 1: 2300 -19499.825 0.039 0.020
Chain 1: 2400 -19272.073 0.039 0.020
Chain 1: 2500 -19073.741 0.025 0.016
Chain 1: 2600 -18704.332 0.023 0.016
Chain 1: 2700 -18661.527 0.018 0.012
Chain 1: 2800 -18378.253 0.019 0.015
Chain 1: 2900 -18659.437 0.019 0.015
Chain 1: 3000 -18645.673 0.012 0.012
Chain 1: 3100 -18730.571 0.011 0.012
Chain 1: 3200 -18421.451 0.012 0.015
Chain 1: 3300 -18626.091 0.011 0.012
Chain 1: 3400 -18101.135 0.012 0.015
Chain 1: 3500 -18712.625 0.015 0.015
Chain 1: 3600 -18019.916 0.017 0.015
Chain 1: 3700 -18406.179 0.018 0.017
Chain 1: 3800 -17366.622 0.023 0.021
Chain 1: 3900 -17362.775 0.021 0.021
Chain 1: 4000 -17480.129 0.022 0.021
Chain 1: 4100 -17393.822 0.022 0.021
Chain 1: 4200 -17210.317 0.021 0.021
Chain 1: 4300 -17348.584 0.021 0.021
Chain 1: 4400 -17305.567 0.018 0.011
Chain 1: 4500 -17208.097 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001594 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.94 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13059.163 1.000 1.000
Chain 1: 200 -9962.684 0.655 1.000
Chain 1: 300 -8484.623 0.495 0.311
Chain 1: 400 -8727.713 0.378 0.311
Chain 1: 500 -8626.956 0.305 0.174
Chain 1: 600 -8458.246 0.257 0.174
Chain 1: 700 -8359.823 0.222 0.028
Chain 1: 800 -8329.902 0.195 0.028
Chain 1: 900 -8366.632 0.174 0.020
Chain 1: 1000 -8443.875 0.157 0.020
Chain 1: 1100 -8468.915 0.058 0.012
Chain 1: 1200 -8380.624 0.028 0.012
Chain 1: 1300 -8322.290 0.011 0.011
Chain 1: 1400 -8340.728 0.008 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001405 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.05 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -63008.164 1.000 1.000
Chain 1: 200 -18513.091 1.702 2.403
Chain 1: 300 -9239.733 1.469 1.004
Chain 1: 400 -8414.207 1.126 1.004
Chain 1: 500 -8481.639 0.903 1.000
Chain 1: 600 -9016.163 0.762 1.000
Chain 1: 700 -8783.915 0.657 0.098
Chain 1: 800 -8285.517 0.582 0.098
Chain 1: 900 -8168.079 0.519 0.060
Chain 1: 1000 -7884.471 0.471 0.060
Chain 1: 1100 -7655.910 0.374 0.059
Chain 1: 1200 -7701.818 0.134 0.036
Chain 1: 1300 -8134.797 0.039 0.036
Chain 1: 1400 -7810.898 0.033 0.036
Chain 1: 1500 -7643.022 0.035 0.036
Chain 1: 1600 -7937.420 0.033 0.036
Chain 1: 1700 -7556.220 0.035 0.037
Chain 1: 1800 -7734.236 0.031 0.036
Chain 1: 1900 -7826.725 0.031 0.036
Chain 1: 2000 -7720.517 0.029 0.030
Chain 1: 2100 -7612.531 0.027 0.023
Chain 1: 2200 -7959.492 0.031 0.037
Chain 1: 2300 -7656.388 0.030 0.037
Chain 1: 2400 -7819.028 0.028 0.023
Chain 1: 2500 -7616.126 0.028 0.027
Chain 1: 2600 -7600.596 0.025 0.023
Chain 1: 2700 -7568.566 0.020 0.021
Chain 1: 2800 -7739.748 0.020 0.021
Chain 1: 2900 -7444.911 0.023 0.022
Chain 1: 3000 -7598.161 0.023 0.022
Chain 1: 3100 -7597.323 0.022 0.022
Chain 1: 3200 -7798.127 0.020 0.022
Chain 1: 3300 -7536.253 0.020 0.022
Chain 1: 3400 -7759.149 0.020 0.026
Chain 1: 3500 -7508.717 0.021 0.026
Chain 1: 3600 -7574.270 0.022 0.026
Chain 1: 3700 -7533.957 0.022 0.026
Chain 1: 3800 -7537.608 0.020 0.026
Chain 1: 3900 -7481.424 0.016 0.020
Chain 1: 4000 -7469.168 0.015 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003124 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.24 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87055.893 1.000 1.000
Chain 1: 200 -14212.344 3.063 5.125
Chain 1: 300 -10487.757 2.160 1.000
Chain 1: 400 -11806.813 1.648 1.000
Chain 1: 500 -9432.911 1.369 0.355
Chain 1: 600 -9607.811 1.144 0.355
Chain 1: 700 -9135.149 0.988 0.252
Chain 1: 800 -8793.462 0.869 0.252
Chain 1: 900 -8835.475 0.773 0.112
Chain 1: 1000 -9116.441 0.699 0.112
Chain 1: 1100 -9222.951 0.600 0.052 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8803.296 0.092 0.048
Chain 1: 1300 -9160.562 0.061 0.039
Chain 1: 1400 -9120.199 0.050 0.039
Chain 1: 1500 -8986.736 0.026 0.031
Chain 1: 1600 -9086.913 0.025 0.031
Chain 1: 1700 -9146.412 0.021 0.015
Chain 1: 1800 -8706.304 0.022 0.015
Chain 1: 1900 -8811.007 0.023 0.015
Chain 1: 2000 -8794.277 0.020 0.012
Chain 1: 2100 -8920.607 0.020 0.014
Chain 1: 2200 -8709.655 0.018 0.014
Chain 1: 2300 -8804.183 0.015 0.012
Chain 1: 2400 -8871.453 0.015 0.012
Chain 1: 2500 -8819.551 0.014 0.011
Chain 1: 2600 -8831.936 0.013 0.011
Chain 1: 2700 -8740.226 0.014 0.011
Chain 1: 2800 -8688.559 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003237 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.37 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8396608.703 1.000 1.000
Chain 1: 200 -1579481.513 2.658 4.316
Chain 1: 300 -889046.647 2.031 1.000
Chain 1: 400 -457037.388 1.759 1.000
Chain 1: 500 -357920.855 1.463 0.945
Chain 1: 600 -233219.854 1.308 0.945
Chain 1: 700 -119761.476 1.257 0.945
Chain 1: 800 -87051.703 1.147 0.945
Chain 1: 900 -67442.609 1.051 0.777
Chain 1: 1000 -52274.442 0.975 0.777
Chain 1: 1100 -39776.579 0.907 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38962.395 0.477 0.376
Chain 1: 1300 -26927.358 0.444 0.376
Chain 1: 1400 -26650.499 0.351 0.314
Chain 1: 1500 -23239.620 0.338 0.314
Chain 1: 1600 -22457.955 0.288 0.291
Chain 1: 1700 -21331.779 0.198 0.290
Chain 1: 1800 -21276.537 0.161 0.147
Chain 1: 1900 -21603.290 0.133 0.053
Chain 1: 2000 -20113.642 0.112 0.053
Chain 1: 2100 -20352.028 0.082 0.035
Chain 1: 2200 -20578.877 0.081 0.035
Chain 1: 2300 -20195.593 0.038 0.019
Chain 1: 2400 -19967.514 0.038 0.019
Chain 1: 2500 -19769.514 0.024 0.015
Chain 1: 2600 -19399.151 0.023 0.015
Chain 1: 2700 -19356.016 0.018 0.012
Chain 1: 2800 -19072.653 0.019 0.015
Chain 1: 2900 -19354.119 0.019 0.015
Chain 1: 3000 -19340.254 0.011 0.012
Chain 1: 3100 -19425.317 0.011 0.011
Chain 1: 3200 -19115.654 0.011 0.015
Chain 1: 3300 -19320.679 0.010 0.011
Chain 1: 3400 -18794.972 0.012 0.015
Chain 1: 3500 -19407.783 0.014 0.015
Chain 1: 3600 -18713.239 0.016 0.015
Chain 1: 3700 -19100.955 0.018 0.016
Chain 1: 3800 -18058.751 0.022 0.020
Chain 1: 3900 -18054.856 0.021 0.020
Chain 1: 4000 -18172.148 0.021 0.020
Chain 1: 4100 -18085.819 0.021 0.020
Chain 1: 4200 -17901.676 0.021 0.020
Chain 1: 4300 -18040.353 0.020 0.020
Chain 1: 4400 -17996.836 0.018 0.010
Chain 1: 4500 -17899.315 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001285 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.85 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12027.529 1.000 1.000
Chain 1: 200 -8940.955 0.673 1.000
Chain 1: 300 -7912.795 0.492 0.345
Chain 1: 400 -8013.470 0.372 0.345
Chain 1: 500 -7867.158 0.301 0.130
Chain 1: 600 -7735.730 0.254 0.130
Chain 1: 700 -7673.512 0.219 0.019
Chain 1: 800 -7683.665 0.192 0.019
Chain 1: 900 -7694.916 0.170 0.017
Chain 1: 1000 -7730.496 0.154 0.017
Chain 1: 1100 -7789.237 0.055 0.013
Chain 1: 1200 -7683.653 0.021 0.013
Chain 1: 1300 -7704.956 0.009 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001364 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.64 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -45599.680 1.000 1.000
Chain 1: 200 -15027.498 1.517 2.034
Chain 1: 300 -8492.657 1.268 1.000
Chain 1: 400 -8337.255 0.956 1.000
Chain 1: 500 -8158.333 0.769 0.769
Chain 1: 600 -7769.983 0.649 0.769
Chain 1: 700 -7775.677 0.556 0.050
Chain 1: 800 -8039.987 0.491 0.050
Chain 1: 900 -7859.347 0.439 0.033
Chain 1: 1000 -7720.547 0.397 0.033
Chain 1: 1100 -7720.034 0.297 0.023
Chain 1: 1200 -7583.415 0.095 0.022
Chain 1: 1300 -7607.559 0.019 0.019
Chain 1: 1400 -7890.931 0.020 0.022
Chain 1: 1500 -7609.493 0.022 0.023
Chain 1: 1600 -7512.985 0.018 0.018
Chain 1: 1700 -7515.771 0.018 0.018
Chain 1: 1800 -7532.399 0.015 0.018
Chain 1: 1900 -7524.947 0.013 0.013
Chain 1: 2000 -7572.810 0.012 0.006 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003082 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.82 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86490.455 1.000 1.000
Chain 1: 200 -13057.161 3.312 5.624
Chain 1: 300 -9526.083 2.332 1.000
Chain 1: 400 -10331.663 1.768 1.000
Chain 1: 500 -8425.621 1.460 0.371
Chain 1: 600 -8218.376 1.221 0.371
Chain 1: 700 -8331.903 1.048 0.226
Chain 1: 800 -8872.822 0.925 0.226
Chain 1: 900 -8385.120 0.829 0.078
Chain 1: 1000 -8156.304 0.748 0.078
Chain 1: 1100 -8440.840 0.652 0.061 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8181.630 0.093 0.058
Chain 1: 1300 -8286.064 0.057 0.034
Chain 1: 1400 -8274.305 0.049 0.032
Chain 1: 1500 -8171.977 0.028 0.028
Chain 1: 1600 -8268.103 0.026 0.028
Chain 1: 1700 -8367.887 0.026 0.028
Chain 1: 1800 -7977.986 0.025 0.028
Chain 1: 1900 -8079.039 0.020 0.013
Chain 1: 2000 -8049.066 0.018 0.013
Chain 1: 2100 -8188.060 0.016 0.013
Chain 1: 2200 -7969.408 0.016 0.013
Chain 1: 2300 -8111.664 0.016 0.013
Chain 1: 2400 -7996.639 0.018 0.014
Chain 1: 2500 -8055.591 0.017 0.014
Chain 1: 2600 -8069.925 0.016 0.014
Chain 1: 2700 -7992.223 0.016 0.014
Chain 1: 2800 -7973.640 0.011 0.013
Chain 1: 2900 -7985.012 0.010 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003156 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.56 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8417319.228 1.000 1.000
Chain 1: 200 -1589092.083 2.648 4.297
Chain 1: 300 -890954.762 2.027 1.000
Chain 1: 400 -456982.591 1.758 1.000
Chain 1: 500 -356837.617 1.462 0.950
Chain 1: 600 -231716.976 1.308 0.950
Chain 1: 700 -118317.603 1.258 0.950
Chain 1: 800 -85632.132 1.149 0.950
Chain 1: 900 -66052.153 1.054 0.784
Chain 1: 1000 -50910.966 0.978 0.784
Chain 1: 1100 -38451.816 0.911 0.540 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37630.311 0.483 0.382
Chain 1: 1300 -25663.026 0.452 0.382
Chain 1: 1400 -25386.271 0.358 0.324
Chain 1: 1500 -21993.992 0.345 0.324
Chain 1: 1600 -21215.518 0.295 0.297
Chain 1: 1700 -20099.163 0.205 0.296
Chain 1: 1800 -20045.206 0.167 0.154
Chain 1: 1900 -20370.696 0.139 0.056
Chain 1: 2000 -18888.144 0.117 0.056
Chain 1: 2100 -19126.210 0.086 0.037
Chain 1: 2200 -19351.414 0.084 0.037
Chain 1: 2300 -18969.869 0.040 0.020
Chain 1: 2400 -18742.280 0.040 0.020
Chain 1: 2500 -18543.981 0.026 0.016
Chain 1: 2600 -18175.224 0.024 0.016
Chain 1: 2700 -18132.486 0.019 0.012
Chain 1: 2800 -17849.532 0.020 0.016
Chain 1: 2900 -18130.322 0.020 0.015
Chain 1: 3000 -18116.648 0.012 0.012
Chain 1: 3100 -18201.532 0.011 0.012
Chain 1: 3200 -17892.739 0.012 0.015
Chain 1: 3300 -18097.029 0.011 0.012
Chain 1: 3400 -17572.808 0.013 0.015
Chain 1: 3500 -18183.313 0.015 0.016
Chain 1: 3600 -17491.717 0.017 0.016
Chain 1: 3700 -17877.189 0.019 0.017
Chain 1: 3800 -16839.535 0.024 0.022
Chain 1: 3900 -16835.683 0.022 0.022
Chain 1: 4000 -16953.034 0.023 0.022
Chain 1: 4100 -16866.929 0.023 0.022
Chain 1: 4200 -16683.729 0.022 0.022
Chain 1: 4300 -16821.764 0.022 0.022
Chain 1: 4400 -16779.056 0.019 0.011
Chain 1: 4500 -16681.623 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001294 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.94 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12356.012 1.000 1.000
Chain 1: 200 -9156.784 0.675 1.000
Chain 1: 300 -8083.809 0.494 0.349
Chain 1: 400 -8205.995 0.374 0.349
Chain 1: 500 -8138.946 0.301 0.133
Chain 1: 600 -7989.249 0.254 0.133
Chain 1: 700 -7907.826 0.219 0.019
Chain 1: 800 -7915.663 0.192 0.019
Chain 1: 900 -7839.565 0.172 0.015
Chain 1: 1000 -7961.422 0.156 0.015
Chain 1: 1100 -7937.103 0.056 0.015
Chain 1: 1200 -7927.214 0.022 0.010
Chain 1: 1300 -7866.305 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001388 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.88 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57316.926 1.000 1.000
Chain 1: 200 -17521.260 1.636 2.271
Chain 1: 300 -8763.110 1.424 1.000
Chain 1: 400 -8289.180 1.082 1.000
Chain 1: 500 -8403.303 0.868 0.999
Chain 1: 600 -8569.947 0.727 0.999
Chain 1: 700 -8168.742 0.630 0.057
Chain 1: 800 -8191.558 0.552 0.057
Chain 1: 900 -7818.794 0.496 0.049
Chain 1: 1000 -7710.794 0.447 0.049
Chain 1: 1100 -7659.559 0.348 0.048
Chain 1: 1200 -7613.523 0.122 0.019
Chain 1: 1300 -7787.379 0.024 0.019
Chain 1: 1400 -7832.775 0.019 0.014
Chain 1: 1500 -7622.186 0.020 0.019
Chain 1: 1600 -7580.826 0.019 0.014
Chain 1: 1700 -7546.616 0.014 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003991 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 39.91 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85983.759 1.000 1.000
Chain 1: 200 -13539.950 3.175 5.350
Chain 1: 300 -9911.566 2.239 1.000
Chain 1: 400 -10874.523 1.701 1.000
Chain 1: 500 -8737.606 1.410 0.366
Chain 1: 600 -8559.178 1.178 0.366
Chain 1: 700 -8601.895 1.011 0.245
Chain 1: 800 -8822.361 0.888 0.245
Chain 1: 900 -8858.811 0.789 0.089
Chain 1: 1000 -8603.593 0.713 0.089
Chain 1: 1100 -8767.838 0.615 0.030 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8369.643 0.085 0.030
Chain 1: 1300 -8594.370 0.051 0.026
Chain 1: 1400 -8608.384 0.042 0.025
Chain 1: 1500 -8462.545 0.020 0.021
Chain 1: 1600 -8575.599 0.019 0.019
Chain 1: 1700 -8655.999 0.019 0.019
Chain 1: 1800 -8238.619 0.022 0.019
Chain 1: 1900 -8336.649 0.023 0.019
Chain 1: 2000 -8310.462 0.020 0.017
Chain 1: 2100 -8434.224 0.020 0.015
Chain 1: 2200 -8249.661 0.017 0.015
Chain 1: 2300 -8331.232 0.015 0.013
Chain 1: 2400 -8400.825 0.016 0.013
Chain 1: 2500 -8346.685 0.015 0.012
Chain 1: 2600 -8346.867 0.014 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003518 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.18 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8401234.600 1.000 1.000
Chain 1: 200 -1582680.536 2.654 4.308
Chain 1: 300 -890565.830 2.028 1.000
Chain 1: 400 -458008.040 1.757 1.000
Chain 1: 500 -358659.188 1.461 0.944
Chain 1: 600 -233405.303 1.307 0.944
Chain 1: 700 -119422.895 1.257 0.944
Chain 1: 800 -86650.558 1.147 0.944
Chain 1: 900 -66940.173 1.052 0.777
Chain 1: 1000 -51709.251 0.977 0.777
Chain 1: 1100 -39168.113 0.909 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38337.685 0.480 0.378
Chain 1: 1300 -26265.682 0.448 0.378
Chain 1: 1400 -25982.543 0.355 0.320
Chain 1: 1500 -22563.885 0.342 0.320
Chain 1: 1600 -21779.302 0.292 0.295
Chain 1: 1700 -20648.947 0.202 0.294
Chain 1: 1800 -20592.304 0.165 0.152
Chain 1: 1900 -20918.546 0.137 0.055
Chain 1: 2000 -19427.808 0.115 0.055
Chain 1: 2100 -19666.042 0.084 0.036
Chain 1: 2200 -19893.170 0.083 0.036
Chain 1: 2300 -19509.766 0.039 0.020
Chain 1: 2400 -19281.763 0.039 0.020
Chain 1: 2500 -19084.194 0.025 0.016
Chain 1: 2600 -18713.996 0.023 0.016
Chain 1: 2700 -18670.794 0.018 0.012
Chain 1: 2800 -18387.869 0.020 0.015
Chain 1: 2900 -18669.124 0.019 0.015
Chain 1: 3000 -18655.150 0.012 0.012
Chain 1: 3100 -18740.239 0.011 0.012
Chain 1: 3200 -18430.818 0.012 0.015
Chain 1: 3300 -18635.594 0.011 0.012
Chain 1: 3400 -18110.523 0.012 0.015
Chain 1: 3500 -18722.602 0.015 0.015
Chain 1: 3600 -18028.907 0.017 0.015
Chain 1: 3700 -18416.052 0.018 0.017
Chain 1: 3800 -17375.420 0.023 0.021
Chain 1: 3900 -17371.590 0.021 0.021
Chain 1: 4000 -17488.813 0.022 0.021
Chain 1: 4100 -17402.687 0.022 0.021
Chain 1: 4200 -17218.755 0.021 0.021
Chain 1: 4300 -17357.206 0.021 0.021
Chain 1: 4400 -17313.940 0.019 0.011
Chain 1: 4500 -17216.493 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001317 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.17 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49018.879 1.000 1.000
Chain 1: 200 -16517.376 1.484 1.968
Chain 1: 300 -19423.573 1.039 1.000
Chain 1: 400 -12198.915 0.927 1.000
Chain 1: 500 -38614.989 0.879 0.684
Chain 1: 600 -13802.724 1.032 1.000
Chain 1: 700 -16165.181 0.905 0.684
Chain 1: 800 -13128.368 0.821 0.684
Chain 1: 900 -12995.996 0.731 0.592
Chain 1: 1000 -14661.473 0.669 0.592
Chain 1: 1100 -14380.753 0.571 0.231 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -10032.332 0.418 0.231
Chain 1: 1300 -9745.591 0.406 0.231
Chain 1: 1400 -12566.445 0.369 0.224
Chain 1: 1500 -10858.661 0.316 0.157
Chain 1: 1600 -11956.766 0.146 0.146
Chain 1: 1700 -10360.101 0.147 0.154
Chain 1: 1800 -9936.491 0.128 0.114
Chain 1: 1900 -12834.353 0.149 0.154
Chain 1: 2000 -10142.427 0.164 0.157
Chain 1: 2100 -11791.461 0.176 0.157
Chain 1: 2200 -10063.642 0.150 0.157
Chain 1: 2300 -9146.570 0.157 0.157
Chain 1: 2400 -9939.322 0.143 0.154
Chain 1: 2500 -8978.687 0.138 0.140
Chain 1: 2600 -9699.409 0.136 0.140
Chain 1: 2700 -10443.170 0.128 0.107
Chain 1: 2800 -10100.758 0.127 0.107
Chain 1: 2900 -15281.994 0.138 0.107
Chain 1: 3000 -8903.098 0.183 0.107
Chain 1: 3100 -10145.996 0.182 0.107
Chain 1: 3200 -14221.000 0.193 0.107
Chain 1: 3300 -9973.961 0.226 0.123
Chain 1: 3400 -13519.820 0.244 0.262
Chain 1: 3500 -12855.096 0.238 0.262
Chain 1: 3600 -9984.018 0.260 0.287
Chain 1: 3700 -11923.399 0.269 0.287
Chain 1: 3800 -8587.002 0.304 0.288
Chain 1: 3900 -9013.905 0.275 0.287
Chain 1: 4000 -12671.462 0.232 0.287
Chain 1: 4100 -9134.577 0.259 0.288
Chain 1: 4200 -9329.554 0.232 0.288
Chain 1: 4300 -9702.826 0.194 0.262
Chain 1: 4400 -8925.373 0.176 0.163
Chain 1: 4500 -8805.526 0.172 0.163
Chain 1: 4600 -12292.324 0.172 0.163
Chain 1: 4700 -12302.881 0.156 0.087
Chain 1: 4800 -8775.252 0.157 0.087
Chain 1: 4900 -9118.226 0.156 0.087
Chain 1: 5000 -12663.895 0.155 0.087
Chain 1: 5100 -8617.706 0.163 0.087
Chain 1: 5200 -8603.546 0.161 0.087
Chain 1: 5300 -9201.160 0.164 0.087
Chain 1: 5400 -9056.007 0.157 0.065
Chain 1: 5500 -8645.876 0.160 0.065
Chain 1: 5600 -8860.276 0.134 0.047
Chain 1: 5700 -8912.716 0.135 0.047
Chain 1: 5800 -8677.236 0.097 0.038
Chain 1: 5900 -10785.370 0.113 0.047
Chain 1: 6000 -8550.934 0.111 0.047
Chain 1: 6100 -8744.970 0.067 0.027
Chain 1: 6200 -8739.465 0.067 0.027
Chain 1: 6300 -12721.926 0.091 0.027
Chain 1: 6400 -14898.481 0.104 0.047
Chain 1: 6500 -9715.127 0.153 0.146
Chain 1: 6600 -9443.221 0.153 0.146
Chain 1: 6700 -9376.460 0.154 0.146
Chain 1: 6800 -8399.460 0.162 0.146
Chain 1: 6900 -8411.272 0.143 0.116
Chain 1: 7000 -16966.763 0.167 0.116
Chain 1: 7100 -11252.886 0.216 0.146
Chain 1: 7200 -11474.646 0.218 0.146
Chain 1: 7300 -9958.486 0.202 0.146
Chain 1: 7400 -11047.338 0.197 0.116
Chain 1: 7500 -10907.573 0.145 0.099
Chain 1: 7600 -8582.997 0.169 0.116
Chain 1: 7700 -9991.214 0.182 0.141
Chain 1: 7800 -9484.870 0.176 0.141
Chain 1: 7900 -8345.182 0.190 0.141
Chain 1: 8000 -8268.509 0.140 0.137
Chain 1: 8100 -8643.555 0.094 0.099
Chain 1: 8200 -8985.250 0.096 0.099
Chain 1: 8300 -9442.222 0.085 0.053
Chain 1: 8400 -9660.766 0.078 0.048
Chain 1: 8500 -8326.496 0.092 0.053
Chain 1: 8600 -9444.151 0.077 0.053
Chain 1: 8700 -11397.983 0.080 0.053
Chain 1: 8800 -8270.614 0.113 0.118
Chain 1: 8900 -8674.713 0.104 0.048
Chain 1: 9000 -8563.112 0.104 0.048
Chain 1: 9100 -8559.633 0.100 0.048
Chain 1: 9200 -8407.805 0.098 0.048
Chain 1: 9300 -8255.781 0.095 0.047
Chain 1: 9400 -8144.244 0.094 0.047
Chain 1: 9500 -8271.787 0.079 0.018
Chain 1: 9600 -8426.895 0.069 0.018
Chain 1: 9700 -8123.438 0.056 0.018
Chain 1: 9800 -8577.483 0.023 0.018
Chain 1: 9900 -9183.022 0.025 0.018
Chain 1: 10000 -9357.117 0.026 0.018
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001381 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.81 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57078.889 1.000 1.000
Chain 1: 200 -17541.099 1.627 2.254
Chain 1: 300 -8819.981 1.414 1.000
Chain 1: 400 -8452.502 1.072 1.000
Chain 1: 500 -8312.299 0.861 0.989
Chain 1: 600 -8739.319 0.725 0.989
Chain 1: 700 -8459.726 0.626 0.049
Chain 1: 800 -8217.355 0.552 0.049
Chain 1: 900 -7993.446 0.494 0.043
Chain 1: 1000 -7859.923 0.446 0.043
Chain 1: 1100 -7867.789 0.346 0.033
Chain 1: 1200 -7768.228 0.122 0.029
Chain 1: 1300 -7812.476 0.024 0.028
Chain 1: 1400 -7741.487 0.020 0.017
Chain 1: 1500 -7653.281 0.020 0.017
Chain 1: 1600 -7887.400 0.018 0.017
Chain 1: 1700 -7587.195 0.018 0.017
Chain 1: 1800 -7664.116 0.016 0.013
Chain 1: 1900 -7660.046 0.014 0.012
Chain 1: 2000 -7708.942 0.013 0.010
Chain 1: 2100 -7671.321 0.013 0.010
Chain 1: 2200 -7779.807 0.013 0.010
Chain 1: 2300 -7654.787 0.014 0.012
Chain 1: 2400 -7720.491 0.014 0.012
Chain 1: 2500 -7558.313 0.015 0.014
Chain 1: 2600 -7583.350 0.012 0.010
Chain 1: 2700 -7643.523 0.009 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003309 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.09 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86456.242 1.000 1.000
Chain 1: 200 -13573.859 3.185 5.369
Chain 1: 300 -9906.914 2.246 1.000
Chain 1: 400 -10902.413 1.708 1.000
Chain 1: 500 -8892.644 1.411 0.370
Chain 1: 600 -8357.617 1.187 0.370
Chain 1: 700 -8324.016 1.018 0.226
Chain 1: 800 -9037.513 0.900 0.226
Chain 1: 900 -8673.837 0.805 0.091
Chain 1: 1000 -8668.326 0.725 0.091
Chain 1: 1100 -8573.458 0.626 0.079 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8340.148 0.092 0.064
Chain 1: 1300 -8586.905 0.057 0.042
Chain 1: 1400 -8558.467 0.049 0.029
Chain 1: 1500 -8456.655 0.027 0.028
Chain 1: 1600 -8562.694 0.022 0.012
Chain 1: 1700 -8644.622 0.023 0.012
Chain 1: 1800 -8219.217 0.020 0.012
Chain 1: 1900 -8321.317 0.017 0.012
Chain 1: 2000 -8295.874 0.017 0.012
Chain 1: 2100 -8422.029 0.018 0.012
Chain 1: 2200 -8222.959 0.017 0.012
Chain 1: 2300 -8316.207 0.015 0.012
Chain 1: 2400 -8384.667 0.016 0.012
Chain 1: 2500 -8330.911 0.015 0.012
Chain 1: 2600 -8332.723 0.014 0.011
Chain 1: 2700 -8249.252 0.014 0.011
Chain 1: 2800 -8208.523 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003435 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.35 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8400219.680 1.000 1.000
Chain 1: 200 -1582825.722 2.654 4.307
Chain 1: 300 -891044.830 2.028 1.000
Chain 1: 400 -457595.007 1.758 1.000
Chain 1: 500 -358100.671 1.462 0.947
Chain 1: 600 -233245.270 1.307 0.947
Chain 1: 700 -119464.151 1.257 0.947
Chain 1: 800 -86616.339 1.147 0.947
Chain 1: 900 -66946.223 1.052 0.776
Chain 1: 1000 -51723.189 0.976 0.776
Chain 1: 1100 -39178.062 0.908 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38354.633 0.480 0.379
Chain 1: 1300 -26291.477 0.448 0.379
Chain 1: 1400 -26008.824 0.354 0.320
Chain 1: 1500 -22589.935 0.342 0.320
Chain 1: 1600 -21804.686 0.292 0.294
Chain 1: 1700 -20676.426 0.202 0.294
Chain 1: 1800 -20620.123 0.164 0.151
Chain 1: 1900 -20946.410 0.137 0.055
Chain 1: 2000 -19455.993 0.115 0.055
Chain 1: 2100 -19694.617 0.084 0.036
Chain 1: 2200 -19921.157 0.083 0.036
Chain 1: 2300 -19538.235 0.039 0.020
Chain 1: 2400 -19310.277 0.039 0.020
Chain 1: 2500 -19112.180 0.025 0.016
Chain 1: 2600 -18742.329 0.023 0.016
Chain 1: 2700 -18699.295 0.018 0.012
Chain 1: 2800 -18416.025 0.019 0.015
Chain 1: 2900 -18697.403 0.019 0.015
Chain 1: 3000 -18683.606 0.012 0.012
Chain 1: 3100 -18768.586 0.011 0.012
Chain 1: 3200 -18459.194 0.012 0.015
Chain 1: 3300 -18663.996 0.011 0.012
Chain 1: 3400 -18138.719 0.012 0.015
Chain 1: 3500 -18750.825 0.015 0.015
Chain 1: 3600 -18057.269 0.017 0.015
Chain 1: 3700 -18444.271 0.018 0.017
Chain 1: 3800 -17403.494 0.023 0.021
Chain 1: 3900 -17399.625 0.021 0.021
Chain 1: 4000 -17516.950 0.022 0.021
Chain 1: 4100 -17430.628 0.022 0.021
Chain 1: 4200 -17246.817 0.021 0.021
Chain 1: 4300 -17385.287 0.021 0.021
Chain 1: 4400 -17342.065 0.018 0.011
Chain 1: 4500 -17244.550 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001415 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.15 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48126.838 1.000 1.000
Chain 1: 200 -22211.679 1.083 1.167
Chain 1: 300 -18500.009 0.789 1.000
Chain 1: 400 -30051.199 0.688 1.000
Chain 1: 500 -18380.874 0.677 0.635
Chain 1: 600 -16201.147 0.587 0.635
Chain 1: 700 -11008.125 0.570 0.472
Chain 1: 800 -9875.536 0.513 0.472
Chain 1: 900 -14315.141 0.491 0.384
Chain 1: 1000 -11498.535 0.466 0.384
Chain 1: 1100 -9430.157 0.388 0.310
Chain 1: 1200 -20765.152 0.326 0.310
Chain 1: 1300 -11588.230 0.385 0.384
Chain 1: 1400 -11841.117 0.349 0.310
Chain 1: 1500 -9602.413 0.309 0.245
Chain 1: 1600 -11185.691 0.309 0.245
Chain 1: 1700 -9461.439 0.281 0.233
Chain 1: 1800 -11764.479 0.289 0.233
Chain 1: 1900 -9573.418 0.280 0.229
Chain 1: 2000 -11169.978 0.270 0.219
Chain 1: 2100 -16023.515 0.279 0.229
Chain 1: 2200 -9397.205 0.295 0.229
Chain 1: 2300 -8541.016 0.225 0.196
Chain 1: 2400 -10016.692 0.238 0.196
Chain 1: 2500 -10652.927 0.221 0.182
Chain 1: 2600 -8487.209 0.232 0.196
Chain 1: 2700 -15629.077 0.260 0.229
Chain 1: 2800 -9708.254 0.301 0.255
Chain 1: 2900 -8898.566 0.287 0.255
Chain 1: 3000 -8248.711 0.281 0.255
Chain 1: 3100 -9243.280 0.261 0.147
Chain 1: 3200 -12384.430 0.216 0.147
Chain 1: 3300 -14790.270 0.222 0.163
Chain 1: 3400 -9719.051 0.260 0.254
Chain 1: 3500 -9772.455 0.254 0.254
Chain 1: 3600 -9461.673 0.232 0.163
Chain 1: 3700 -8772.610 0.194 0.108
Chain 1: 3800 -9439.866 0.140 0.091
Chain 1: 3900 -13193.161 0.160 0.108
Chain 1: 4000 -8967.950 0.199 0.163
Chain 1: 4100 -8418.335 0.195 0.163
Chain 1: 4200 -10765.787 0.191 0.163
Chain 1: 4300 -9395.597 0.189 0.146
Chain 1: 4400 -8404.188 0.149 0.118
Chain 1: 4500 -8861.483 0.154 0.118
Chain 1: 4600 -13511.618 0.185 0.146
Chain 1: 4700 -11101.147 0.199 0.217
Chain 1: 4800 -8139.055 0.228 0.218
Chain 1: 4900 -10852.178 0.225 0.218
Chain 1: 5000 -9161.718 0.196 0.217
Chain 1: 5100 -16397.128 0.233 0.218
Chain 1: 5200 -14119.089 0.228 0.217
Chain 1: 5300 -11084.805 0.241 0.250
Chain 1: 5400 -7956.304 0.268 0.274
Chain 1: 5500 -8105.871 0.265 0.274
Chain 1: 5600 -10263.680 0.251 0.250
Chain 1: 5700 -9148.309 0.242 0.250
Chain 1: 5800 -8139.469 0.218 0.210
Chain 1: 5900 -13715.175 0.234 0.210
Chain 1: 6000 -8917.106 0.269 0.274
Chain 1: 6100 -8000.056 0.236 0.210
Chain 1: 6200 -7885.872 0.222 0.210
Chain 1: 6300 -8666.863 0.203 0.124
Chain 1: 6400 -10576.434 0.182 0.124
Chain 1: 6500 -8774.969 0.201 0.181
Chain 1: 6600 -8187.299 0.187 0.124
Chain 1: 6700 -11794.687 0.205 0.181
Chain 1: 6800 -7828.469 0.243 0.205
Chain 1: 6900 -8002.470 0.205 0.181
Chain 1: 7000 -8403.055 0.156 0.115
Chain 1: 7100 -7814.887 0.152 0.090
Chain 1: 7200 -10254.111 0.174 0.181
Chain 1: 7300 -8503.874 0.186 0.205
Chain 1: 7400 -7922.111 0.175 0.205
Chain 1: 7500 -7900.298 0.155 0.075
Chain 1: 7600 -8547.284 0.155 0.076
Chain 1: 7700 -8059.768 0.131 0.075
Chain 1: 7800 -11302.199 0.109 0.075
Chain 1: 7900 -7817.988 0.151 0.076
Chain 1: 8000 -11195.876 0.177 0.206
Chain 1: 8100 -10761.264 0.173 0.206
Chain 1: 8200 -7745.325 0.188 0.206
Chain 1: 8300 -7663.132 0.169 0.076
Chain 1: 8400 -10524.287 0.189 0.272
Chain 1: 8500 -7733.848 0.224 0.287
Chain 1: 8600 -7885.245 0.219 0.287
Chain 1: 8700 -9133.753 0.226 0.287
Chain 1: 8800 -9170.861 0.198 0.272
Chain 1: 8900 -10939.879 0.170 0.162
Chain 1: 9000 -8116.075 0.174 0.162
Chain 1: 9100 -9536.565 0.185 0.162
Chain 1: 9200 -8292.639 0.161 0.150
Chain 1: 9300 -7940.331 0.165 0.150
Chain 1: 9400 -8101.982 0.139 0.149
Chain 1: 9500 -8009.390 0.104 0.137
Chain 1: 9600 -7875.819 0.104 0.137
Chain 1: 9700 -10066.069 0.112 0.149
Chain 1: 9800 -7871.383 0.140 0.150
Chain 1: 9900 -10384.564 0.148 0.150
Chain 1: 10000 -9250.826 0.125 0.149
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001399 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.99 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56474.636 1.000 1.000
Chain 1: 200 -16724.023 1.688 2.377
Chain 1: 300 -8370.130 1.458 1.000
Chain 1: 400 -8471.397 1.097 1.000
Chain 1: 500 -7876.143 0.892 0.998
Chain 1: 600 -8181.493 0.750 0.998
Chain 1: 700 -7567.069 0.654 0.081
Chain 1: 800 -7869.407 0.577 0.081
Chain 1: 900 -7738.569 0.515 0.076
Chain 1: 1000 -7693.270 0.464 0.076
Chain 1: 1100 -7549.817 0.366 0.038
Chain 1: 1200 -7449.074 0.130 0.037
Chain 1: 1300 -7525.725 0.031 0.019
Chain 1: 1400 -7724.683 0.032 0.026
Chain 1: 1500 -7481.050 0.028 0.026
Chain 1: 1600 -7395.417 0.026 0.019
Chain 1: 1700 -7371.945 0.018 0.017
Chain 1: 1800 -7396.703 0.014 0.014
Chain 1: 1900 -7460.183 0.013 0.012
Chain 1: 2000 -7469.316 0.013 0.012
Chain 1: 2100 -7398.452 0.012 0.010
Chain 1: 2200 -7512.826 0.012 0.010
Chain 1: 2300 -7450.482 0.012 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004035 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 40.35 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85908.830 1.000 1.000
Chain 1: 200 -12801.299 3.355 5.711
Chain 1: 300 -9306.035 2.362 1.000
Chain 1: 400 -10233.899 1.794 1.000
Chain 1: 500 -8139.204 1.487 0.376
Chain 1: 600 -7921.547 1.244 0.376
Chain 1: 700 -7935.432 1.066 0.257
Chain 1: 800 -8167.415 0.937 0.257
Chain 1: 900 -8183.459 0.833 0.091
Chain 1: 1000 -7925.593 0.753 0.091
Chain 1: 1100 -8185.922 0.656 0.033 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -7921.151 0.088 0.033
Chain 1: 1300 -8066.330 0.052 0.032
Chain 1: 1400 -8061.765 0.043 0.028
Chain 1: 1500 -7978.427 0.019 0.027
Chain 1: 1600 -8058.744 0.017 0.018
Chain 1: 1700 -8164.043 0.018 0.018
Chain 1: 1800 -7782.443 0.020 0.018
Chain 1: 1900 -7880.679 0.021 0.018
Chain 1: 2000 -7850.901 0.018 0.013
Chain 1: 2100 -7996.188 0.017 0.013
Chain 1: 2200 -7773.829 0.016 0.013
Chain 1: 2300 -7907.694 0.016 0.013
Chain 1: 2400 -7798.324 0.018 0.014
Chain 1: 2500 -7857.795 0.017 0.014
Chain 1: 2600 -7870.744 0.017 0.014
Chain 1: 2700 -7791.704 0.016 0.014
Chain 1: 2800 -7776.890 0.012 0.012
Chain 1: 2900 -7765.249 0.010 0.010
Chain 1: 3000 -7781.667 0.010 0.010
Chain 1: 3100 -7844.715 0.009 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003166 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.66 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8407678.947 1.000 1.000
Chain 1: 200 -1582912.611 2.656 4.312
Chain 1: 300 -889948.746 2.030 1.000
Chain 1: 400 -457297.045 1.759 1.000
Chain 1: 500 -357547.327 1.463 0.946
Chain 1: 600 -232360.458 1.309 0.946
Chain 1: 700 -118511.084 1.259 0.946
Chain 1: 800 -85721.706 1.150 0.946
Chain 1: 900 -66047.836 1.055 0.779
Chain 1: 1000 -50825.479 0.979 0.779
Chain 1: 1100 -38302.336 0.912 0.539 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37470.855 0.483 0.383
Chain 1: 1300 -25441.832 0.453 0.383
Chain 1: 1400 -25158.185 0.359 0.327
Chain 1: 1500 -21750.767 0.347 0.327
Chain 1: 1600 -20968.012 0.297 0.300
Chain 1: 1700 -19844.207 0.206 0.298
Chain 1: 1800 -19788.558 0.168 0.157
Chain 1: 1900 -20113.853 0.140 0.057
Chain 1: 2000 -18628.216 0.118 0.057
Chain 1: 2100 -18866.167 0.087 0.037
Chain 1: 2200 -19091.953 0.086 0.037
Chain 1: 2300 -18709.994 0.041 0.020
Chain 1: 2400 -18482.432 0.041 0.020
Chain 1: 2500 -18284.504 0.026 0.016
Chain 1: 2600 -17915.425 0.024 0.016
Chain 1: 2700 -17872.639 0.019 0.013
Chain 1: 2800 -17589.899 0.020 0.016
Chain 1: 2900 -17870.790 0.020 0.016
Chain 1: 3000 -17856.936 0.012 0.013
Chain 1: 3100 -17941.818 0.012 0.012
Chain 1: 3200 -17633.031 0.012 0.016
Chain 1: 3300 -17837.375 0.011 0.012
Chain 1: 3400 -17313.282 0.013 0.016
Chain 1: 3500 -17923.640 0.015 0.016
Chain 1: 3600 -17232.311 0.017 0.016
Chain 1: 3700 -17617.638 0.019 0.018
Chain 1: 3800 -16580.419 0.024 0.022
Chain 1: 3900 -16576.671 0.022 0.022
Chain 1: 4000 -16693.952 0.023 0.022
Chain 1: 4100 -16607.853 0.023 0.022
Chain 1: 4200 -16424.797 0.022 0.022
Chain 1: 4300 -16562.685 0.022 0.022
Chain 1: 4400 -16520.055 0.019 0.011
Chain 1: 4500 -16422.716 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003763 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 37.63 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -11875.857 1.000 1.000
Chain 1: 200 -8900.827 0.667 1.000
Chain 1: 300 -7875.438 0.488 0.334
Chain 1: 400 -7913.982 0.367 0.334
Chain 1: 500 -7773.106 0.297 0.130
Chain 1: 600 -7718.945 0.249 0.130
Chain 1: 700 -7660.829 0.215 0.018
Chain 1: 800 -7718.198 0.189 0.018
Chain 1: 900 -7888.417 0.170 0.018
Chain 1: 1000 -7697.690 0.156 0.022
Chain 1: 1100 -7752.233 0.056 0.018
Chain 1: 1200 -7663.405 0.024 0.012
Chain 1: 1300 -7693.946 0.011 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00162 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 16.2 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57450.639 1.000 1.000
Chain 1: 200 -17081.318 1.682 2.363
Chain 1: 300 -8443.565 1.462 1.023
Chain 1: 400 -8504.925 1.098 1.023
Chain 1: 500 -8422.742 0.881 1.000
Chain 1: 600 -8228.011 0.738 1.000
Chain 1: 700 -7922.799 0.638 0.039
Chain 1: 800 -8028.385 0.560 0.039
Chain 1: 900 -7809.656 0.501 0.028
Chain 1: 1000 -7730.874 0.452 0.028
Chain 1: 1100 -7725.627 0.352 0.024
Chain 1: 1200 -7580.086 0.117 0.019
Chain 1: 1300 -7674.711 0.016 0.013
Chain 1: 1400 -7695.938 0.016 0.013
Chain 1: 1500 -7595.524 0.016 0.013
Chain 1: 1600 -7510.140 0.015 0.013
Chain 1: 1700 -7495.928 0.011 0.012
Chain 1: 1800 -7516.519 0.010 0.011
Chain 1: 1900 -7583.107 0.008 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003255 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.55 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85764.062 1.000 1.000
Chain 1: 200 -12935.684 3.315 5.630
Chain 1: 300 -9454.518 2.333 1.000
Chain 1: 400 -10220.204 1.768 1.000
Chain 1: 500 -8305.723 1.461 0.368
Chain 1: 600 -8182.399 1.220 0.368
Chain 1: 700 -8363.949 1.049 0.231
Chain 1: 800 -8780.482 0.923 0.231
Chain 1: 900 -8341.554 0.827 0.075
Chain 1: 1000 -8115.451 0.747 0.075
Chain 1: 1100 -8391.853 0.650 0.053 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8158.838 0.090 0.047
Chain 1: 1300 -8260.341 0.054 0.033
Chain 1: 1400 -8241.915 0.047 0.029
Chain 1: 1500 -8152.927 0.025 0.028
Chain 1: 1600 -8230.701 0.025 0.028
Chain 1: 1700 -8330.308 0.024 0.028
Chain 1: 1800 -7961.839 0.024 0.028
Chain 1: 1900 -8057.656 0.019 0.012
Chain 1: 2000 -8029.300 0.017 0.012
Chain 1: 2100 -8177.844 0.016 0.012
Chain 1: 2200 -7953.328 0.015 0.012
Chain 1: 2300 -8034.987 0.015 0.012
Chain 1: 2400 -8102.383 0.016 0.012
Chain 1: 2500 -8063.933 0.015 0.012
Chain 1: 2600 -8057.615 0.014 0.012
Chain 1: 2700 -7969.974 0.014 0.011
Chain 1: 2800 -7956.675 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003519 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.19 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8417704.700 1.000 1.000
Chain 1: 200 -1588525.070 2.650 4.299
Chain 1: 300 -891628.651 2.027 1.000
Chain 1: 400 -457421.320 1.757 1.000
Chain 1: 500 -357438.895 1.462 0.949
Chain 1: 600 -232164.766 1.308 0.949
Chain 1: 700 -118492.401 1.258 0.949
Chain 1: 800 -85712.033 1.149 0.949
Chain 1: 900 -66076.459 1.054 0.782
Chain 1: 1000 -50884.516 0.979 0.782
Chain 1: 1100 -38382.851 0.911 0.540 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37552.441 0.484 0.382
Chain 1: 1300 -25550.775 0.452 0.382
Chain 1: 1400 -25268.013 0.359 0.326
Chain 1: 1500 -21867.421 0.346 0.326
Chain 1: 1600 -21085.952 0.296 0.299
Chain 1: 1700 -19966.165 0.206 0.297
Chain 1: 1800 -19910.997 0.168 0.156
Chain 1: 1900 -20236.251 0.139 0.056
Chain 1: 2000 -18752.458 0.118 0.056
Chain 1: 2100 -18990.397 0.086 0.037
Chain 1: 2200 -19215.778 0.085 0.037
Chain 1: 2300 -18834.194 0.040 0.020
Chain 1: 2400 -18606.740 0.040 0.020
Chain 1: 2500 -18408.564 0.026 0.016
Chain 1: 2600 -18039.935 0.024 0.016
Chain 1: 2700 -17997.215 0.019 0.013
Chain 1: 2800 -17714.482 0.020 0.016
Chain 1: 2900 -17995.204 0.020 0.016
Chain 1: 3000 -17981.475 0.012 0.013
Chain 1: 3100 -18066.326 0.011 0.012
Chain 1: 3200 -17757.685 0.012 0.016
Chain 1: 3300 -17961.866 0.011 0.012
Chain 1: 3400 -17437.951 0.013 0.016
Chain 1: 3500 -18048.003 0.015 0.016
Chain 1: 3600 -17357.073 0.017 0.016
Chain 1: 3700 -17742.113 0.019 0.017
Chain 1: 3800 -16705.449 0.024 0.022
Chain 1: 3900 -16701.678 0.022 0.022
Chain 1: 4000 -16818.993 0.023 0.022
Chain 1: 4100 -16732.957 0.023 0.022
Chain 1: 4200 -16549.983 0.022 0.022
Chain 1: 4300 -16687.831 0.022 0.022
Chain 1: 4400 -16645.324 0.019 0.011
Chain 1: 4500 -16547.976 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001352 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.52 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12811.830 1.000 1.000
Chain 1: 200 -9722.141 0.659 1.000
Chain 1: 300 -8216.213 0.500 0.318
Chain 1: 400 -8471.773 0.383 0.318
Chain 1: 500 -8351.704 0.309 0.183
Chain 1: 600 -8202.899 0.261 0.183
Chain 1: 700 -8296.892 0.225 0.030
Chain 1: 800 -8143.043 0.199 0.030
Chain 1: 900 -8194.872 0.178 0.019
Chain 1: 1000 -8142.088 0.161 0.019
Chain 1: 1100 -8237.048 0.062 0.018
Chain 1: 1200 -8127.860 0.031 0.014
Chain 1: 1300 -8059.533 0.014 0.013
Chain 1: 1400 -8081.453 0.011 0.012
Chain 1: 1500 -8177.478 0.011 0.012
Chain 1: 1600 -8088.732 0.010 0.011
Chain 1: 1700 -8059.052 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001395 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.95 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -62146.359 1.000 1.000
Chain 1: 200 -18294.566 1.698 2.397
Chain 1: 300 -9082.632 1.470 1.014
Chain 1: 400 -7885.647 1.141 1.014
Chain 1: 500 -7836.171 0.914 1.000
Chain 1: 600 -9395.863 0.789 1.000
Chain 1: 700 -7921.248 0.703 0.186
Chain 1: 800 -8273.299 0.621 0.186
Chain 1: 900 -8032.765 0.555 0.166
Chain 1: 1000 -7661.769 0.504 0.166
Chain 1: 1100 -7725.575 0.405 0.152
Chain 1: 1200 -8259.383 0.172 0.065
Chain 1: 1300 -7811.891 0.076 0.057
Chain 1: 1400 -7629.858 0.063 0.048
Chain 1: 1500 -7515.201 0.064 0.048
Chain 1: 1600 -7823.921 0.052 0.043
Chain 1: 1700 -7446.155 0.038 0.043
Chain 1: 1800 -7551.132 0.035 0.039
Chain 1: 1900 -7578.483 0.033 0.039
Chain 1: 2000 -7675.676 0.029 0.024
Chain 1: 2100 -7527.373 0.030 0.024
Chain 1: 2200 -7734.579 0.026 0.024
Chain 1: 2300 -7543.635 0.023 0.024
Chain 1: 2400 -7646.565 0.022 0.020
Chain 1: 2500 -7537.576 0.022 0.020
Chain 1: 2600 -7495.711 0.019 0.014
Chain 1: 2700 -7432.975 0.014 0.014
Chain 1: 2800 -7616.118 0.015 0.014
Chain 1: 2900 -7385.723 0.018 0.020
Chain 1: 3000 -7504.835 0.018 0.020
Chain 1: 3100 -7505.148 0.017 0.016
Chain 1: 3200 -7711.298 0.017 0.016
Chain 1: 3300 -7414.611 0.018 0.016
Chain 1: 3400 -7666.656 0.020 0.024
Chain 1: 3500 -7411.198 0.022 0.027
Chain 1: 3600 -7468.514 0.022 0.027
Chain 1: 3700 -7425.458 0.022 0.027
Chain 1: 3800 -7432.399 0.020 0.027
Chain 1: 3900 -7397.284 0.017 0.016
Chain 1: 4000 -7369.232 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003873 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 38.73 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86280.215 1.000 1.000
Chain 1: 200 -13982.452 3.085 5.171
Chain 1: 300 -10228.175 2.179 1.000
Chain 1: 400 -11800.825 1.668 1.000
Chain 1: 500 -8982.643 1.397 0.367
Chain 1: 600 -9737.359 1.177 0.367
Chain 1: 700 -8585.361 1.028 0.314
Chain 1: 800 -9277.035 0.909 0.314
Chain 1: 900 -9046.691 0.811 0.134
Chain 1: 1000 -8587.957 0.735 0.134
Chain 1: 1100 -9042.618 0.640 0.133 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8566.292 0.129 0.078
Chain 1: 1300 -8853.448 0.095 0.075
Chain 1: 1400 -8783.735 0.083 0.056
Chain 1: 1500 -8711.039 0.052 0.053
Chain 1: 1600 -8822.661 0.045 0.050
Chain 1: 1700 -8875.082 0.033 0.032
Chain 1: 1800 -8429.755 0.030 0.032
Chain 1: 1900 -8534.995 0.029 0.032
Chain 1: 2000 -8516.841 0.024 0.013
Chain 1: 2100 -8658.379 0.021 0.013
Chain 1: 2200 -8428.691 0.018 0.013
Chain 1: 2300 -8585.176 0.016 0.013
Chain 1: 2400 -8427.889 0.017 0.016
Chain 1: 2500 -8505.806 0.018 0.016
Chain 1: 2600 -8538.743 0.017 0.016
Chain 1: 2700 -8458.589 0.017 0.016
Chain 1: 2800 -8410.201 0.012 0.012
Chain 1: 2900 -8520.531 0.012 0.013
Chain 1: 3000 -8407.240 0.014 0.013
Chain 1: 3100 -8395.041 0.012 0.013
Chain 1: 3200 -8370.466 0.010 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003589 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.89 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8391132.666 1.000 1.000
Chain 1: 200 -1578620.745 2.658 4.315
Chain 1: 300 -890312.462 2.030 1.000
Chain 1: 400 -457909.195 1.758 1.000
Chain 1: 500 -358804.531 1.462 0.944
Chain 1: 600 -233995.638 1.307 0.944
Chain 1: 700 -120036.906 1.256 0.944
Chain 1: 800 -87185.973 1.146 0.944
Chain 1: 900 -67480.525 1.051 0.773
Chain 1: 1000 -52235.604 0.975 0.773
Chain 1: 1100 -39669.632 0.907 0.533 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38845.611 0.477 0.377
Chain 1: 1300 -26742.656 0.445 0.377
Chain 1: 1400 -26459.192 0.352 0.317
Chain 1: 1500 -23030.182 0.339 0.317
Chain 1: 1600 -22243.116 0.290 0.292
Chain 1: 1700 -21108.945 0.200 0.292
Chain 1: 1800 -21051.664 0.163 0.149
Chain 1: 1900 -21378.411 0.135 0.054
Chain 1: 2000 -19884.292 0.113 0.054
Chain 1: 2100 -20123.010 0.083 0.035
Chain 1: 2200 -20350.545 0.082 0.035
Chain 1: 2300 -19966.611 0.038 0.019
Chain 1: 2400 -19738.401 0.038 0.019
Chain 1: 2500 -19540.662 0.025 0.015
Chain 1: 2600 -19170.014 0.023 0.015
Chain 1: 2700 -19126.723 0.018 0.012
Chain 1: 2800 -18843.462 0.019 0.015
Chain 1: 2900 -19125.009 0.019 0.015
Chain 1: 3000 -19111.120 0.012 0.012
Chain 1: 3100 -19196.233 0.011 0.012
Chain 1: 3200 -18886.440 0.011 0.015
Chain 1: 3300 -19091.522 0.011 0.012
Chain 1: 3400 -18565.726 0.012 0.015
Chain 1: 3500 -19178.801 0.014 0.015
Chain 1: 3600 -18483.890 0.016 0.015
Chain 1: 3700 -18871.942 0.018 0.016
Chain 1: 3800 -17829.283 0.022 0.021
Chain 1: 3900 -17825.397 0.021 0.021
Chain 1: 4000 -17942.653 0.022 0.021
Chain 1: 4100 -17856.329 0.022 0.021
Chain 1: 4200 -17672.058 0.021 0.021
Chain 1: 4300 -17810.807 0.021 0.021
Chain 1: 4400 -17767.214 0.018 0.010
Chain 1: 4500 -17669.677 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00197 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 19.7 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48520.518 1.000 1.000
Chain 1: 200 -16490.556 1.471 1.942
Chain 1: 300 -18421.860 1.016 1.000
Chain 1: 400 -14219.451 0.836 1.000
Chain 1: 500 -11398.235 0.718 0.296
Chain 1: 600 -26016.219 0.692 0.562
Chain 1: 700 -15897.738 0.684 0.562
Chain 1: 800 -12335.862 0.635 0.562
Chain 1: 900 -10680.241 0.581 0.296
Chain 1: 1000 -11093.433 0.527 0.296
Chain 1: 1100 -12371.646 0.437 0.289
Chain 1: 1200 -15705.235 0.264 0.248
Chain 1: 1300 -10256.641 0.307 0.289
Chain 1: 1400 -16547.824 0.315 0.289
Chain 1: 1500 -12105.308 0.327 0.367
Chain 1: 1600 -30159.961 0.331 0.367
Chain 1: 1700 -10933.139 0.443 0.367
Chain 1: 1800 -13976.489 0.436 0.367
Chain 1: 1900 -9864.506 0.462 0.380
Chain 1: 2000 -12001.963 0.476 0.380
Chain 1: 2100 -9145.479 0.497 0.380
Chain 1: 2200 -11310.590 0.495 0.380
Chain 1: 2300 -9545.099 0.461 0.367
Chain 1: 2400 -8939.148 0.429 0.312
Chain 1: 2500 -10028.681 0.404 0.218
Chain 1: 2600 -10794.029 0.351 0.191
Chain 1: 2700 -8866.864 0.197 0.191
Chain 1: 2800 -8964.328 0.176 0.185
Chain 1: 2900 -9710.546 0.142 0.178
Chain 1: 3000 -9082.016 0.131 0.109
Chain 1: 3100 -9404.013 0.103 0.077
Chain 1: 3200 -9844.260 0.089 0.071
Chain 1: 3300 -9440.758 0.074 0.069
Chain 1: 3400 -10445.781 0.077 0.071
Chain 1: 3500 -11996.538 0.079 0.071
Chain 1: 3600 -10157.914 0.090 0.077
Chain 1: 3700 -9171.899 0.079 0.077
Chain 1: 3800 -8446.632 0.087 0.086
Chain 1: 3900 -10397.973 0.098 0.096
Chain 1: 4000 -11829.327 0.103 0.108
Chain 1: 4100 -8578.035 0.138 0.121
Chain 1: 4200 -9206.180 0.140 0.121
Chain 1: 4300 -12524.693 0.162 0.129
Chain 1: 4400 -12111.124 0.156 0.129
Chain 1: 4500 -8589.560 0.184 0.181
Chain 1: 4600 -9728.740 0.178 0.121
Chain 1: 4700 -10271.214 0.172 0.121
Chain 1: 4800 -10664.039 0.167 0.121
Chain 1: 4900 -8609.824 0.172 0.121
Chain 1: 5000 -13509.034 0.196 0.239
Chain 1: 5100 -8326.157 0.221 0.239
Chain 1: 5200 -9216.238 0.224 0.239
Chain 1: 5300 -11622.560 0.218 0.207
Chain 1: 5400 -9071.560 0.243 0.239
Chain 1: 5500 -11606.384 0.223 0.218
Chain 1: 5600 -8327.312 0.251 0.239
Chain 1: 5700 -14798.749 0.289 0.281
Chain 1: 5800 -8594.104 0.358 0.363
Chain 1: 5900 -14611.763 0.375 0.394
Chain 1: 6000 -8177.210 0.418 0.412
Chain 1: 6100 -9626.437 0.371 0.394
Chain 1: 6200 -8192.550 0.378 0.394
Chain 1: 6300 -12374.354 0.391 0.394
Chain 1: 6400 -10756.265 0.378 0.394
Chain 1: 6500 -8603.956 0.382 0.394
Chain 1: 6600 -11330.441 0.366 0.338
Chain 1: 6700 -8943.127 0.349 0.267
Chain 1: 6800 -11591.340 0.300 0.250
Chain 1: 6900 -11315.851 0.261 0.241
Chain 1: 7000 -13748.445 0.200 0.228
Chain 1: 7100 -10475.440 0.216 0.241
Chain 1: 7200 -8210.002 0.226 0.250
Chain 1: 7300 -8551.201 0.197 0.241
Chain 1: 7400 -8630.727 0.182 0.241
Chain 1: 7500 -11199.118 0.180 0.229
Chain 1: 7600 -9962.688 0.169 0.228
Chain 1: 7700 -8399.811 0.161 0.186
Chain 1: 7800 -11603.889 0.165 0.186
Chain 1: 7900 -8424.327 0.201 0.229
Chain 1: 8000 -10333.475 0.202 0.229
Chain 1: 8100 -9272.273 0.182 0.186
Chain 1: 8200 -8739.312 0.160 0.185
Chain 1: 8300 -9210.100 0.161 0.185
Chain 1: 8400 -8125.517 0.174 0.185
Chain 1: 8500 -8216.932 0.152 0.133
Chain 1: 8600 -8811.486 0.146 0.133
Chain 1: 8700 -9185.509 0.132 0.114
Chain 1: 8800 -8053.020 0.118 0.114
Chain 1: 8900 -8232.788 0.083 0.067
Chain 1: 9000 -8640.966 0.069 0.061
Chain 1: 9100 -8509.413 0.059 0.051
Chain 1: 9200 -8635.522 0.054 0.047
Chain 1: 9300 -10062.831 0.063 0.047
Chain 1: 9400 -9839.154 0.052 0.041
Chain 1: 9500 -10686.902 0.059 0.047
Chain 1: 9600 -8904.563 0.072 0.047
Chain 1: 9700 -11148.288 0.089 0.079
Chain 1: 9800 -8089.215 0.112 0.079
Chain 1: 9900 -9255.286 0.123 0.126
Chain 1: 10000 -8327.494 0.129 0.126
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001373 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.73 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57014.472 1.000 1.000
Chain 1: 200 -17231.531 1.654 2.309
Chain 1: 300 -8608.603 1.437 1.002
Chain 1: 400 -8207.254 1.090 1.002
Chain 1: 500 -8074.591 0.875 1.000
Chain 1: 600 -8766.513 0.742 1.000
Chain 1: 700 -7863.140 0.653 0.115
Chain 1: 800 -8049.337 0.574 0.115
Chain 1: 900 -7874.114 0.513 0.079
Chain 1: 1000 -7838.229 0.462 0.079
Chain 1: 1100 -7582.109 0.365 0.049
Chain 1: 1200 -7536.879 0.135 0.034
Chain 1: 1300 -7703.090 0.037 0.023
Chain 1: 1400 -7618.528 0.033 0.022
Chain 1: 1500 -7570.830 0.032 0.022
Chain 1: 1600 -7531.828 0.025 0.022
Chain 1: 1700 -7479.032 0.014 0.011
Chain 1: 1800 -7545.607 0.013 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.005709 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 57.09 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85888.248 1.000 1.000
Chain 1: 200 -13264.965 3.237 5.475
Chain 1: 300 -9681.367 2.282 1.000
Chain 1: 400 -10616.797 1.733 1.000
Chain 1: 500 -8618.165 1.433 0.370
Chain 1: 600 -8227.411 1.202 0.370
Chain 1: 700 -8403.661 1.033 0.232
Chain 1: 800 -8659.521 0.908 0.232
Chain 1: 900 -8495.261 0.809 0.088
Chain 1: 1000 -8265.479 0.731 0.088
Chain 1: 1100 -8520.956 0.634 0.047 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8291.066 0.089 0.030
Chain 1: 1300 -8424.686 0.054 0.030
Chain 1: 1400 -8428.524 0.045 0.028
Chain 1: 1500 -8292.516 0.024 0.028
Chain 1: 1600 -8400.666 0.020 0.021
Chain 1: 1700 -8486.155 0.019 0.019
Chain 1: 1800 -8092.324 0.021 0.019
Chain 1: 1900 -8193.071 0.020 0.016
Chain 1: 2000 -8163.744 0.018 0.016
Chain 1: 2100 -8286.235 0.016 0.015
Chain 1: 2200 -8067.501 0.016 0.015
Chain 1: 2300 -8221.890 0.017 0.015
Chain 1: 2400 -8235.762 0.017 0.015
Chain 1: 2500 -8205.176 0.015 0.013
Chain 1: 2600 -8207.799 0.014 0.012
Chain 1: 2700 -8114.046 0.014 0.012
Chain 1: 2800 -8085.213 0.010 0.012 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.007477 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 74.77 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8394711.439 1.000 1.000
Chain 1: 200 -1583805.993 2.650 4.300
Chain 1: 300 -890285.034 2.026 1.000
Chain 1: 400 -457192.025 1.757 1.000
Chain 1: 500 -357473.242 1.461 0.947
Chain 1: 600 -232601.115 1.307 0.947
Chain 1: 700 -118910.285 1.257 0.947
Chain 1: 800 -86149.956 1.147 0.947
Chain 1: 900 -66507.881 1.053 0.779
Chain 1: 1000 -51310.153 0.977 0.779
Chain 1: 1100 -38793.420 0.909 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37968.541 0.481 0.380
Chain 1: 1300 -25934.577 0.450 0.380
Chain 1: 1400 -25652.869 0.356 0.323
Chain 1: 1500 -22243.217 0.344 0.323
Chain 1: 1600 -21460.360 0.294 0.296
Chain 1: 1700 -20335.414 0.204 0.295
Chain 1: 1800 -20279.787 0.166 0.153
Chain 1: 1900 -20605.555 0.138 0.055
Chain 1: 2000 -19118.245 0.116 0.055
Chain 1: 2100 -19356.422 0.085 0.036
Chain 1: 2200 -19582.604 0.084 0.036
Chain 1: 2300 -19200.190 0.040 0.020
Chain 1: 2400 -18972.435 0.040 0.020
Chain 1: 2500 -18774.504 0.025 0.016
Chain 1: 2600 -18405.023 0.024 0.016
Chain 1: 2700 -18362.114 0.019 0.012
Chain 1: 2800 -18079.143 0.020 0.016
Chain 1: 2900 -18360.250 0.020 0.015
Chain 1: 3000 -18346.412 0.012 0.012
Chain 1: 3100 -18431.349 0.011 0.012
Chain 1: 3200 -18122.282 0.012 0.015
Chain 1: 3300 -18326.841 0.011 0.012
Chain 1: 3400 -17802.215 0.013 0.015
Chain 1: 3500 -18413.399 0.015 0.016
Chain 1: 3600 -17721.025 0.017 0.016
Chain 1: 3700 -18107.112 0.019 0.017
Chain 1: 3800 -17068.279 0.023 0.021
Chain 1: 3900 -17064.489 0.022 0.021
Chain 1: 4000 -17181.777 0.022 0.021
Chain 1: 4100 -17095.586 0.022 0.021
Chain 1: 4200 -16912.187 0.022 0.021
Chain 1: 4300 -17050.333 0.021 0.021
Chain 1: 4400 -17007.414 0.019 0.011
Chain 1: 4500 -16910.011 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001303 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.03 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48961.536 1.000 1.000
Chain 1: 200 -15138.424 1.617 2.234
Chain 1: 300 -21035.746 1.172 1.000
Chain 1: 400 -13564.869 1.016 1.000
Chain 1: 500 -14764.397 0.829 0.551
Chain 1: 600 -12466.779 0.722 0.551
Chain 1: 700 -23636.476 0.686 0.473
Chain 1: 800 -13133.188 0.700 0.551
Chain 1: 900 -12076.285 0.632 0.473
Chain 1: 1000 -10526.453 0.584 0.473
Chain 1: 1100 -16768.631 0.521 0.372 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -12177.308 0.335 0.372
Chain 1: 1300 -10196.212 0.327 0.372
Chain 1: 1400 -16955.149 0.311 0.372
Chain 1: 1500 -11832.446 0.347 0.377
Chain 1: 1600 -12072.189 0.330 0.377
Chain 1: 1700 -10405.925 0.299 0.372
Chain 1: 1800 -10721.599 0.222 0.194
Chain 1: 1900 -15532.902 0.244 0.310
Chain 1: 2000 -10158.495 0.282 0.372
Chain 1: 2100 -10007.485 0.247 0.310
Chain 1: 2200 -9668.737 0.212 0.194
Chain 1: 2300 -16307.664 0.234 0.310
Chain 1: 2400 -13889.000 0.211 0.174
Chain 1: 2500 -10695.231 0.198 0.174
Chain 1: 2600 -9279.411 0.211 0.174
Chain 1: 2700 -16963.333 0.240 0.299
Chain 1: 2800 -9355.508 0.319 0.310
Chain 1: 2900 -16811.165 0.332 0.407
Chain 1: 3000 -8945.372 0.367 0.407
Chain 1: 3100 -12028.991 0.391 0.407
Chain 1: 3200 -10734.841 0.400 0.407
Chain 1: 3300 -16958.128 0.396 0.367
Chain 1: 3400 -9315.933 0.460 0.443
Chain 1: 3500 -9523.149 0.433 0.443
Chain 1: 3600 -9901.116 0.421 0.443
Chain 1: 3700 -10098.066 0.378 0.367
Chain 1: 3800 -10100.808 0.297 0.256
Chain 1: 3900 -9577.719 0.258 0.121
Chain 1: 4000 -9345.464 0.172 0.055
Chain 1: 4100 -9120.648 0.149 0.038
Chain 1: 4200 -13146.655 0.168 0.038
Chain 1: 4300 -9639.353 0.167 0.038
Chain 1: 4400 -8970.749 0.093 0.038
Chain 1: 4500 -8579.831 0.095 0.046
Chain 1: 4600 -14077.873 0.130 0.055
Chain 1: 4700 -10656.940 0.161 0.075
Chain 1: 4800 -8987.912 0.179 0.186
Chain 1: 4900 -9196.397 0.176 0.186
Chain 1: 5000 -8563.123 0.181 0.186
Chain 1: 5100 -16486.923 0.226 0.306
Chain 1: 5200 -10196.231 0.258 0.321
Chain 1: 5300 -9448.860 0.229 0.186
Chain 1: 5400 -16652.135 0.265 0.321
Chain 1: 5500 -12594.365 0.293 0.322
Chain 1: 5600 -9047.736 0.293 0.322
Chain 1: 5700 -10663.073 0.276 0.322
Chain 1: 5800 -11561.335 0.265 0.322
Chain 1: 5900 -10908.582 0.269 0.322
Chain 1: 6000 -9226.157 0.279 0.322
Chain 1: 6100 -8762.868 0.237 0.182
Chain 1: 6200 -8459.307 0.179 0.151
Chain 1: 6300 -10889.362 0.193 0.182
Chain 1: 6400 -13005.674 0.166 0.163
Chain 1: 6500 -9197.321 0.175 0.163
Chain 1: 6600 -8863.061 0.140 0.151
Chain 1: 6700 -8535.704 0.128 0.078
Chain 1: 6800 -10335.568 0.138 0.163
Chain 1: 6900 -8866.125 0.149 0.166
Chain 1: 7000 -8928.545 0.131 0.163
Chain 1: 7100 -11766.831 0.150 0.166
Chain 1: 7200 -8779.488 0.180 0.174
Chain 1: 7300 -8530.475 0.161 0.166
Chain 1: 7400 -8642.915 0.146 0.166
Chain 1: 7500 -10728.287 0.124 0.166
Chain 1: 7600 -9127.483 0.138 0.174
Chain 1: 7700 -9012.352 0.135 0.174
Chain 1: 7800 -12778.628 0.147 0.175
Chain 1: 7900 -8271.217 0.185 0.194
Chain 1: 8000 -8600.971 0.188 0.194
Chain 1: 8100 -9022.491 0.169 0.175
Chain 1: 8200 -10709.698 0.151 0.158
Chain 1: 8300 -13239.708 0.167 0.175
Chain 1: 8400 -9133.634 0.211 0.191
Chain 1: 8500 -9398.983 0.194 0.175
Chain 1: 8600 -8345.003 0.189 0.158
Chain 1: 8700 -9641.835 0.201 0.158
Chain 1: 8800 -8731.314 0.182 0.135
Chain 1: 8900 -10086.593 0.141 0.134
Chain 1: 9000 -9241.052 0.146 0.134
Chain 1: 9100 -10620.358 0.155 0.134
Chain 1: 9200 -9277.702 0.153 0.134
Chain 1: 9300 -9217.176 0.135 0.130
Chain 1: 9400 -9498.670 0.093 0.126
Chain 1: 9500 -8344.560 0.104 0.130
Chain 1: 9600 -8490.303 0.093 0.130
Chain 1: 9700 -8865.042 0.084 0.104
Chain 1: 9800 -11227.113 0.094 0.130
Chain 1: 9900 -8902.202 0.107 0.130
Chain 1: 10000 -9308.289 0.102 0.130
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001376 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.76 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -62135.255 1.000 1.000
Chain 1: 200 -18184.182 1.708 2.417
Chain 1: 300 -8979.773 1.481 1.025
Chain 1: 400 -8439.983 1.126 1.025
Chain 1: 500 -8646.905 0.906 1.000
Chain 1: 600 -8324.588 0.761 1.000
Chain 1: 700 -7857.932 0.661 0.064
Chain 1: 800 -8197.938 0.584 0.064
Chain 1: 900 -7804.857 0.524 0.059
Chain 1: 1000 -7894.561 0.473 0.059
Chain 1: 1100 -7709.320 0.376 0.050
Chain 1: 1200 -7542.499 0.136 0.041
Chain 1: 1300 -7801.198 0.037 0.039
Chain 1: 1400 -7820.120 0.031 0.033
Chain 1: 1500 -7580.964 0.031 0.033
Chain 1: 1600 -7634.724 0.028 0.032
Chain 1: 1700 -7558.211 0.023 0.024
Chain 1: 1800 -7573.495 0.019 0.022
Chain 1: 1900 -7580.484 0.014 0.011
Chain 1: 2000 -7645.168 0.014 0.010
Chain 1: 2100 -7571.647 0.013 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00342 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.2 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86623.014 1.000 1.000
Chain 1: 200 -13762.660 3.147 5.294
Chain 1: 300 -10077.133 2.220 1.000
Chain 1: 400 -11265.527 1.691 1.000
Chain 1: 500 -9025.392 1.403 0.366
Chain 1: 600 -8608.651 1.177 0.366
Chain 1: 700 -8829.842 1.012 0.248
Chain 1: 800 -9618.923 0.896 0.248
Chain 1: 900 -8731.098 0.808 0.105
Chain 1: 1000 -8795.966 0.728 0.105
Chain 1: 1100 -8715.322 0.629 0.102 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8471.878 0.102 0.082
Chain 1: 1300 -8756.925 0.069 0.048
Chain 1: 1400 -8718.613 0.059 0.033
Chain 1: 1500 -8607.020 0.035 0.029
Chain 1: 1600 -8715.382 0.032 0.025
Chain 1: 1700 -8790.114 0.030 0.013
Chain 1: 1800 -8359.665 0.027 0.013
Chain 1: 1900 -8463.568 0.018 0.012
Chain 1: 2000 -8438.758 0.018 0.012
Chain 1: 2100 -8573.690 0.018 0.013
Chain 1: 2200 -8367.853 0.018 0.013
Chain 1: 2300 -8463.587 0.016 0.012
Chain 1: 2400 -8527.465 0.016 0.012
Chain 1: 2500 -8472.158 0.015 0.012
Chain 1: 2600 -8476.259 0.014 0.011
Chain 1: 2700 -8391.559 0.014 0.011
Chain 1: 2800 -8348.316 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00489 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 48.9 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8399258.934 1.000 1.000
Chain 1: 200 -1582046.828 2.655 4.309
Chain 1: 300 -891237.591 2.028 1.000
Chain 1: 400 -458167.892 1.757 1.000
Chain 1: 500 -358810.350 1.461 0.945
Chain 1: 600 -233806.156 1.307 0.945
Chain 1: 700 -119815.262 1.256 0.945
Chain 1: 800 -86957.235 1.146 0.945
Chain 1: 900 -67243.255 1.051 0.775
Chain 1: 1000 -51992.335 0.976 0.775
Chain 1: 1100 -39422.213 0.908 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38595.397 0.479 0.378
Chain 1: 1300 -26500.998 0.447 0.378
Chain 1: 1400 -26216.033 0.353 0.319
Chain 1: 1500 -22790.230 0.341 0.319
Chain 1: 1600 -22003.276 0.291 0.293
Chain 1: 1700 -20870.819 0.201 0.293
Chain 1: 1800 -20813.688 0.164 0.150
Chain 1: 1900 -21140.085 0.136 0.054
Chain 1: 2000 -19647.514 0.114 0.054
Chain 1: 2100 -19886.104 0.084 0.036
Chain 1: 2200 -20113.265 0.083 0.036
Chain 1: 2300 -19729.799 0.039 0.019
Chain 1: 2400 -19501.736 0.039 0.019
Chain 1: 2500 -19303.948 0.025 0.015
Chain 1: 2600 -18933.674 0.023 0.015
Chain 1: 2700 -18890.473 0.018 0.012
Chain 1: 2800 -18607.301 0.019 0.015
Chain 1: 2900 -18888.789 0.019 0.015
Chain 1: 3000 -18874.836 0.012 0.012
Chain 1: 3100 -18959.906 0.011 0.012
Chain 1: 3200 -18650.337 0.012 0.015
Chain 1: 3300 -18855.269 0.011 0.012
Chain 1: 3400 -18329.840 0.012 0.015
Chain 1: 3500 -18942.274 0.015 0.015
Chain 1: 3600 -18248.259 0.016 0.015
Chain 1: 3700 -18635.662 0.018 0.017
Chain 1: 3800 -17594.276 0.023 0.021
Chain 1: 3900 -17590.426 0.021 0.021
Chain 1: 4000 -17707.704 0.022 0.021
Chain 1: 4100 -17621.406 0.022 0.021
Chain 1: 4200 -17437.417 0.021 0.021
Chain 1: 4300 -17575.959 0.021 0.021
Chain 1: 4400 -17532.610 0.018 0.011
Chain 1: 4500 -17435.109 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001393 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.93 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12300.399 1.000 1.000
Chain 1: 200 -9163.155 0.671 1.000
Chain 1: 300 -8045.447 0.494 0.342
Chain 1: 400 -8215.824 0.376 0.342
Chain 1: 500 -8076.457 0.304 0.139
Chain 1: 600 -7998.079 0.255 0.139
Chain 1: 700 -7917.445 0.220 0.021
Chain 1: 800 -7948.840 0.193 0.021
Chain 1: 900 -8114.424 0.174 0.020
Chain 1: 1000 -7957.690 0.158 0.020
Chain 1: 1100 -7998.352 0.059 0.020
Chain 1: 1200 -7951.974 0.025 0.017
Chain 1: 1300 -7892.271 0.012 0.010
Chain 1: 1400 -7901.125 0.010 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001376 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.76 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61507.360 1.000 1.000
Chain 1: 200 -17652.562 1.742 2.484
Chain 1: 300 -8793.806 1.497 1.007
Chain 1: 400 -8286.539 1.138 1.007
Chain 1: 500 -8293.530 0.911 1.000
Chain 1: 600 -8345.951 0.760 1.000
Chain 1: 700 -8042.836 0.657 0.061
Chain 1: 800 -7921.098 0.577 0.061
Chain 1: 900 -7860.170 0.513 0.038
Chain 1: 1000 -8068.232 0.465 0.038
Chain 1: 1100 -7661.736 0.370 0.038
Chain 1: 1200 -7759.534 0.123 0.026
Chain 1: 1300 -7592.352 0.024 0.022
Chain 1: 1400 -7774.378 0.020 0.022
Chain 1: 1500 -7563.488 0.023 0.023
Chain 1: 1600 -7583.580 0.023 0.023
Chain 1: 1700 -7499.441 0.020 0.022
Chain 1: 1800 -7521.486 0.019 0.022
Chain 1: 1900 -7536.424 0.018 0.022
Chain 1: 2000 -7574.222 0.016 0.013
Chain 1: 2100 -7554.435 0.011 0.011
Chain 1: 2200 -7671.797 0.012 0.011
Chain 1: 2300 -7578.880 0.011 0.011
Chain 1: 2400 -7606.326 0.009 0.005 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002975 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.75 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86339.323 1.000 1.000
Chain 1: 200 -13389.592 3.224 5.448
Chain 1: 300 -9822.840 2.270 1.000
Chain 1: 400 -10653.889 1.722 1.000
Chain 1: 500 -8742.936 1.422 0.363
Chain 1: 600 -8347.675 1.193 0.363
Chain 1: 700 -8429.719 1.024 0.219
Chain 1: 800 -8779.210 0.901 0.219
Chain 1: 900 -8610.909 0.803 0.078
Chain 1: 1000 -8347.996 0.726 0.078
Chain 1: 1100 -8717.040 0.630 0.047 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8342.456 0.089 0.045
Chain 1: 1300 -8503.963 0.055 0.042
Chain 1: 1400 -8564.940 0.048 0.040
Chain 1: 1500 -8425.612 0.028 0.031
Chain 1: 1600 -8528.867 0.024 0.020
Chain 1: 1700 -8623.062 0.024 0.020
Chain 1: 1800 -8226.652 0.025 0.020
Chain 1: 1900 -8329.458 0.024 0.019
Chain 1: 2000 -8299.545 0.022 0.017
Chain 1: 2100 -8424.335 0.019 0.015
Chain 1: 2200 -8208.545 0.017 0.015
Chain 1: 2300 -8357.895 0.017 0.015
Chain 1: 2400 -8372.843 0.016 0.015
Chain 1: 2500 -8340.558 0.015 0.012
Chain 1: 2600 -8342.763 0.014 0.012
Chain 1: 2700 -8249.322 0.014 0.012
Chain 1: 2800 -8221.546 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003038 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.38 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8439175.657 1.000 1.000
Chain 1: 200 -1588910.820 2.656 4.311
Chain 1: 300 -890785.046 2.032 1.000
Chain 1: 400 -457333.410 1.761 1.000
Chain 1: 500 -357248.561 1.465 0.948
Chain 1: 600 -232150.924 1.310 0.948
Chain 1: 700 -118684.947 1.260 0.948
Chain 1: 800 -86019.726 1.150 0.948
Chain 1: 900 -66429.801 1.055 0.784
Chain 1: 1000 -51283.386 0.979 0.784
Chain 1: 1100 -38823.631 0.911 0.539 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38001.635 0.482 0.380
Chain 1: 1300 -26022.086 0.450 0.380
Chain 1: 1400 -25745.774 0.356 0.321
Chain 1: 1500 -22351.082 0.343 0.321
Chain 1: 1600 -21572.875 0.293 0.295
Chain 1: 1700 -20454.371 0.203 0.295
Chain 1: 1800 -20400.136 0.165 0.152
Chain 1: 1900 -20725.955 0.137 0.055
Chain 1: 2000 -19242.162 0.115 0.055
Chain 1: 2100 -19480.017 0.084 0.036
Chain 1: 2200 -19705.771 0.083 0.036
Chain 1: 2300 -19323.716 0.039 0.020
Chain 1: 2400 -19096.033 0.039 0.020
Chain 1: 2500 -18897.964 0.025 0.016
Chain 1: 2600 -18528.638 0.024 0.016
Chain 1: 2700 -18485.771 0.018 0.012
Chain 1: 2800 -18202.832 0.020 0.016
Chain 1: 2900 -18483.739 0.020 0.015
Chain 1: 3000 -18469.975 0.012 0.012
Chain 1: 3100 -18554.920 0.011 0.012
Chain 1: 3200 -18245.890 0.012 0.015
Chain 1: 3300 -18450.385 0.011 0.012
Chain 1: 3400 -17925.814 0.013 0.015
Chain 1: 3500 -18536.878 0.015 0.016
Chain 1: 3600 -17844.549 0.017 0.016
Chain 1: 3700 -18230.582 0.019 0.017
Chain 1: 3800 -17191.852 0.023 0.021
Chain 1: 3900 -17188.021 0.022 0.021
Chain 1: 4000 -17305.329 0.022 0.021
Chain 1: 4100 -17219.207 0.022 0.021
Chain 1: 4200 -17035.772 0.022 0.021
Chain 1: 4300 -17173.933 0.021 0.021
Chain 1: 4400 -17131.016 0.019 0.011
Chain 1: 4500 -17033.602 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001135 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 11.35 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12719.268 1.000 1.000
Chain 1: 200 -9498.366 0.670 1.000
Chain 1: 300 -8225.049 0.498 0.339
Chain 1: 400 -8385.721 0.378 0.339
Chain 1: 500 -8267.504 0.305 0.155
Chain 1: 600 -8184.258 0.256 0.155
Chain 1: 700 -8065.487 0.222 0.019
Chain 1: 800 -8061.623 0.194 0.019
Chain 1: 900 -8219.664 0.175 0.019
Chain 1: 1000 -8133.942 0.158 0.019
Chain 1: 1100 -8106.180 0.059 0.015
Chain 1: 1200 -8097.706 0.025 0.014
Chain 1: 1300 -8043.281 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001382 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.82 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -62065.856 1.000 1.000
Chain 1: 200 -18279.799 1.698 2.395
Chain 1: 300 -9113.199 1.467 1.006
Chain 1: 400 -8428.244 1.121 1.006
Chain 1: 500 -8827.110 0.906 1.000
Chain 1: 600 -8127.359 0.769 1.000
Chain 1: 700 -8351.745 0.663 0.086
Chain 1: 800 -8218.185 0.582 0.086
Chain 1: 900 -8098.878 0.519 0.081
Chain 1: 1000 -7800.724 0.471 0.081
Chain 1: 1100 -7642.240 0.373 0.045
Chain 1: 1200 -7826.228 0.136 0.038
Chain 1: 1300 -7856.807 0.036 0.027
Chain 1: 1400 -8074.733 0.030 0.027
Chain 1: 1500 -7583.737 0.032 0.027
Chain 1: 1600 -7821.388 0.027 0.027
Chain 1: 1700 -7591.752 0.027 0.027
Chain 1: 1800 -7710.874 0.027 0.027
Chain 1: 1900 -7725.118 0.026 0.027
Chain 1: 2000 -7682.551 0.022 0.024
Chain 1: 2100 -7525.121 0.022 0.024
Chain 1: 2200 -7816.132 0.024 0.027
Chain 1: 2300 -7648.315 0.026 0.027
Chain 1: 2400 -7709.615 0.024 0.022
Chain 1: 2500 -7649.454 0.018 0.021
Chain 1: 2600 -7588.860 0.016 0.015
Chain 1: 2700 -7581.346 0.013 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004961 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 49.61 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86689.232 1.000 1.000
Chain 1: 200 -13870.522 3.125 5.250
Chain 1: 300 -10153.961 2.205 1.000
Chain 1: 400 -11591.601 1.685 1.000
Chain 1: 500 -9021.114 1.405 0.366
Chain 1: 600 -9216.393 1.174 0.366
Chain 1: 700 -8839.095 1.013 0.285
Chain 1: 800 -9506.498 0.895 0.285
Chain 1: 900 -8989.554 0.802 0.124
Chain 1: 1000 -8823.889 0.724 0.124
Chain 1: 1100 -8908.099 0.624 0.070 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8421.325 0.105 0.058
Chain 1: 1300 -9014.125 0.075 0.058
Chain 1: 1400 -8747.079 0.066 0.058
Chain 1: 1500 -8675.752 0.038 0.043
Chain 1: 1600 -8781.763 0.037 0.043
Chain 1: 1700 -8844.280 0.034 0.031
Chain 1: 1800 -8406.944 0.032 0.031
Chain 1: 1900 -8510.894 0.027 0.019
Chain 1: 2000 -8486.542 0.026 0.012
Chain 1: 2100 -8628.874 0.027 0.016
Chain 1: 2200 -8416.986 0.023 0.016
Chain 1: 2300 -8577.142 0.019 0.016
Chain 1: 2400 -8412.803 0.017 0.016
Chain 1: 2500 -8484.316 0.017 0.016
Chain 1: 2600 -8396.353 0.017 0.016
Chain 1: 2700 -8430.560 0.017 0.016
Chain 1: 2800 -8390.335 0.012 0.012
Chain 1: 2900 -8483.974 0.012 0.011
Chain 1: 3000 -8317.894 0.014 0.016
Chain 1: 3100 -8473.049 0.014 0.018
Chain 1: 3200 -8344.858 0.013 0.015
Chain 1: 3300 -8352.781 0.011 0.011
Chain 1: 3400 -8514.180 0.011 0.011
Chain 1: 3500 -8524.590 0.011 0.011
Chain 1: 3600 -8301.280 0.012 0.015
Chain 1: 3700 -8447.750 0.013 0.017
Chain 1: 3800 -8307.701 0.015 0.017
Chain 1: 3900 -8242.095 0.014 0.017
Chain 1: 4000 -8318.528 0.013 0.017
Chain 1: 4100 -8313.429 0.012 0.015
Chain 1: 4200 -8297.440 0.010 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003077 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.77 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8422390.370 1.000 1.000
Chain 1: 200 -1585265.649 2.656 4.313
Chain 1: 300 -891325.891 2.030 1.000
Chain 1: 400 -458192.208 1.759 1.000
Chain 1: 500 -358268.745 1.463 0.945
Chain 1: 600 -233250.644 1.309 0.945
Chain 1: 700 -119535.060 1.258 0.945
Chain 1: 800 -86766.330 1.148 0.945
Chain 1: 900 -67124.709 1.053 0.779
Chain 1: 1000 -51940.776 0.977 0.779
Chain 1: 1100 -39434.464 0.908 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38616.443 0.479 0.378
Chain 1: 1300 -26580.160 0.447 0.378
Chain 1: 1400 -26302.901 0.353 0.317
Chain 1: 1500 -22891.379 0.340 0.317
Chain 1: 1600 -22109.148 0.290 0.293
Chain 1: 1700 -20983.035 0.200 0.292
Chain 1: 1800 -20927.727 0.163 0.149
Chain 1: 1900 -21254.210 0.135 0.054
Chain 1: 2000 -19764.866 0.113 0.054
Chain 1: 2100 -20003.285 0.083 0.035
Chain 1: 2200 -20229.965 0.082 0.035
Chain 1: 2300 -19846.893 0.038 0.019
Chain 1: 2400 -19618.826 0.039 0.019
Chain 1: 2500 -19420.823 0.025 0.015
Chain 1: 2600 -19050.608 0.023 0.015
Chain 1: 2700 -19007.512 0.018 0.012
Chain 1: 2800 -18724.148 0.019 0.015
Chain 1: 2900 -19005.567 0.019 0.015
Chain 1: 3000 -18991.779 0.012 0.012
Chain 1: 3100 -19076.803 0.011 0.012
Chain 1: 3200 -18767.231 0.011 0.015
Chain 1: 3300 -18972.163 0.011 0.012
Chain 1: 3400 -18446.604 0.012 0.015
Chain 1: 3500 -19059.186 0.014 0.015
Chain 1: 3600 -18364.949 0.016 0.015
Chain 1: 3700 -18752.377 0.018 0.016
Chain 1: 3800 -17710.679 0.023 0.021
Chain 1: 3900 -17706.774 0.021 0.021
Chain 1: 4000 -17824.089 0.022 0.021
Chain 1: 4100 -17737.752 0.022 0.021
Chain 1: 4200 -17553.715 0.021 0.021
Chain 1: 4300 -17692.322 0.021 0.021
Chain 1: 4400 -17648.879 0.018 0.010
Chain 1: 4500 -17551.366 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001444 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.44 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12808.543 1.000 1.000
Chain 1: 200 -9791.885 0.654 1.000
Chain 1: 300 -8398.373 0.491 0.308
Chain 1: 400 -8588.297 0.374 0.308
Chain 1: 500 -8585.030 0.299 0.166
Chain 1: 600 -8357.157 0.254 0.166
Chain 1: 700 -8263.749 0.219 0.027
Chain 1: 800 -8287.868 0.192 0.027
Chain 1: 900 -8393.045 0.172 0.022
Chain 1: 1000 -8296.975 0.156 0.022
Chain 1: 1100 -8339.061 0.057 0.013
Chain 1: 1200 -8294.358 0.026 0.012
Chain 1: 1300 -8216.466 0.011 0.011
Chain 1: 1400 -8246.460 0.009 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001477 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.77 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -60687.297 1.000 1.000
Chain 1: 200 -18562.526 1.635 2.269
Chain 1: 300 -9017.048 1.443 1.059
Chain 1: 400 -8307.035 1.103 1.059
Chain 1: 500 -8567.798 0.889 1.000
Chain 1: 600 -8209.978 0.748 1.000
Chain 1: 700 -8054.757 0.644 0.085
Chain 1: 800 -8495.146 0.570 0.085
Chain 1: 900 -8105.197 0.512 0.052
Chain 1: 1000 -7795.393 0.465 0.052
Chain 1: 1100 -7789.026 0.365 0.048
Chain 1: 1200 -7633.818 0.140 0.044
Chain 1: 1300 -7901.787 0.037 0.040
Chain 1: 1400 -7886.441 0.029 0.034
Chain 1: 1500 -7657.338 0.029 0.034
Chain 1: 1600 -7886.593 0.027 0.030
Chain 1: 1700 -7665.850 0.028 0.030
Chain 1: 1800 -7770.563 0.025 0.029
Chain 1: 1900 -7603.202 0.022 0.029
Chain 1: 2000 -7740.090 0.020 0.022
Chain 1: 2100 -7653.272 0.021 0.022
Chain 1: 2200 -7802.476 0.021 0.022
Chain 1: 2300 -7630.287 0.020 0.022
Chain 1: 2400 -7714.130 0.020 0.022
Chain 1: 2500 -7737.926 0.018 0.019
Chain 1: 2600 -7600.423 0.017 0.018
Chain 1: 2700 -7600.063 0.014 0.018
Chain 1: 2800 -7583.950 0.013 0.018
Chain 1: 2900 -7444.386 0.012 0.018
Chain 1: 3000 -7593.699 0.013 0.018
Chain 1: 3100 -7594.106 0.011 0.018
Chain 1: 3200 -7811.319 0.012 0.018
Chain 1: 3300 -7535.654 0.014 0.018
Chain 1: 3400 -7770.084 0.016 0.019
Chain 1: 3500 -7504.826 0.019 0.020
Chain 1: 3600 -7572.384 0.018 0.020
Chain 1: 3700 -7522.178 0.019 0.020
Chain 1: 3800 -7524.006 0.018 0.020
Chain 1: 3900 -7481.115 0.017 0.020
Chain 1: 4000 -7473.481 0.015 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003009 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.09 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87053.757 1.000 1.000
Chain 1: 200 -14011.462 3.107 5.213
Chain 1: 300 -10310.830 2.191 1.000
Chain 1: 400 -11502.515 1.669 1.000
Chain 1: 500 -9303.622 1.382 0.359
Chain 1: 600 -8774.630 1.162 0.359
Chain 1: 700 -8679.045 0.998 0.236
Chain 1: 800 -9845.317 0.888 0.236
Chain 1: 900 -9081.794 0.798 0.118
Chain 1: 1000 -8964.690 0.720 0.118
Chain 1: 1100 -9102.506 0.621 0.104 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8697.680 0.105 0.084
Chain 1: 1300 -8944.627 0.072 0.060
Chain 1: 1400 -8956.207 0.061 0.047
Chain 1: 1500 -8844.919 0.039 0.028
Chain 1: 1600 -8953.181 0.034 0.015
Chain 1: 1700 -9023.849 0.034 0.015
Chain 1: 1800 -8591.605 0.027 0.015
Chain 1: 1900 -8695.910 0.020 0.013
Chain 1: 2000 -8671.296 0.019 0.013
Chain 1: 2100 -8620.262 0.018 0.012
Chain 1: 2200 -8614.814 0.013 0.012
Chain 1: 2300 -8745.169 0.012 0.012
Chain 1: 2400 -8599.204 0.014 0.012
Chain 1: 2500 -8666.452 0.013 0.012
Chain 1: 2600 -8587.287 0.013 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003745 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 37.45 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8435429.779 1.000 1.000
Chain 1: 200 -1589296.579 2.654 4.308
Chain 1: 300 -892559.628 2.029 1.000
Chain 1: 400 -458799.661 1.758 1.000
Chain 1: 500 -358885.516 1.462 0.945
Chain 1: 600 -233539.754 1.308 0.945
Chain 1: 700 -119743.611 1.257 0.945
Chain 1: 800 -86957.766 1.147 0.945
Chain 1: 900 -67291.192 1.052 0.781
Chain 1: 1000 -52093.730 0.976 0.781
Chain 1: 1100 -39578.566 0.908 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38757.928 0.479 0.377
Chain 1: 1300 -26718.109 0.446 0.377
Chain 1: 1400 -26438.590 0.353 0.316
Chain 1: 1500 -23027.352 0.339 0.316
Chain 1: 1600 -22245.108 0.289 0.292
Chain 1: 1700 -21119.001 0.200 0.292
Chain 1: 1800 -21063.442 0.162 0.148
Chain 1: 1900 -21389.892 0.134 0.053
Chain 1: 2000 -19900.803 0.113 0.053
Chain 1: 2100 -20139.119 0.082 0.035
Chain 1: 2200 -20365.804 0.081 0.035
Chain 1: 2300 -19982.759 0.038 0.019
Chain 1: 2400 -19754.748 0.038 0.019
Chain 1: 2500 -19556.798 0.025 0.015
Chain 1: 2600 -19186.627 0.023 0.015
Chain 1: 2700 -19143.512 0.018 0.012
Chain 1: 2800 -18860.218 0.019 0.015
Chain 1: 2900 -19141.656 0.019 0.015
Chain 1: 3000 -19127.724 0.012 0.012
Chain 1: 3100 -19212.795 0.011 0.012
Chain 1: 3200 -18903.252 0.011 0.015
Chain 1: 3300 -19108.186 0.011 0.012
Chain 1: 3400 -18582.714 0.012 0.015
Chain 1: 3500 -19195.115 0.014 0.015
Chain 1: 3600 -18501.111 0.016 0.015
Chain 1: 3700 -18888.442 0.018 0.016
Chain 1: 3800 -17847.051 0.022 0.021
Chain 1: 3900 -17843.174 0.021 0.021
Chain 1: 4000 -17960.483 0.021 0.021
Chain 1: 4100 -17874.167 0.022 0.021
Chain 1: 4200 -17690.189 0.021 0.021
Chain 1: 4300 -17828.731 0.021 0.021
Chain 1: 4400 -17785.372 0.018 0.010
Chain 1: 4500 -17687.863 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001247 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.47 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12111.376 1.000 1.000
Chain 1: 200 -9015.550 0.672 1.000
Chain 1: 300 -7828.349 0.498 0.343
Chain 1: 400 -7995.063 0.379 0.343
Chain 1: 500 -7870.935 0.306 0.152
Chain 1: 600 -7745.760 0.258 0.152
Chain 1: 700 -7754.876 0.221 0.021
Chain 1: 800 -7665.827 0.195 0.021
Chain 1: 900 -7667.181 0.173 0.016
Chain 1: 1000 -7726.193 0.157 0.016
Chain 1: 1100 -7811.319 0.058 0.016
Chain 1: 1200 -7686.512 0.025 0.016
Chain 1: 1300 -7699.639 0.010 0.012
Chain 1: 1400 -7672.525 0.008 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001497 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.97 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -46073.042 1.000 1.000
Chain 1: 200 -15235.984 1.512 2.024
Chain 1: 300 -8490.696 1.273 1.000
Chain 1: 400 -8352.285 0.959 1.000
Chain 1: 500 -8138.214 0.772 0.794
Chain 1: 600 -8797.160 0.656 0.794
Chain 1: 700 -8069.480 0.575 0.090
Chain 1: 800 -7869.372 0.506 0.090
Chain 1: 900 -7601.878 0.454 0.075
Chain 1: 1000 -7651.342 0.409 0.075
Chain 1: 1100 -7552.518 0.311 0.035
Chain 1: 1200 -7504.461 0.109 0.026
Chain 1: 1300 -7471.415 0.030 0.025
Chain 1: 1400 -7783.022 0.032 0.026
Chain 1: 1500 -7506.770 0.033 0.035
Chain 1: 1600 -7393.717 0.027 0.025
Chain 1: 1700 -7405.376 0.018 0.015
Chain 1: 1800 -7451.451 0.017 0.013
Chain 1: 1900 -7475.184 0.013 0.006 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003301 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.01 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86165.013 1.000 1.000
Chain 1: 200 -13174.264 3.270 5.540
Chain 1: 300 -9590.834 2.305 1.000
Chain 1: 400 -10382.243 1.748 1.000
Chain 1: 500 -8544.985 1.441 0.374
Chain 1: 600 -8088.868 1.210 0.374
Chain 1: 700 -8282.651 1.041 0.215
Chain 1: 800 -8572.988 0.915 0.215
Chain 1: 900 -8401.792 0.815 0.076
Chain 1: 1000 -8150.430 0.737 0.076
Chain 1: 1100 -8462.328 0.641 0.056 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8018.681 0.092 0.055
Chain 1: 1300 -8301.136 0.058 0.037
Chain 1: 1400 -8295.336 0.051 0.034
Chain 1: 1500 -8203.603 0.030 0.034
Chain 1: 1600 -8305.812 0.026 0.031
Chain 1: 1700 -8389.467 0.025 0.031
Chain 1: 1800 -7994.969 0.026 0.031
Chain 1: 1900 -8096.213 0.025 0.031
Chain 1: 2000 -8066.948 0.023 0.013
Chain 1: 2100 -8189.031 0.020 0.013
Chain 1: 2200 -7969.475 0.018 0.013
Chain 1: 2300 -8125.071 0.016 0.013
Chain 1: 2400 -8138.685 0.016 0.013
Chain 1: 2500 -8108.471 0.015 0.013
Chain 1: 2600 -8111.171 0.014 0.013
Chain 1: 2700 -8017.385 0.014 0.013
Chain 1: 2800 -7988.292 0.010 0.012 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004603 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 46.03 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8407861.146 1.000 1.000
Chain 1: 200 -1587599.042 2.648 4.296
Chain 1: 300 -891543.100 2.026 1.000
Chain 1: 400 -457764.490 1.756 1.000
Chain 1: 500 -357806.818 1.461 0.948
Chain 1: 600 -232714.652 1.307 0.948
Chain 1: 700 -118895.243 1.257 0.948
Chain 1: 800 -86089.461 1.147 0.948
Chain 1: 900 -66429.796 1.053 0.781
Chain 1: 1000 -51223.831 0.977 0.781
Chain 1: 1100 -38700.878 0.910 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37875.013 0.482 0.381
Chain 1: 1300 -25840.629 0.451 0.381
Chain 1: 1400 -25559.372 0.357 0.324
Chain 1: 1500 -22149.126 0.344 0.324
Chain 1: 1600 -21365.582 0.294 0.297
Chain 1: 1700 -20240.924 0.204 0.296
Chain 1: 1800 -20185.246 0.166 0.154
Chain 1: 1900 -20511.030 0.138 0.056
Chain 1: 2000 -19023.501 0.117 0.056
Chain 1: 2100 -19261.893 0.085 0.037
Chain 1: 2200 -19487.933 0.084 0.037
Chain 1: 2300 -19105.575 0.040 0.020
Chain 1: 2400 -18877.779 0.040 0.020
Chain 1: 2500 -18679.733 0.026 0.016
Chain 1: 2600 -18310.419 0.024 0.016
Chain 1: 2700 -18267.467 0.019 0.012
Chain 1: 2800 -17984.450 0.020 0.016
Chain 1: 2900 -18265.543 0.020 0.015
Chain 1: 3000 -18251.797 0.012 0.012
Chain 1: 3100 -18336.721 0.011 0.012
Chain 1: 3200 -18027.657 0.012 0.015
Chain 1: 3300 -18232.158 0.011 0.012
Chain 1: 3400 -17707.525 0.013 0.015
Chain 1: 3500 -18318.715 0.015 0.016
Chain 1: 3600 -17626.278 0.017 0.016
Chain 1: 3700 -18012.432 0.019 0.017
Chain 1: 3800 -16973.479 0.023 0.021
Chain 1: 3900 -16969.623 0.022 0.021
Chain 1: 4000 -17086.961 0.023 0.021
Chain 1: 4100 -17000.783 0.023 0.021
Chain 1: 4200 -16817.288 0.022 0.021
Chain 1: 4300 -16955.505 0.022 0.021
Chain 1: 4400 -16912.575 0.019 0.011
Chain 1: 4500 -16815.127 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001378 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.78 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -11727.893 1.000 1.000
Chain 1: 200 -8666.699 0.677 1.000
Chain 1: 300 -7770.041 0.490 0.353
Chain 1: 400 -7830.280 0.369 0.353
Chain 1: 500 -7692.609 0.299 0.115
Chain 1: 600 -7629.124 0.250 0.115
Chain 1: 700 -7578.338 0.216 0.018
Chain 1: 800 -7531.778 0.189 0.018
Chain 1: 900 -7497.359 0.169 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001693 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 16.93 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56336.528 1.000 1.000
Chain 1: 200 -16686.348 1.688 2.376
Chain 1: 300 -8415.418 1.453 1.000
Chain 1: 400 -8038.101 1.101 1.000
Chain 1: 500 -8275.319 0.887 0.983
Chain 1: 600 -7950.152 0.746 0.983
Chain 1: 700 -7738.390 0.643 0.047
Chain 1: 800 -7970.715 0.567 0.047
Chain 1: 900 -7814.018 0.506 0.041
Chain 1: 1000 -7690.176 0.457 0.041
Chain 1: 1100 -7673.095 0.357 0.029
Chain 1: 1200 -7541.686 0.121 0.029
Chain 1: 1300 -7580.516 0.023 0.027
Chain 1: 1400 -7817.679 0.022 0.027
Chain 1: 1500 -7574.467 0.022 0.027
Chain 1: 1600 -7498.821 0.019 0.020
Chain 1: 1700 -7461.731 0.017 0.017
Chain 1: 1800 -7503.925 0.014 0.016
Chain 1: 1900 -7463.762 0.013 0.010
Chain 1: 2000 -7553.390 0.013 0.010
Chain 1: 2100 -7479.913 0.013 0.010
Chain 1: 2200 -7604.039 0.013 0.010
Chain 1: 2300 -7730.900 0.014 0.012
Chain 1: 2400 -7559.800 0.014 0.012
Chain 1: 2500 -7454.303 0.012 0.012
Chain 1: 2600 -7473.383 0.011 0.012
Chain 1: 2700 -7516.143 0.011 0.012
Chain 1: 2800 -7462.563 0.011 0.012
Chain 1: 2900 -7396.735 0.012 0.012
Chain 1: 3000 -7519.276 0.012 0.014
Chain 1: 3100 -7481.191 0.012 0.014
Chain 1: 3200 -7622.314 0.012 0.014
Chain 1: 3300 -7442.979 0.013 0.014
Chain 1: 3400 -7557.063 0.012 0.014
Chain 1: 3500 -7424.675 0.012 0.015
Chain 1: 3600 -7460.107 0.012 0.015
Chain 1: 3700 -7424.184 0.012 0.015
Chain 1: 3800 -7448.631 0.012 0.015
Chain 1: 3900 -7432.926 0.011 0.015
Chain 1: 4000 -7405.344 0.010 0.005 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003099 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.99 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85366.519 1.000 1.000
Chain 1: 200 -12781.686 3.339 5.679
Chain 1: 300 -9339.589 2.349 1.000
Chain 1: 400 -9789.504 1.773 1.000
Chain 1: 500 -8241.983 1.456 0.369
Chain 1: 600 -8365.737 1.216 0.369
Chain 1: 700 -8087.528 1.047 0.188
Chain 1: 800 -8319.856 0.920 0.188
Chain 1: 900 -8254.137 0.818 0.046
Chain 1: 1000 -8015.765 0.740 0.046
Chain 1: 1100 -8283.100 0.643 0.034 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8068.093 0.078 0.032
Chain 1: 1300 -8176.647 0.042 0.030
Chain 1: 1400 -8065.649 0.039 0.028
Chain 1: 1500 -8042.577 0.020 0.027
Chain 1: 1600 -8159.741 0.020 0.027
Chain 1: 1700 -8224.667 0.018 0.014
Chain 1: 1800 -7881.587 0.019 0.014
Chain 1: 1900 -7975.588 0.020 0.014
Chain 1: 2000 -7947.376 0.017 0.014
Chain 1: 2100 -8101.400 0.016 0.014
Chain 1: 2200 -7873.623 0.016 0.014
Chain 1: 2300 -7952.381 0.016 0.014
Chain 1: 2400 -8012.421 0.015 0.012
Chain 1: 2500 -7979.447 0.015 0.012
Chain 1: 2600 -7970.914 0.014 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003958 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 39.58 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8398173.222 1.000 1.000
Chain 1: 200 -1584109.578 2.651 4.302
Chain 1: 300 -890359.421 2.027 1.000
Chain 1: 400 -457420.862 1.757 1.000
Chain 1: 500 -357653.548 1.461 0.946
Chain 1: 600 -232503.527 1.307 0.946
Chain 1: 700 -118554.371 1.258 0.946
Chain 1: 800 -85743.318 1.149 0.946
Chain 1: 900 -66051.064 1.054 0.779
Chain 1: 1000 -50815.568 0.979 0.779
Chain 1: 1100 -38280.979 0.911 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37443.452 0.483 0.383
Chain 1: 1300 -25408.852 0.453 0.383
Chain 1: 1400 -25123.074 0.359 0.327
Chain 1: 1500 -21714.300 0.347 0.327
Chain 1: 1600 -20930.414 0.297 0.300
Chain 1: 1700 -19806.474 0.207 0.298
Chain 1: 1800 -19750.454 0.169 0.157
Chain 1: 1900 -20075.371 0.140 0.057
Chain 1: 2000 -18590.204 0.118 0.057
Chain 1: 2100 -18828.169 0.087 0.037
Chain 1: 2200 -19053.704 0.086 0.037
Chain 1: 2300 -18672.037 0.041 0.020
Chain 1: 2400 -18444.583 0.041 0.020
Chain 1: 2500 -18246.704 0.026 0.016
Chain 1: 2600 -17878.001 0.024 0.016
Chain 1: 2700 -17835.314 0.019 0.013
Chain 1: 2800 -17552.754 0.020 0.016
Chain 1: 2900 -17833.439 0.020 0.016
Chain 1: 3000 -17819.668 0.012 0.013
Chain 1: 3100 -17904.486 0.012 0.012
Chain 1: 3200 -17595.944 0.012 0.016
Chain 1: 3300 -17800.067 0.011 0.012
Chain 1: 3400 -17276.431 0.013 0.016
Chain 1: 3500 -17886.149 0.015 0.016
Chain 1: 3600 -17195.639 0.017 0.016
Chain 1: 3700 -17580.344 0.019 0.018
Chain 1: 3800 -16544.451 0.024 0.022
Chain 1: 3900 -16540.732 0.022 0.022
Chain 1: 4000 -16657.994 0.023 0.022
Chain 1: 4100 -16571.978 0.023 0.022
Chain 1: 4200 -16389.201 0.022 0.022
Chain 1: 4300 -16526.891 0.022 0.022
Chain 1: 4400 -16484.486 0.019 0.011
Chain 1: 4500 -16387.191 0.017 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001757 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 17.57 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12619.725 1.000 1.000
Chain 1: 200 -9610.217 0.657 1.000
Chain 1: 300 -8318.406 0.489 0.313
Chain 1: 400 -8356.865 0.368 0.313
Chain 1: 500 -8276.860 0.297 0.155
Chain 1: 600 -8225.187 0.248 0.155
Chain 1: 700 -8121.667 0.215 0.013
Chain 1: 800 -8128.753 0.188 0.013
Chain 1: 900 -8100.484 0.167 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001437 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.37 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57504.644 1.000 1.000
Chain 1: 200 -17799.139 1.615 2.231
Chain 1: 300 -8880.327 1.412 1.004
Chain 1: 400 -8330.462 1.075 1.004
Chain 1: 500 -8630.119 0.867 1.000
Chain 1: 600 -9244.282 0.734 1.000
Chain 1: 700 -8049.973 0.650 0.148
Chain 1: 800 -8342.378 0.573 0.148
Chain 1: 900 -8013.850 0.514 0.066
Chain 1: 1000 -7791.267 0.466 0.066
Chain 1: 1100 -7884.962 0.367 0.066
Chain 1: 1200 -7653.188 0.147 0.041
Chain 1: 1300 -7715.591 0.047 0.035
Chain 1: 1400 -7797.038 0.041 0.035
Chain 1: 1500 -7613.870 0.040 0.030
Chain 1: 1600 -7604.416 0.034 0.029
Chain 1: 1700 -7589.529 0.019 0.024
Chain 1: 1800 -7573.445 0.016 0.012
Chain 1: 1900 -7609.563 0.012 0.010
Chain 1: 2000 -7663.802 0.010 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003671 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 36.71 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86636.348 1.000 1.000
Chain 1: 200 -13791.266 3.141 5.282
Chain 1: 300 -10125.294 2.215 1.000
Chain 1: 400 -11136.570 1.684 1.000
Chain 1: 500 -9108.199 1.392 0.362
Chain 1: 600 -8762.385 1.166 0.362
Chain 1: 700 -8624.883 1.002 0.223
Chain 1: 800 -9260.241 0.885 0.223
Chain 1: 900 -8816.392 0.792 0.091
Chain 1: 1000 -8913.033 0.714 0.091
Chain 1: 1100 -8799.196 0.616 0.069 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8607.720 0.090 0.050
Chain 1: 1300 -8810.501 0.056 0.039
Chain 1: 1400 -8808.142 0.047 0.023
Chain 1: 1500 -8680.477 0.026 0.022
Chain 1: 1600 -8789.632 0.023 0.016
Chain 1: 1700 -8868.862 0.022 0.015
Chain 1: 1800 -8445.981 0.021 0.015
Chain 1: 1900 -8546.789 0.017 0.013
Chain 1: 2000 -8521.289 0.016 0.013
Chain 1: 2100 -8646.717 0.016 0.015
Chain 1: 2200 -8450.128 0.016 0.015
Chain 1: 2300 -8541.618 0.015 0.012
Chain 1: 2400 -8610.455 0.016 0.012
Chain 1: 2500 -8556.703 0.015 0.012
Chain 1: 2600 -8558.014 0.014 0.011
Chain 1: 2700 -8474.754 0.014 0.011
Chain 1: 2800 -8434.715 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.005451 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 54.51 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8391479.163 1.000 1.000
Chain 1: 200 -1582025.420 2.652 4.304
Chain 1: 300 -892454.223 2.026 1.000
Chain 1: 400 -458724.228 1.756 1.000
Chain 1: 500 -359394.280 1.460 0.946
Chain 1: 600 -234254.200 1.306 0.946
Chain 1: 700 -120057.677 1.255 0.946
Chain 1: 800 -87124.337 1.145 0.946
Chain 1: 900 -67372.145 1.051 0.773
Chain 1: 1000 -52087.001 0.975 0.773
Chain 1: 1100 -39489.903 0.907 0.534 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38657.701 0.479 0.378
Chain 1: 1300 -26540.211 0.447 0.378
Chain 1: 1400 -26252.526 0.353 0.319
Chain 1: 1500 -22819.838 0.341 0.319
Chain 1: 1600 -22030.600 0.291 0.293
Chain 1: 1700 -20895.595 0.201 0.293
Chain 1: 1800 -20837.678 0.164 0.150
Chain 1: 1900 -21163.914 0.136 0.054
Chain 1: 2000 -19670.128 0.114 0.054
Chain 1: 2100 -19908.891 0.084 0.036
Chain 1: 2200 -20136.066 0.083 0.036
Chain 1: 2300 -19752.587 0.039 0.019
Chain 1: 2400 -19524.534 0.039 0.019
Chain 1: 2500 -19326.808 0.025 0.015
Chain 1: 2600 -18956.697 0.023 0.015
Chain 1: 2700 -18913.545 0.018 0.012
Chain 1: 2800 -18630.451 0.019 0.015
Chain 1: 2900 -18911.875 0.019 0.015
Chain 1: 3000 -18897.976 0.012 0.012
Chain 1: 3100 -18983.000 0.011 0.012
Chain 1: 3200 -18673.577 0.011 0.015
Chain 1: 3300 -18878.386 0.011 0.012
Chain 1: 3400 -18353.189 0.012 0.015
Chain 1: 3500 -18965.351 0.015 0.015
Chain 1: 3600 -18271.695 0.016 0.015
Chain 1: 3700 -18658.819 0.018 0.017
Chain 1: 3800 -17618.057 0.023 0.021
Chain 1: 3900 -17614.221 0.021 0.021
Chain 1: 4000 -17731.485 0.022 0.021
Chain 1: 4100 -17645.228 0.022 0.021
Chain 1: 4200 -17461.359 0.021 0.021
Chain 1: 4300 -17599.815 0.021 0.021
Chain 1: 4400 -17556.586 0.018 0.011
Chain 1: 4500 -17459.094 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00179 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 17.9 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48395.223 1.000 1.000
Chain 1: 200 -39867.815 0.607 1.000
Chain 1: 300 -20942.904 0.706 0.904
Chain 1: 400 -12215.901 0.708 0.904
Chain 1: 500 -11328.112 0.582 0.714
Chain 1: 600 -12761.772 0.504 0.714
Chain 1: 700 -11662.939 0.445 0.214
Chain 1: 800 -14508.066 0.414 0.214
Chain 1: 900 -13837.758 0.373 0.196
Chain 1: 1000 -13580.738 0.338 0.196
Chain 1: 1100 -10291.264 0.270 0.196
Chain 1: 1200 -10411.887 0.250 0.112
Chain 1: 1300 -19947.796 0.207 0.112
Chain 1: 1400 -11496.335 0.209 0.112
Chain 1: 1500 -12377.501 0.209 0.112
Chain 1: 1600 -13842.203 0.208 0.106
Chain 1: 1700 -9244.725 0.248 0.196
Chain 1: 1800 -9897.789 0.235 0.106
Chain 1: 1900 -13571.290 0.257 0.271
Chain 1: 2000 -16584.374 0.274 0.271
Chain 1: 2100 -9181.041 0.322 0.271
Chain 1: 2200 -9485.329 0.324 0.271
Chain 1: 2300 -8888.408 0.283 0.182
Chain 1: 2400 -8764.165 0.211 0.106
Chain 1: 2500 -8860.346 0.205 0.106
Chain 1: 2600 -8983.156 0.196 0.067
Chain 1: 2700 -16081.062 0.190 0.067
Chain 1: 2800 -10119.594 0.243 0.182
Chain 1: 2900 -13829.642 0.242 0.182
Chain 1: 3000 -10171.386 0.260 0.268
Chain 1: 3100 -9445.566 0.187 0.077
Chain 1: 3200 -8379.662 0.197 0.127
Chain 1: 3300 -10720.216 0.212 0.218
Chain 1: 3400 -12370.485 0.224 0.218
Chain 1: 3500 -10628.328 0.239 0.218
Chain 1: 3600 -10990.385 0.241 0.218
Chain 1: 3700 -16110.592 0.229 0.218
Chain 1: 3800 -8702.919 0.255 0.218
Chain 1: 3900 -15321.623 0.271 0.218
Chain 1: 4000 -9144.714 0.303 0.218
Chain 1: 4100 -11734.992 0.317 0.221
Chain 1: 4200 -12717.889 0.312 0.221
Chain 1: 4300 -9154.386 0.329 0.318
Chain 1: 4400 -9174.376 0.316 0.318
Chain 1: 4500 -9330.071 0.302 0.318
Chain 1: 4600 -13026.599 0.327 0.318
Chain 1: 4700 -8464.584 0.349 0.389
Chain 1: 4800 -8601.058 0.265 0.284
Chain 1: 4900 -8562.091 0.222 0.221
Chain 1: 5000 -8987.054 0.160 0.077
Chain 1: 5100 -8693.431 0.141 0.047
Chain 1: 5200 -8569.835 0.135 0.034
Chain 1: 5300 -8988.346 0.100 0.034
Chain 1: 5400 -10482.503 0.114 0.047
Chain 1: 5500 -8672.647 0.134 0.047
Chain 1: 5600 -8434.819 0.108 0.047
Chain 1: 5700 -8329.833 0.055 0.034
Chain 1: 5800 -8261.210 0.055 0.034
Chain 1: 5900 -8694.238 0.059 0.047
Chain 1: 6000 -11520.317 0.079 0.047
Chain 1: 6100 -9379.069 0.098 0.050
Chain 1: 6200 -8080.194 0.113 0.143
Chain 1: 6300 -8357.621 0.112 0.143
Chain 1: 6400 -11039.545 0.122 0.161
Chain 1: 6500 -8779.724 0.127 0.161
Chain 1: 6600 -8312.639 0.129 0.161
Chain 1: 6700 -8864.451 0.134 0.161
Chain 1: 6800 -8074.431 0.143 0.161
Chain 1: 6900 -8172.172 0.140 0.161
Chain 1: 7000 -9452.749 0.129 0.135
Chain 1: 7100 -11208.350 0.121 0.135
Chain 1: 7200 -8950.125 0.131 0.135
Chain 1: 7300 -8708.786 0.130 0.135
Chain 1: 7400 -9939.565 0.118 0.124
Chain 1: 7500 -9140.286 0.101 0.098
Chain 1: 7600 -8444.423 0.104 0.098
Chain 1: 7700 -8252.363 0.100 0.098
Chain 1: 7800 -8136.864 0.092 0.087
Chain 1: 7900 -8084.019 0.091 0.087
Chain 1: 8000 -9486.298 0.092 0.087
Chain 1: 8100 -8155.642 0.093 0.087
Chain 1: 8200 -8348.489 0.070 0.082
Chain 1: 8300 -11339.932 0.094 0.087
Chain 1: 8400 -8266.375 0.118 0.087
Chain 1: 8500 -7977.683 0.113 0.082
Chain 1: 8600 -8496.180 0.111 0.061
Chain 1: 8700 -8970.012 0.114 0.061
Chain 1: 8800 -10121.628 0.124 0.114
Chain 1: 8900 -9991.777 0.125 0.114
Chain 1: 9000 -11028.052 0.119 0.094
Chain 1: 9100 -8652.137 0.130 0.094
Chain 1: 9200 -8193.812 0.134 0.094
Chain 1: 9300 -8995.266 0.116 0.089
Chain 1: 9400 -8116.655 0.090 0.089
Chain 1: 9500 -7900.481 0.089 0.089
Chain 1: 9600 -7937.525 0.083 0.089
Chain 1: 9700 -7867.934 0.079 0.089
Chain 1: 9800 -8279.102 0.073 0.056
Chain 1: 9900 -9199.611 0.081 0.089
Chain 1: 10000 -8027.027 0.086 0.089
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00139 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.9 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57748.165 1.000 1.000
Chain 1: 200 -17282.602 1.671 2.341
Chain 1: 300 -8484.775 1.459 1.037
Chain 1: 400 -8068.009 1.107 1.037
Chain 1: 500 -8454.656 0.895 1.000
Chain 1: 600 -8563.928 0.748 1.000
Chain 1: 700 -7824.458 0.655 0.095
Chain 1: 800 -8029.943 0.576 0.095
Chain 1: 900 -7859.319 0.514 0.052
Chain 1: 1000 -7654.947 0.466 0.052
Chain 1: 1100 -7609.057 0.366 0.046
Chain 1: 1200 -7722.855 0.134 0.027
Chain 1: 1300 -7643.860 0.031 0.026
Chain 1: 1400 -7795.389 0.028 0.022
Chain 1: 1500 -7579.399 0.026 0.022
Chain 1: 1600 -7474.523 0.026 0.022
Chain 1: 1700 -7461.525 0.017 0.019
Chain 1: 1800 -7501.716 0.015 0.015
Chain 1: 1900 -7536.097 0.013 0.014
Chain 1: 2000 -7529.543 0.011 0.010
Chain 1: 2100 -7543.320 0.010 0.010
Chain 1: 2200 -7622.554 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003127 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.27 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86025.826 1.000 1.000
Chain 1: 200 -13118.681 3.279 5.558
Chain 1: 300 -9596.579 2.308 1.000
Chain 1: 400 -10295.893 1.748 1.000
Chain 1: 500 -8535.253 1.440 0.367
Chain 1: 600 -8120.727 1.208 0.367
Chain 1: 700 -8308.471 1.039 0.206
Chain 1: 800 -8452.160 0.911 0.206
Chain 1: 900 -8473.522 0.810 0.068
Chain 1: 1000 -8253.187 0.732 0.068
Chain 1: 1100 -8496.812 0.635 0.051 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8270.187 0.082 0.029
Chain 1: 1300 -8360.279 0.046 0.027
Chain 1: 1400 -8344.640 0.039 0.027
Chain 1: 1500 -8247.993 0.020 0.023
Chain 1: 1600 -8339.396 0.016 0.017
Chain 1: 1700 -8443.360 0.015 0.012
Chain 1: 1800 -8055.779 0.018 0.012
Chain 1: 1900 -8155.458 0.019 0.012
Chain 1: 2000 -8125.663 0.017 0.012
Chain 1: 2100 -8267.621 0.016 0.012
Chain 1: 2200 -8047.297 0.016 0.012
Chain 1: 2300 -8189.599 0.016 0.012
Chain 1: 2400 -8074.865 0.018 0.014
Chain 1: 2500 -8132.505 0.017 0.014
Chain 1: 2600 -8146.236 0.016 0.014
Chain 1: 2700 -8067.798 0.016 0.014
Chain 1: 2800 -8050.882 0.011 0.012
Chain 1: 2900 -8058.790 0.010 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003408 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.08 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8397996.523 1.000 1.000
Chain 1: 200 -1582019.638 2.654 4.308
Chain 1: 300 -889926.926 2.029 1.000
Chain 1: 400 -457302.147 1.758 1.000
Chain 1: 500 -357893.740 1.462 0.946
Chain 1: 600 -232820.658 1.308 0.946
Chain 1: 700 -118934.687 1.258 0.946
Chain 1: 800 -86109.932 1.148 0.946
Chain 1: 900 -66424.893 1.054 0.778
Chain 1: 1000 -51188.891 0.978 0.778
Chain 1: 1100 -38646.931 0.910 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37812.107 0.482 0.381
Chain 1: 1300 -25764.748 0.451 0.381
Chain 1: 1400 -25478.792 0.357 0.325
Chain 1: 1500 -22066.491 0.345 0.325
Chain 1: 1600 -21281.843 0.295 0.298
Chain 1: 1700 -20156.178 0.205 0.296
Chain 1: 1800 -20099.855 0.167 0.155
Chain 1: 1900 -20425.291 0.139 0.056
Chain 1: 2000 -18938.063 0.117 0.056
Chain 1: 2100 -19176.201 0.086 0.037
Chain 1: 2200 -19402.271 0.085 0.037
Chain 1: 2300 -19019.996 0.040 0.020
Chain 1: 2400 -18792.373 0.040 0.020
Chain 1: 2500 -18594.444 0.026 0.016
Chain 1: 2600 -18225.380 0.024 0.016
Chain 1: 2700 -18182.473 0.019 0.012
Chain 1: 2800 -17899.774 0.020 0.016
Chain 1: 2900 -18180.642 0.020 0.015
Chain 1: 3000 -18166.807 0.012 0.012
Chain 1: 3100 -18251.744 0.011 0.012
Chain 1: 3200 -17942.904 0.012 0.015
Chain 1: 3300 -18147.207 0.011 0.012
Chain 1: 3400 -17623.062 0.013 0.015
Chain 1: 3500 -18233.622 0.015 0.016
Chain 1: 3600 -17541.961 0.017 0.016
Chain 1: 3700 -17927.608 0.019 0.017
Chain 1: 3800 -16889.944 0.023 0.022
Chain 1: 3900 -16886.162 0.022 0.022
Chain 1: 4000 -17003.433 0.023 0.022
Chain 1: 4100 -16917.409 0.023 0.022
Chain 1: 4200 -16734.155 0.022 0.022
Chain 1: 4300 -16872.174 0.022 0.022
Chain 1: 4400 -16829.478 0.019 0.011
Chain 1: 4500 -16732.100 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00138 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.8 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12621.568 1.000 1.000
Chain 1: 200 -9459.141 0.667 1.000
Chain 1: 300 -8103.618 0.501 0.334
Chain 1: 400 -8339.113 0.382 0.334
Chain 1: 500 -8110.644 0.312 0.167
Chain 1: 600 -8062.792 0.261 0.167
Chain 1: 700 -8131.475 0.225 0.028
Chain 1: 800 -8065.004 0.198 0.028
Chain 1: 900 -7886.018 0.178 0.028
Chain 1: 1000 -8019.503 0.162 0.028
Chain 1: 1100 -8053.460 0.062 0.023
Chain 1: 1200 -8001.698 0.030 0.017
Chain 1: 1300 -7948.951 0.014 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001416 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.16 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -46516.789 1.000 1.000
Chain 1: 200 -15784.000 1.474 1.947
Chain 1: 300 -8825.318 1.245 1.000
Chain 1: 400 -8308.924 0.949 1.000
Chain 1: 500 -8281.845 0.760 0.788
Chain 1: 600 -8594.671 0.640 0.788
Chain 1: 700 -7647.502 0.566 0.124
Chain 1: 800 -8296.953 0.505 0.124
Chain 1: 900 -8020.994 0.453 0.078
Chain 1: 1000 -7835.502 0.410 0.078
Chain 1: 1100 -7759.543 0.311 0.062
Chain 1: 1200 -7830.194 0.117 0.036
Chain 1: 1300 -7732.136 0.039 0.034
Chain 1: 1400 -7658.236 0.034 0.024
Chain 1: 1500 -7605.822 0.034 0.024
Chain 1: 1600 -7777.644 0.033 0.022
Chain 1: 1700 -7550.556 0.024 0.022
Chain 1: 1800 -7659.267 0.017 0.014
Chain 1: 1900 -7667.175 0.014 0.013
Chain 1: 2000 -7686.030 0.012 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003034 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.34 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86685.105 1.000 1.000
Chain 1: 200 -13749.110 3.152 5.305
Chain 1: 300 -10044.674 2.225 1.000
Chain 1: 400 -11214.858 1.694 1.000
Chain 1: 500 -9040.337 1.404 0.369
Chain 1: 600 -8461.521 1.181 0.369
Chain 1: 700 -8644.111 1.015 0.241
Chain 1: 800 -8846.640 0.891 0.241
Chain 1: 900 -8836.555 0.792 0.104
Chain 1: 1000 -8698.338 0.715 0.104
Chain 1: 1100 -8783.738 0.616 0.068 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8416.102 0.090 0.044
Chain 1: 1300 -8712.591 0.056 0.034
Chain 1: 1400 -8709.171 0.046 0.023
Chain 1: 1500 -8568.691 0.023 0.021
Chain 1: 1600 -8683.943 0.018 0.016
Chain 1: 1700 -8749.348 0.016 0.016
Chain 1: 1800 -8315.821 0.019 0.016
Chain 1: 1900 -8419.867 0.021 0.016
Chain 1: 2000 -8395.288 0.019 0.013
Chain 1: 2100 -8532.363 0.020 0.016
Chain 1: 2200 -8326.015 0.018 0.016
Chain 1: 2300 -8472.543 0.016 0.016
Chain 1: 2400 -8324.188 0.018 0.016
Chain 1: 2500 -8393.803 0.017 0.016
Chain 1: 2600 -8306.973 0.017 0.016
Chain 1: 2700 -8340.042 0.017 0.016
Chain 1: 2800 -8301.150 0.012 0.012
Chain 1: 2900 -8393.272 0.012 0.011
Chain 1: 3000 -8219.953 0.014 0.016
Chain 1: 3100 -8383.253 0.014 0.017
Chain 1: 3200 -8256.057 0.013 0.015
Chain 1: 3300 -8265.022 0.011 0.011
Chain 1: 3400 -8417.057 0.011 0.011
Chain 1: 3500 -8409.810 0.011 0.011
Chain 1: 3600 -8213.245 0.012 0.015
Chain 1: 3700 -8356.583 0.013 0.017
Chain 1: 3800 -8220.077 0.014 0.017
Chain 1: 3900 -8155.250 0.014 0.017
Chain 1: 4000 -8229.488 0.013 0.017
Chain 1: 4100 -8220.642 0.011 0.015
Chain 1: 4200 -8209.229 0.010 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003296 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.96 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8390533.192 1.000 1.000
Chain 1: 200 -1588722.076 2.641 4.281
Chain 1: 300 -892958.029 2.020 1.000
Chain 1: 400 -458510.116 1.752 1.000
Chain 1: 500 -358630.427 1.457 0.948
Chain 1: 600 -233312.115 1.304 0.948
Chain 1: 700 -119518.442 1.254 0.948
Chain 1: 800 -86668.899 1.144 0.948
Chain 1: 900 -67026.583 1.050 0.779
Chain 1: 1000 -51839.865 0.974 0.779
Chain 1: 1100 -39317.547 0.906 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38501.592 0.480 0.379
Chain 1: 1300 -26460.470 0.448 0.379
Chain 1: 1400 -26181.404 0.354 0.318
Chain 1: 1500 -22767.640 0.341 0.318
Chain 1: 1600 -21983.809 0.291 0.293
Chain 1: 1700 -20858.241 0.201 0.293
Chain 1: 1800 -20802.752 0.163 0.150
Chain 1: 1900 -21129.200 0.136 0.054
Chain 1: 2000 -19639.595 0.114 0.054
Chain 1: 2100 -19878.319 0.083 0.036
Chain 1: 2200 -20104.717 0.082 0.036
Chain 1: 2300 -19721.863 0.039 0.019
Chain 1: 2400 -19493.835 0.039 0.019
Chain 1: 2500 -19295.583 0.025 0.015
Chain 1: 2600 -18925.727 0.023 0.015
Chain 1: 2700 -18882.688 0.018 0.012
Chain 1: 2800 -18599.196 0.019 0.015
Chain 1: 2900 -18880.691 0.019 0.015
Chain 1: 3000 -18866.937 0.012 0.012
Chain 1: 3100 -18951.912 0.011 0.012
Chain 1: 3200 -18642.455 0.012 0.015
Chain 1: 3300 -18847.313 0.011 0.012
Chain 1: 3400 -18321.786 0.012 0.015
Chain 1: 3500 -18934.233 0.015 0.015
Chain 1: 3600 -18240.282 0.016 0.015
Chain 1: 3700 -18627.494 0.018 0.017
Chain 1: 3800 -17586.071 0.023 0.021
Chain 1: 3900 -17582.158 0.021 0.021
Chain 1: 4000 -17699.520 0.022 0.021
Chain 1: 4100 -17613.130 0.022 0.021
Chain 1: 4200 -17429.184 0.021 0.021
Chain 1: 4300 -17567.762 0.021 0.021
Chain 1: 4400 -17524.409 0.018 0.011
Chain 1: 4500 -17426.875 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00127 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.7 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12523.700 1.000 1.000
Chain 1: 200 -9386.459 0.667 1.000
Chain 1: 300 -7920.942 0.506 0.334
Chain 1: 400 -8073.253 0.385 0.334
Chain 1: 500 -7939.536 0.311 0.185
Chain 1: 600 -7906.076 0.260 0.185
Chain 1: 700 -7800.180 0.225 0.019
Chain 1: 800 -7808.993 0.197 0.019
Chain 1: 900 -7743.986 0.176 0.017
Chain 1: 1000 -7922.279 0.160 0.019
Chain 1: 1100 -7939.817 0.061 0.017
Chain 1: 1200 -7821.643 0.029 0.015
Chain 1: 1300 -7774.026 0.011 0.014
Chain 1: 1400 -7796.665 0.009 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004467 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 44.67 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56914.692 1.000 1.000
Chain 1: 200 -17432.478 1.632 2.265
Chain 1: 300 -8650.647 1.427 1.015
Chain 1: 400 -8286.497 1.081 1.015
Chain 1: 500 -8713.799 0.875 1.000
Chain 1: 600 -8876.144 0.732 1.000
Chain 1: 700 -7793.253 0.647 0.139
Chain 1: 800 -8204.191 0.573 0.139
Chain 1: 900 -7502.450 0.519 0.094
Chain 1: 1000 -7589.666 0.469 0.094
Chain 1: 1100 -7712.999 0.370 0.050
Chain 1: 1200 -7517.873 0.146 0.049
Chain 1: 1300 -7493.650 0.045 0.044
Chain 1: 1400 -7505.032 0.041 0.026
Chain 1: 1500 -7495.510 0.036 0.018
Chain 1: 1600 -7673.458 0.037 0.023
Chain 1: 1700 -7337.961 0.027 0.023
Chain 1: 1800 -7535.833 0.025 0.023
Chain 1: 1900 -7533.795 0.015 0.016
Chain 1: 2000 -7519.267 0.015 0.016
Chain 1: 2100 -7477.648 0.013 0.006 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002976 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.76 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85390.927 1.000 1.000
Chain 1: 200 -13450.607 3.174 5.348
Chain 1: 300 -9804.735 2.240 1.000
Chain 1: 400 -10624.918 1.699 1.000
Chain 1: 500 -8771.423 1.402 0.372
Chain 1: 600 -8316.085 1.177 0.372
Chain 1: 700 -8300.742 1.009 0.211
Chain 1: 800 -8521.685 0.886 0.211
Chain 1: 900 -8561.947 0.788 0.077
Chain 1: 1000 -8483.298 0.711 0.077
Chain 1: 1100 -8678.116 0.613 0.055 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8267.786 0.083 0.050
Chain 1: 1300 -8499.421 0.048 0.027
Chain 1: 1400 -8510.496 0.041 0.026
Chain 1: 1500 -8358.369 0.022 0.022
Chain 1: 1600 -8472.470 0.017 0.018
Chain 1: 1700 -8550.803 0.018 0.018
Chain 1: 1800 -8129.520 0.021 0.018
Chain 1: 1900 -8229.642 0.021 0.018
Chain 1: 2000 -8203.804 0.021 0.018
Chain 1: 2100 -8328.670 0.020 0.015
Chain 1: 2200 -8135.564 0.018 0.015
Chain 1: 2300 -8224.288 0.016 0.013
Chain 1: 2400 -8293.420 0.017 0.013
Chain 1: 2500 -8239.533 0.015 0.012
Chain 1: 2600 -8240.412 0.014 0.011
Chain 1: 2700 -8157.381 0.014 0.011
Chain 1: 2800 -8117.978 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003292 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.92 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8352459.830 1.000 1.000
Chain 1: 200 -1576512.972 2.649 4.298
Chain 1: 300 -890584.403 2.023 1.000
Chain 1: 400 -457826.339 1.753 1.000
Chain 1: 500 -358867.174 1.458 0.945
Chain 1: 600 -233912.268 1.304 0.945
Chain 1: 700 -119729.830 1.254 0.945
Chain 1: 800 -86815.609 1.145 0.945
Chain 1: 900 -67062.507 1.050 0.770
Chain 1: 1000 -51778.301 0.975 0.770
Chain 1: 1100 -39175.881 0.907 0.534 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38344.743 0.479 0.379
Chain 1: 1300 -26217.150 0.448 0.379
Chain 1: 1400 -25929.220 0.355 0.322
Chain 1: 1500 -22493.949 0.343 0.322
Chain 1: 1600 -21703.981 0.293 0.295
Chain 1: 1700 -20567.435 0.203 0.295
Chain 1: 1800 -20509.334 0.165 0.153
Chain 1: 1900 -20835.563 0.138 0.055
Chain 1: 2000 -19341.080 0.116 0.055
Chain 1: 2100 -19579.819 0.085 0.036
Chain 1: 2200 -19807.190 0.084 0.036
Chain 1: 2300 -19423.553 0.039 0.020
Chain 1: 2400 -19195.471 0.040 0.020
Chain 1: 2500 -18997.868 0.025 0.016
Chain 1: 2600 -18627.649 0.024 0.016
Chain 1: 2700 -18584.491 0.018 0.012
Chain 1: 2800 -18301.437 0.020 0.015
Chain 1: 2900 -18582.882 0.020 0.015
Chain 1: 3000 -18568.974 0.012 0.012
Chain 1: 3100 -18653.982 0.011 0.012
Chain 1: 3200 -18344.565 0.012 0.015
Chain 1: 3300 -18549.376 0.011 0.012
Chain 1: 3400 -18024.209 0.013 0.015
Chain 1: 3500 -18636.393 0.015 0.015
Chain 1: 3600 -17942.727 0.017 0.015
Chain 1: 3700 -18329.844 0.019 0.017
Chain 1: 3800 -17289.110 0.023 0.021
Chain 1: 3900 -17285.300 0.022 0.021
Chain 1: 4000 -17402.545 0.022 0.021
Chain 1: 4100 -17316.295 0.022 0.021
Chain 1: 4200 -17132.444 0.022 0.021
Chain 1: 4300 -17270.879 0.021 0.021
Chain 1: 4400 -17227.628 0.019 0.011
Chain 1: 4500 -17130.177 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001273 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.73 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12956.770 1.000 1.000
Chain 1: 200 -9829.256 0.659 1.000
Chain 1: 300 -8237.313 0.504 0.318
Chain 1: 400 -8487.126 0.385 0.318
Chain 1: 500 -8190.222 0.315 0.193
Chain 1: 600 -8190.290 0.263 0.193
Chain 1: 700 -8044.740 0.228 0.036
Chain 1: 800 -8018.501 0.200 0.036
Chain 1: 900 -7989.892 0.178 0.029
Chain 1: 1000 -8184.367 0.163 0.029
Chain 1: 1100 -8496.924 0.066 0.029
Chain 1: 1200 -8068.148 0.040 0.029
Chain 1: 1300 -8027.574 0.021 0.024
Chain 1: 1400 -8040.133 0.018 0.018
Chain 1: 1500 -8149.204 0.016 0.013
Chain 1: 1600 -8036.656 0.017 0.014
Chain 1: 1700 -8014.059 0.016 0.013
Chain 1: 1800 -7986.333 0.016 0.013
Chain 1: 1900 -8015.053 0.016 0.013
Chain 1: 2000 -7947.026 0.014 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001618 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 16.18 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -59075.076 1.000 1.000
Chain 1: 200 -18320.867 1.612 2.224
Chain 1: 300 -8956.577 1.423 1.046
Chain 1: 400 -8068.179 1.095 1.046
Chain 1: 500 -8106.685 0.877 1.000
Chain 1: 600 -8156.500 0.732 1.000
Chain 1: 700 -8786.180 0.638 0.110
Chain 1: 800 -8412.584 0.563 0.110
Chain 1: 900 -7832.018 0.509 0.074
Chain 1: 1000 -7775.670 0.459 0.074
Chain 1: 1100 -7701.659 0.360 0.072
Chain 1: 1200 -7569.629 0.139 0.044
Chain 1: 1300 -7550.849 0.035 0.017
Chain 1: 1400 -7958.829 0.029 0.017
Chain 1: 1500 -7567.473 0.034 0.044
Chain 1: 1600 -7779.007 0.036 0.044
Chain 1: 1700 -7604.612 0.031 0.027
Chain 1: 1800 -7699.905 0.028 0.023
Chain 1: 1900 -7595.100 0.022 0.017
Chain 1: 2000 -7693.251 0.022 0.017
Chain 1: 2100 -7540.016 0.023 0.020
Chain 1: 2200 -7824.099 0.025 0.023
Chain 1: 2300 -7615.086 0.028 0.027
Chain 1: 2400 -7516.444 0.024 0.023
Chain 1: 2500 -7575.423 0.019 0.020
Chain 1: 2600 -7522.455 0.017 0.014
Chain 1: 2700 -7502.685 0.015 0.013
Chain 1: 2800 -7531.233 0.015 0.013
Chain 1: 2900 -7381.263 0.015 0.013
Chain 1: 3000 -7539.569 0.016 0.020
Chain 1: 3100 -7524.888 0.014 0.013
Chain 1: 3200 -7745.774 0.013 0.013
Chain 1: 3300 -7437.158 0.015 0.013
Chain 1: 3400 -7695.279 0.017 0.020
Chain 1: 3500 -7442.415 0.019 0.021
Chain 1: 3600 -7495.892 0.019 0.021
Chain 1: 3700 -7453.087 0.020 0.021
Chain 1: 3800 -7420.432 0.020 0.021
Chain 1: 3900 -7401.588 0.018 0.021
Chain 1: 4000 -7397.555 0.016 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003952 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 39.52 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87259.631 1.000 1.000
Chain 1: 200 -14050.897 3.105 5.210
Chain 1: 300 -10250.028 2.194 1.000
Chain 1: 400 -12066.655 1.683 1.000
Chain 1: 500 -8673.412 1.425 0.391
Chain 1: 600 -8656.579 1.187 0.391
Chain 1: 700 -9187.040 1.026 0.371
Chain 1: 800 -8873.986 0.902 0.371
Chain 1: 900 -8984.375 0.803 0.151
Chain 1: 1000 -8565.400 0.728 0.151
Chain 1: 1100 -8721.014 0.630 0.058 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8599.646 0.110 0.049
Chain 1: 1300 -8817.831 0.075 0.035
Chain 1: 1400 -8753.620 0.061 0.025
Chain 1: 1500 -8730.940 0.022 0.018
Chain 1: 1600 -8764.424 0.022 0.018
Chain 1: 1700 -8847.908 0.018 0.014
Chain 1: 1800 -8407.086 0.019 0.014
Chain 1: 1900 -8504.713 0.019 0.014
Chain 1: 2000 -8524.342 0.015 0.011
Chain 1: 2100 -8621.086 0.014 0.011
Chain 1: 2200 -8394.572 0.015 0.011
Chain 1: 2300 -8613.857 0.015 0.011
Chain 1: 2400 -8399.859 0.017 0.011
Chain 1: 2500 -8476.574 0.018 0.011
Chain 1: 2600 -8386.071 0.018 0.011
Chain 1: 2700 -8419.675 0.018 0.011
Chain 1: 2800 -8370.431 0.013 0.011
Chain 1: 2900 -8485.487 0.013 0.011
Chain 1: 3000 -8397.431 0.014 0.011
Chain 1: 3100 -8362.691 0.014 0.011
Chain 1: 3200 -8334.324 0.011 0.010
Chain 1: 3300 -8595.018 0.012 0.010
Chain 1: 3400 -8637.962 0.010 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003372 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.72 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8437955.408 1.000 1.000
Chain 1: 200 -1589464.959 2.654 4.309
Chain 1: 300 -891035.255 2.031 1.000
Chain 1: 400 -458397.541 1.759 1.000
Chain 1: 500 -358442.281 1.463 0.944
Chain 1: 600 -233350.745 1.309 0.944
Chain 1: 700 -119658.671 1.257 0.944
Chain 1: 800 -86910.648 1.147 0.944
Chain 1: 900 -67280.371 1.052 0.784
Chain 1: 1000 -52113.880 0.976 0.784
Chain 1: 1100 -39620.009 0.908 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38806.000 0.479 0.377
Chain 1: 1300 -26778.895 0.445 0.377
Chain 1: 1400 -26502.778 0.352 0.315
Chain 1: 1500 -23094.403 0.339 0.315
Chain 1: 1600 -22313.447 0.289 0.292
Chain 1: 1700 -21188.196 0.199 0.291
Chain 1: 1800 -21133.185 0.162 0.148
Chain 1: 1900 -21460.183 0.134 0.053
Chain 1: 2000 -19970.430 0.112 0.053
Chain 1: 2100 -20208.839 0.082 0.035
Chain 1: 2200 -20435.841 0.081 0.035
Chain 1: 2300 -20052.337 0.038 0.019
Chain 1: 2400 -19824.158 0.038 0.019
Chain 1: 2500 -19626.095 0.024 0.015
Chain 1: 2600 -19255.495 0.023 0.015
Chain 1: 2700 -19212.188 0.018 0.012
Chain 1: 2800 -18928.659 0.019 0.015
Chain 1: 2900 -19210.276 0.019 0.015
Chain 1: 3000 -19196.379 0.012 0.012
Chain 1: 3100 -19281.542 0.011 0.012
Chain 1: 3200 -18971.653 0.011 0.015
Chain 1: 3300 -19176.811 0.010 0.012
Chain 1: 3400 -18650.722 0.012 0.015
Chain 1: 3500 -19264.097 0.014 0.015
Chain 1: 3600 -18568.759 0.016 0.015
Chain 1: 3700 -18957.070 0.018 0.016
Chain 1: 3800 -17913.660 0.022 0.020
Chain 1: 3900 -17909.703 0.021 0.020
Chain 1: 4000 -18027.023 0.021 0.020
Chain 1: 4100 -17940.661 0.021 0.020
Chain 1: 4200 -17756.187 0.021 0.020
Chain 1: 4300 -17895.096 0.021 0.020
Chain 1: 4400 -17851.363 0.018 0.010
Chain 1: 4500 -17753.779 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001387 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.87 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12461.077 1.000 1.000
Chain 1: 200 -9374.102 0.665 1.000
Chain 1: 300 -8061.622 0.497 0.329
Chain 1: 400 -8288.141 0.380 0.329
Chain 1: 500 -7916.646 0.313 0.163
Chain 1: 600 -8011.260 0.263 0.163
Chain 1: 700 -8130.812 0.228 0.047
Chain 1: 800 -7987.465 0.201 0.047
Chain 1: 900 -7868.910 0.181 0.027
Chain 1: 1000 -8052.276 0.165 0.027
Chain 1: 1100 -8041.753 0.065 0.023
Chain 1: 1200 -7963.934 0.033 0.018
Chain 1: 1300 -7913.176 0.017 0.015
Chain 1: 1400 -7926.969 0.015 0.015
Chain 1: 1500 -8016.738 0.011 0.012
Chain 1: 1600 -7969.876 0.011 0.011
Chain 1: 1700 -7909.602 0.010 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001402 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.02 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56969.645 1.000 1.000
Chain 1: 200 -17615.572 1.617 2.234
Chain 1: 300 -8773.263 1.414 1.008
Chain 1: 400 -8372.694 1.072 1.008
Chain 1: 500 -8557.000 0.862 1.000
Chain 1: 600 -8709.909 0.721 1.000
Chain 1: 700 -8079.134 0.630 0.078
Chain 1: 800 -7949.413 0.553 0.078
Chain 1: 900 -7898.978 0.492 0.048
Chain 1: 1000 -7990.030 0.444 0.048
Chain 1: 1100 -7780.393 0.347 0.027
Chain 1: 1200 -7622.219 0.125 0.022
Chain 1: 1300 -7780.686 0.027 0.021
Chain 1: 1400 -7899.110 0.023 0.020
Chain 1: 1500 -7608.264 0.025 0.020
Chain 1: 1600 -7772.570 0.025 0.021
Chain 1: 1700 -7576.401 0.020 0.021
Chain 1: 1800 -7682.200 0.020 0.021
Chain 1: 1900 -7494.270 0.022 0.021
Chain 1: 2000 -7642.237 0.023 0.021
Chain 1: 2100 -7627.744 0.020 0.021
Chain 1: 2200 -7745.816 0.020 0.020
Chain 1: 2300 -7613.976 0.019 0.019
Chain 1: 2400 -7664.777 0.018 0.019
Chain 1: 2500 -7609.585 0.015 0.017
Chain 1: 2600 -7544.464 0.014 0.015
Chain 1: 2700 -7582.448 0.012 0.014
Chain 1: 2800 -7522.715 0.011 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003314 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.14 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86269.216 1.000 1.000
Chain 1: 200 -13605.630 3.170 5.341
Chain 1: 300 -9950.729 2.236 1.000
Chain 1: 400 -10881.786 1.698 1.000
Chain 1: 500 -8891.836 1.403 0.367
Chain 1: 600 -8750.385 1.172 0.367
Chain 1: 700 -8493.394 1.009 0.224
Chain 1: 800 -8933.955 0.889 0.224
Chain 1: 900 -8705.489 0.793 0.086
Chain 1: 1000 -8518.812 0.716 0.086
Chain 1: 1100 -8626.916 0.617 0.049 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8393.332 0.086 0.030
Chain 1: 1300 -8635.954 0.052 0.028
Chain 1: 1400 -8567.369 0.044 0.028
Chain 1: 1500 -8498.565 0.023 0.026
Chain 1: 1600 -8604.669 0.022 0.026
Chain 1: 1700 -8687.428 0.020 0.022
Chain 1: 1800 -8263.927 0.021 0.022
Chain 1: 1900 -8365.129 0.019 0.013
Chain 1: 2000 -8339.408 0.017 0.012
Chain 1: 2100 -8464.590 0.018 0.012
Chain 1: 2200 -8269.471 0.017 0.012
Chain 1: 2300 -8359.706 0.015 0.012
Chain 1: 2400 -8428.710 0.015 0.012
Chain 1: 2500 -8374.896 0.015 0.012
Chain 1: 2600 -8375.986 0.014 0.011
Chain 1: 2700 -8292.846 0.014 0.011
Chain 1: 2800 -8253.121 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003407 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.07 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8379314.680 1.000 1.000
Chain 1: 200 -1580778.031 2.650 4.301
Chain 1: 300 -890916.846 2.025 1.000
Chain 1: 400 -457964.775 1.755 1.000
Chain 1: 500 -358799.346 1.459 0.945
Chain 1: 600 -233796.608 1.305 0.945
Chain 1: 700 -119704.230 1.255 0.945
Chain 1: 800 -86838.477 1.145 0.945
Chain 1: 900 -67113.617 1.051 0.774
Chain 1: 1000 -51854.807 0.975 0.774
Chain 1: 1100 -39276.522 0.907 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38446.904 0.479 0.378
Chain 1: 1300 -26342.979 0.448 0.378
Chain 1: 1400 -26057.381 0.354 0.320
Chain 1: 1500 -22628.719 0.342 0.320
Chain 1: 1600 -21840.754 0.292 0.294
Chain 1: 1700 -20706.979 0.202 0.294
Chain 1: 1800 -20649.452 0.165 0.152
Chain 1: 1900 -20975.723 0.137 0.055
Chain 1: 2000 -19482.676 0.115 0.055
Chain 1: 2100 -19721.351 0.084 0.036
Chain 1: 2200 -19948.532 0.083 0.036
Chain 1: 2300 -19565.027 0.039 0.020
Chain 1: 2400 -19336.972 0.039 0.020
Chain 1: 2500 -19139.295 0.025 0.016
Chain 1: 2600 -18769.158 0.023 0.016
Chain 1: 2700 -18725.981 0.018 0.012
Chain 1: 2800 -18442.925 0.019 0.015
Chain 1: 2900 -18724.298 0.019 0.015
Chain 1: 3000 -18710.413 0.012 0.012
Chain 1: 3100 -18795.447 0.011 0.012
Chain 1: 3200 -18486.017 0.012 0.015
Chain 1: 3300 -18690.814 0.011 0.012
Chain 1: 3400 -18165.644 0.012 0.015
Chain 1: 3500 -18777.832 0.015 0.015
Chain 1: 3600 -18084.097 0.017 0.015
Chain 1: 3700 -18471.262 0.018 0.017
Chain 1: 3800 -17430.460 0.023 0.021
Chain 1: 3900 -17426.615 0.021 0.021
Chain 1: 4000 -17543.866 0.022 0.021
Chain 1: 4100 -17457.655 0.022 0.021
Chain 1: 4200 -17273.739 0.021 0.021
Chain 1: 4300 -17412.215 0.021 0.021
Chain 1: 4400 -17368.947 0.018 0.011
Chain 1: 4500 -17271.472 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001241 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.41 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49422.913 1.000 1.000
Chain 1: 200 -24277.066 1.018 1.036
Chain 1: 300 -14176.825 0.916 1.000
Chain 1: 400 -12838.633 0.713 1.000
Chain 1: 500 -13034.591 0.573 0.712
Chain 1: 600 -16672.820 0.514 0.712
Chain 1: 700 -15171.293 0.455 0.218
Chain 1: 800 -15454.421 0.400 0.218
Chain 1: 900 -19785.619 0.380 0.218
Chain 1: 1000 -12881.033 0.396 0.219
Chain 1: 1100 -20743.547 0.334 0.219
Chain 1: 1200 -14992.673 0.268 0.219
Chain 1: 1300 -13915.524 0.205 0.218
Chain 1: 1400 -13348.878 0.199 0.218
Chain 1: 1500 -11386.091 0.215 0.218
Chain 1: 1600 -11954.177 0.197 0.172
Chain 1: 1700 -10002.994 0.207 0.195
Chain 1: 1800 -10237.683 0.208 0.195
Chain 1: 1900 -10532.418 0.188 0.172
Chain 1: 2000 -24010.635 0.191 0.172
Chain 1: 2100 -10102.855 0.291 0.172
Chain 1: 2200 -10028.320 0.253 0.077
Chain 1: 2300 -11924.749 0.261 0.159
Chain 1: 2400 -10255.129 0.273 0.163
Chain 1: 2500 -16978.870 0.296 0.163
Chain 1: 2600 -9721.769 0.366 0.195
Chain 1: 2700 -9554.166 0.348 0.163
Chain 1: 2800 -20024.872 0.398 0.396
Chain 1: 2900 -15524.802 0.424 0.396
Chain 1: 3000 -9378.747 0.433 0.396
Chain 1: 3100 -10215.110 0.304 0.290
Chain 1: 3200 -9634.362 0.309 0.290
Chain 1: 3300 -17083.820 0.337 0.396
Chain 1: 3400 -10498.625 0.383 0.436
Chain 1: 3500 -9576.900 0.353 0.436
Chain 1: 3600 -19718.380 0.330 0.436
Chain 1: 3700 -9015.956 0.447 0.514 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 3800 -10910.791 0.412 0.436
Chain 1: 3900 -15560.842 0.413 0.436
Chain 1: 4000 -10401.133 0.397 0.436
Chain 1: 4100 -9617.690 0.397 0.436
Chain 1: 4200 -9022.431 0.398 0.436
Chain 1: 4300 -15470.320 0.396 0.417
Chain 1: 4400 -9504.221 0.396 0.417
Chain 1: 4500 -9343.737 0.388 0.417
Chain 1: 4600 -8901.752 0.341 0.299
Chain 1: 4700 -9535.456 0.229 0.174
Chain 1: 4800 -9156.090 0.216 0.081
Chain 1: 4900 -9340.995 0.188 0.066
Chain 1: 5000 -10542.743 0.150 0.066
Chain 1: 5100 -9322.792 0.155 0.066
Chain 1: 5200 -17005.411 0.194 0.114
Chain 1: 5300 -12534.896 0.188 0.114
Chain 1: 5400 -16533.360 0.149 0.114
Chain 1: 5500 -13084.912 0.174 0.131
Chain 1: 5600 -13295.572 0.170 0.131
Chain 1: 5700 -12633.183 0.169 0.131
Chain 1: 5800 -9279.544 0.201 0.242
Chain 1: 5900 -8783.258 0.204 0.242
Chain 1: 6000 -9313.198 0.199 0.242
Chain 1: 6100 -9097.366 0.188 0.242
Chain 1: 6200 -8933.154 0.145 0.057
Chain 1: 6300 -9109.987 0.111 0.057
Chain 1: 6400 -8975.462 0.088 0.052
Chain 1: 6500 -8787.894 0.064 0.024
Chain 1: 6600 -8966.383 0.064 0.024
Chain 1: 6700 -10975.785 0.078 0.024
Chain 1: 6800 -9579.697 0.056 0.024
Chain 1: 6900 -9211.982 0.054 0.024
Chain 1: 7000 -12949.314 0.078 0.024
Chain 1: 7100 -8502.119 0.127 0.040
Chain 1: 7200 -10278.220 0.143 0.146
Chain 1: 7300 -10861.778 0.146 0.146
Chain 1: 7400 -8424.503 0.174 0.173
Chain 1: 7500 -10393.289 0.191 0.183
Chain 1: 7600 -9646.188 0.196 0.183
Chain 1: 7700 -9333.574 0.181 0.173
Chain 1: 7800 -11969.212 0.189 0.189
Chain 1: 7900 -8893.601 0.219 0.220
Chain 1: 8000 -13101.394 0.223 0.220
Chain 1: 8100 -8981.894 0.216 0.220
Chain 1: 8200 -8973.393 0.199 0.220
Chain 1: 8300 -8536.342 0.199 0.220
Chain 1: 8400 -8332.239 0.172 0.189
Chain 1: 8500 -11684.661 0.182 0.220
Chain 1: 8600 -8438.205 0.213 0.287
Chain 1: 8700 -8339.775 0.211 0.287
Chain 1: 8800 -8411.822 0.189 0.287
Chain 1: 8900 -10222.943 0.173 0.177
Chain 1: 9000 -9351.277 0.150 0.093
Chain 1: 9100 -10206.908 0.112 0.084
Chain 1: 9200 -10113.725 0.113 0.084
Chain 1: 9300 -9358.657 0.116 0.084
Chain 1: 9400 -8458.431 0.124 0.093
Chain 1: 9500 -9211.048 0.104 0.084
Chain 1: 9600 -8952.773 0.068 0.082
Chain 1: 9700 -8388.316 0.074 0.082
Chain 1: 9800 -8574.047 0.075 0.082
Chain 1: 9900 -10970.921 0.079 0.082
Chain 1: 10000 -9697.220 0.083 0.082
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001375 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.75 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58385.774 1.000 1.000
Chain 1: 200 -18040.147 1.618 2.236
Chain 1: 300 -8846.126 1.425 1.039
Chain 1: 400 -8030.060 1.094 1.039
Chain 1: 500 -8164.704 0.879 1.000
Chain 1: 600 -8740.080 0.743 1.000
Chain 1: 700 -8551.408 0.640 0.102
Chain 1: 800 -8324.917 0.564 0.102
Chain 1: 900 -7798.458 0.508 0.068
Chain 1: 1000 -7914.514 0.459 0.068
Chain 1: 1100 -7942.568 0.359 0.066
Chain 1: 1200 -7590.850 0.140 0.046
Chain 1: 1300 -7854.579 0.040 0.034
Chain 1: 1400 -7792.296 0.031 0.027
Chain 1: 1500 -7580.676 0.032 0.028
Chain 1: 1600 -7660.182 0.026 0.027
Chain 1: 1700 -7586.271 0.025 0.027
Chain 1: 1800 -7674.436 0.023 0.015
Chain 1: 1900 -7604.415 0.017 0.011
Chain 1: 2000 -7657.963 0.017 0.010
Chain 1: 2100 -7588.053 0.017 0.010
Chain 1: 2200 -7928.770 0.017 0.010
Chain 1: 2300 -7530.476 0.019 0.010
Chain 1: 2400 -7700.850 0.020 0.011
Chain 1: 2500 -7568.952 0.019 0.011
Chain 1: 2600 -7537.298 0.019 0.011
Chain 1: 2700 -7507.129 0.018 0.011
Chain 1: 2800 -7506.879 0.017 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003148 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.48 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86689.640 1.000 1.000
Chain 1: 200 -13877.452 3.123 5.247
Chain 1: 300 -10147.231 2.205 1.000
Chain 1: 400 -11483.973 1.683 1.000
Chain 1: 500 -9163.636 1.397 0.368
Chain 1: 600 -8496.189 1.177 0.368
Chain 1: 700 -8490.106 1.009 0.253
Chain 1: 800 -9560.775 0.897 0.253
Chain 1: 900 -8951.161 0.805 0.116
Chain 1: 1000 -8979.170 0.725 0.116
Chain 1: 1100 -8911.964 0.625 0.112 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8372.845 0.107 0.079
Chain 1: 1300 -8784.210 0.075 0.068
Chain 1: 1400 -8798.232 0.064 0.064
Chain 1: 1500 -8644.147 0.040 0.047
Chain 1: 1600 -8758.105 0.034 0.018
Chain 1: 1700 -8813.224 0.034 0.018
Chain 1: 1800 -8373.077 0.028 0.018
Chain 1: 1900 -8477.159 0.023 0.013
Chain 1: 2000 -8460.909 0.022 0.013
Chain 1: 2100 -8582.258 0.023 0.014
Chain 1: 2200 -8374.654 0.019 0.014
Chain 1: 2300 -8469.466 0.016 0.013
Chain 1: 2400 -8536.765 0.016 0.013
Chain 1: 2500 -8484.953 0.015 0.012
Chain 1: 2600 -8498.227 0.014 0.011
Chain 1: 2700 -8406.066 0.014 0.011
Chain 1: 2800 -8353.985 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003021 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.21 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8401813.652 1.000 1.000
Chain 1: 200 -1586954.056 2.647 4.294
Chain 1: 300 -891831.409 2.025 1.000
Chain 1: 400 -458520.299 1.755 1.000
Chain 1: 500 -358511.532 1.460 0.945
Chain 1: 600 -233385.671 1.306 0.945
Chain 1: 700 -119600.655 1.255 0.945
Chain 1: 800 -86811.511 1.145 0.945
Chain 1: 900 -67160.670 1.051 0.779
Chain 1: 1000 -51972.605 0.975 0.779
Chain 1: 1100 -39459.637 0.906 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38643.642 0.479 0.378
Chain 1: 1300 -26602.545 0.446 0.378
Chain 1: 1400 -26325.205 0.353 0.317
Chain 1: 1500 -22911.843 0.340 0.317
Chain 1: 1600 -22129.130 0.290 0.293
Chain 1: 1700 -21002.454 0.200 0.292
Chain 1: 1800 -20947.054 0.163 0.149
Chain 1: 1900 -21273.682 0.135 0.054
Chain 1: 2000 -19783.605 0.113 0.054
Chain 1: 2100 -20022.271 0.083 0.035
Chain 1: 2200 -20248.960 0.082 0.035
Chain 1: 2300 -19865.769 0.038 0.019
Chain 1: 2400 -19637.628 0.039 0.019
Chain 1: 2500 -19439.590 0.025 0.015
Chain 1: 2600 -19069.343 0.023 0.015
Chain 1: 2700 -19026.200 0.018 0.012
Chain 1: 2800 -18742.725 0.019 0.015
Chain 1: 2900 -19024.246 0.019 0.015
Chain 1: 3000 -19010.455 0.012 0.012
Chain 1: 3100 -19095.502 0.011 0.012
Chain 1: 3200 -18785.875 0.011 0.015
Chain 1: 3300 -18990.851 0.011 0.012
Chain 1: 3400 -18465.152 0.012 0.015
Chain 1: 3500 -19077.957 0.014 0.015
Chain 1: 3600 -18383.417 0.016 0.015
Chain 1: 3700 -18771.065 0.018 0.016
Chain 1: 3800 -17728.906 0.023 0.021
Chain 1: 3900 -17724.974 0.021 0.021
Chain 1: 4000 -17842.303 0.022 0.021
Chain 1: 4100 -17755.928 0.022 0.021
Chain 1: 4200 -17571.784 0.021 0.021
Chain 1: 4300 -17710.479 0.021 0.021
Chain 1: 4400 -17666.962 0.018 0.010
Chain 1: 4500 -17569.408 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001139 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 11.39 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12467.344 1.000 1.000
Chain 1: 200 -9342.389 0.667 1.000
Chain 1: 300 -8297.026 0.487 0.334
Chain 1: 400 -8401.840 0.368 0.334
Chain 1: 500 -8263.551 0.298 0.126
Chain 1: 600 -8122.161 0.251 0.126
Chain 1: 700 -8036.817 0.217 0.017
Chain 1: 800 -8046.310 0.190 0.017
Chain 1: 900 -7966.265 0.170 0.017
Chain 1: 1000 -8145.500 0.155 0.017
Chain 1: 1100 -8177.067 0.055 0.017
Chain 1: 1200 -8074.282 0.023 0.013
Chain 1: 1300 -8019.064 0.011 0.012
Chain 1: 1400 -8035.107 0.010 0.011
Chain 1: 1500 -8120.893 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001438 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.38 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57142.865 1.000 1.000
Chain 1: 200 -17531.853 1.630 2.259
Chain 1: 300 -8821.807 1.416 1.000
Chain 1: 400 -8311.732 1.077 1.000
Chain 1: 500 -8464.583 0.865 0.987
Chain 1: 600 -8493.766 0.722 0.987
Chain 1: 700 -8068.983 0.626 0.061
Chain 1: 800 -8352.850 0.552 0.061
Chain 1: 900 -8006.927 0.495 0.053
Chain 1: 1000 -8065.976 0.447 0.053
Chain 1: 1100 -7723.806 0.351 0.044
Chain 1: 1200 -7836.170 0.127 0.043
Chain 1: 1300 -7828.019 0.028 0.034
Chain 1: 1400 -7995.139 0.024 0.021
Chain 1: 1500 -7676.889 0.026 0.034
Chain 1: 1600 -7745.963 0.027 0.034
Chain 1: 1700 -7608.042 0.023 0.021
Chain 1: 1800 -7695.124 0.021 0.018
Chain 1: 1900 -7724.574 0.017 0.014
Chain 1: 2000 -7677.914 0.017 0.014
Chain 1: 2100 -7668.043 0.013 0.011
Chain 1: 2200 -7780.414 0.013 0.011
Chain 1: 2300 -7678.803 0.014 0.013
Chain 1: 2400 -7713.164 0.012 0.011
Chain 1: 2500 -7707.374 0.008 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00303 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.3 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86696.952 1.000 1.000
Chain 1: 200 -13603.514 3.187 5.373
Chain 1: 300 -9987.140 2.245 1.000
Chain 1: 400 -10979.706 1.706 1.000
Chain 1: 500 -8934.674 1.411 0.362
Chain 1: 600 -8519.171 1.184 0.362
Chain 1: 700 -8541.363 1.015 0.229
Chain 1: 800 -9039.989 0.895 0.229
Chain 1: 900 -8842.093 0.798 0.090
Chain 1: 1000 -8498.513 0.722 0.090
Chain 1: 1100 -8868.243 0.627 0.055 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8473.354 0.094 0.049
Chain 1: 1300 -8746.389 0.061 0.047
Chain 1: 1400 -8495.269 0.055 0.042
Chain 1: 1500 -8541.090 0.032 0.040
Chain 1: 1600 -8536.219 0.028 0.031
Chain 1: 1700 -8446.840 0.028 0.031
Chain 1: 1800 -8341.276 0.024 0.030
Chain 1: 1900 -8463.139 0.023 0.030
Chain 1: 2000 -8424.722 0.020 0.014
Chain 1: 2100 -8550.678 0.017 0.014
Chain 1: 2200 -8353.704 0.015 0.014
Chain 1: 2300 -8493.591 0.013 0.014
Chain 1: 2400 -8366.483 0.012 0.014
Chain 1: 2500 -8431.552 0.012 0.014
Chain 1: 2600 -8456.079 0.012 0.014
Chain 1: 2700 -8373.635 0.012 0.014
Chain 1: 2800 -8345.217 0.011 0.014
Chain 1: 2900 -8400.936 0.011 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003344 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.44 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8410965.622 1.000 1.000
Chain 1: 200 -1587421.383 2.649 4.299
Chain 1: 300 -892285.217 2.026 1.000
Chain 1: 400 -459032.420 1.755 1.000
Chain 1: 500 -359057.220 1.460 0.944
Chain 1: 600 -233592.520 1.306 0.944
Chain 1: 700 -119484.268 1.256 0.944
Chain 1: 800 -86659.111 1.146 0.944
Chain 1: 900 -66949.850 1.052 0.779
Chain 1: 1000 -51719.877 0.976 0.779
Chain 1: 1100 -39179.223 0.908 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38349.266 0.480 0.379
Chain 1: 1300 -26290.803 0.448 0.379
Chain 1: 1400 -26006.956 0.355 0.320
Chain 1: 1500 -22592.045 0.342 0.320
Chain 1: 1600 -21807.614 0.292 0.294
Chain 1: 1700 -20679.759 0.202 0.294
Chain 1: 1800 -20623.397 0.164 0.151
Chain 1: 1900 -20949.338 0.137 0.055
Chain 1: 2000 -19460.321 0.115 0.055
Chain 1: 2100 -19698.466 0.084 0.036
Chain 1: 2200 -19925.095 0.083 0.036
Chain 1: 2300 -19542.214 0.039 0.020
Chain 1: 2400 -19314.386 0.039 0.020
Chain 1: 2500 -19116.580 0.025 0.016
Chain 1: 2600 -18746.757 0.023 0.016
Chain 1: 2700 -18703.700 0.018 0.012
Chain 1: 2800 -18420.775 0.019 0.015
Chain 1: 2900 -18701.943 0.019 0.015
Chain 1: 3000 -18688.040 0.012 0.012
Chain 1: 3100 -18773.052 0.011 0.012
Chain 1: 3200 -18463.792 0.012 0.015
Chain 1: 3300 -18668.460 0.011 0.012
Chain 1: 3400 -18143.618 0.012 0.015
Chain 1: 3500 -18755.189 0.015 0.015
Chain 1: 3600 -18062.250 0.017 0.015
Chain 1: 3700 -18448.810 0.018 0.017
Chain 1: 3800 -17409.154 0.023 0.021
Chain 1: 3900 -17405.343 0.021 0.021
Chain 1: 4000 -17522.613 0.022 0.021
Chain 1: 4100 -17436.460 0.022 0.021
Chain 1: 4200 -17252.810 0.021 0.021
Chain 1: 4300 -17391.083 0.021 0.021
Chain 1: 4400 -17348.011 0.018 0.011
Chain 1: 4500 -17250.611 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001242 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.42 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48736.918 1.000 1.000
Chain 1: 200 -21371.346 1.140 1.280
Chain 1: 300 -17824.293 0.826 1.000
Chain 1: 400 -21581.803 0.663 1.000
Chain 1: 500 -18223.381 0.568 0.199
Chain 1: 600 -16714.266 0.488 0.199
Chain 1: 700 -19493.676 0.439 0.184
Chain 1: 800 -11822.221 0.465 0.199
Chain 1: 900 -14235.147 0.432 0.184
Chain 1: 1000 -10834.621 0.420 0.199
Chain 1: 1100 -22866.760 0.373 0.199
Chain 1: 1200 -12198.041 0.332 0.199
Chain 1: 1300 -11989.768 0.314 0.184
Chain 1: 1400 -12385.568 0.300 0.184
Chain 1: 1500 -10026.577 0.305 0.235
Chain 1: 1600 -13165.263 0.320 0.238
Chain 1: 1700 -10247.809 0.334 0.285
Chain 1: 1800 -11567.377 0.281 0.238
Chain 1: 1900 -10101.070 0.278 0.238
Chain 1: 2000 -10788.942 0.253 0.235
Chain 1: 2100 -9345.208 0.216 0.154
Chain 1: 2200 -12346.780 0.153 0.154
Chain 1: 2300 -9474.597 0.181 0.235
Chain 1: 2400 -9139.543 0.182 0.235
Chain 1: 2500 -9273.047 0.160 0.154
Chain 1: 2600 -15386.105 0.176 0.154
Chain 1: 2700 -9126.004 0.216 0.154
Chain 1: 2800 -18855.784 0.256 0.243
Chain 1: 2900 -10526.923 0.321 0.303
Chain 1: 3000 -11168.489 0.320 0.303
Chain 1: 3100 -8803.715 0.331 0.303
Chain 1: 3200 -9370.227 0.313 0.303
Chain 1: 3300 -12461.178 0.308 0.269
Chain 1: 3400 -13381.496 0.311 0.269
Chain 1: 3500 -11056.091 0.330 0.269
Chain 1: 3600 -10019.114 0.301 0.248
Chain 1: 3700 -16824.146 0.273 0.248
Chain 1: 3800 -8927.050 0.310 0.248
Chain 1: 3900 -9651.387 0.238 0.210
Chain 1: 4000 -14130.364 0.264 0.248
Chain 1: 4100 -10101.787 0.277 0.248
Chain 1: 4200 -9669.155 0.276 0.248
Chain 1: 4300 -8709.131 0.262 0.210
Chain 1: 4400 -10945.646 0.275 0.210
Chain 1: 4500 -8786.928 0.279 0.246
Chain 1: 4600 -9944.211 0.280 0.246
Chain 1: 4700 -11284.809 0.252 0.204
Chain 1: 4800 -9603.639 0.181 0.175
Chain 1: 4900 -11313.800 0.188 0.175
Chain 1: 5000 -9549.766 0.175 0.175
Chain 1: 5100 -14318.755 0.168 0.175
Chain 1: 5200 -8889.910 0.225 0.185
Chain 1: 5300 -12392.901 0.242 0.204
Chain 1: 5400 -8955.574 0.260 0.246
Chain 1: 5500 -13746.408 0.270 0.283
Chain 1: 5600 -9053.828 0.311 0.333
Chain 1: 5700 -8504.796 0.305 0.333
Chain 1: 5800 -10796.679 0.309 0.333
Chain 1: 5900 -10119.621 0.301 0.333
Chain 1: 6000 -8664.021 0.299 0.333
Chain 1: 6100 -8797.987 0.267 0.283
Chain 1: 6200 -8448.294 0.210 0.212
Chain 1: 6300 -8412.974 0.182 0.168
Chain 1: 6400 -11408.029 0.170 0.168
Chain 1: 6500 -9324.716 0.158 0.168
Chain 1: 6600 -12464.185 0.131 0.168
Chain 1: 6700 -8325.265 0.174 0.212
Chain 1: 6800 -12707.201 0.188 0.223
Chain 1: 6900 -8595.203 0.229 0.252
Chain 1: 7000 -8205.244 0.217 0.252
Chain 1: 7100 -8330.853 0.217 0.252
Chain 1: 7200 -8134.010 0.215 0.252
Chain 1: 7300 -9268.174 0.227 0.252
Chain 1: 7400 -13705.146 0.233 0.252
Chain 1: 7500 -11166.775 0.233 0.252
Chain 1: 7600 -8940.111 0.233 0.249
Chain 1: 7700 -8243.149 0.192 0.227
Chain 1: 7800 -10983.113 0.182 0.227
Chain 1: 7900 -8128.511 0.169 0.227
Chain 1: 8000 -8109.409 0.165 0.227
Chain 1: 8100 -8793.189 0.171 0.227
Chain 1: 8200 -8265.896 0.175 0.227
Chain 1: 8300 -9043.650 0.172 0.227
Chain 1: 8400 -8470.191 0.146 0.086
Chain 1: 8500 -8171.642 0.127 0.085
Chain 1: 8600 -9190.307 0.113 0.085
Chain 1: 8700 -7999.941 0.119 0.086
Chain 1: 8800 -8344.210 0.099 0.078
Chain 1: 8900 -8400.535 0.064 0.068
Chain 1: 9000 -8358.860 0.064 0.068
Chain 1: 9100 -11816.230 0.086 0.068
Chain 1: 9200 -8982.184 0.111 0.086
Chain 1: 9300 -9495.783 0.108 0.068
Chain 1: 9400 -13457.240 0.131 0.111
Chain 1: 9500 -7974.774 0.196 0.149
Chain 1: 9600 -8752.592 0.193 0.149
Chain 1: 9700 -8164.531 0.186 0.089
Chain 1: 9800 -10075.471 0.201 0.190
Chain 1: 9900 -8664.990 0.216 0.190
Chain 1: 10000 -10788.928 0.235 0.197
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001426 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.26 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58183.132 1.000 1.000
Chain 1: 200 -17464.472 1.666 2.332
Chain 1: 300 -8621.204 1.452 1.026
Chain 1: 400 -8040.019 1.107 1.026
Chain 1: 500 -8073.572 0.887 1.000
Chain 1: 600 -8902.579 0.754 1.000
Chain 1: 700 -7933.895 0.664 0.122
Chain 1: 800 -8091.939 0.584 0.122
Chain 1: 900 -7945.906 0.521 0.093
Chain 1: 1000 -7770.763 0.471 0.093
Chain 1: 1100 -7731.772 0.371 0.072
Chain 1: 1200 -7624.566 0.140 0.023
Chain 1: 1300 -7739.073 0.039 0.020
Chain 1: 1400 -7979.935 0.034 0.020
Chain 1: 1500 -7628.550 0.039 0.023
Chain 1: 1600 -7611.211 0.029 0.020
Chain 1: 1700 -7525.468 0.018 0.018
Chain 1: 1800 -7603.384 0.017 0.015
Chain 1: 1900 -7572.205 0.016 0.014
Chain 1: 2000 -7655.077 0.015 0.011
Chain 1: 2100 -7605.603 0.015 0.011
Chain 1: 2200 -7704.842 0.015 0.011
Chain 1: 2300 -7616.093 0.015 0.011
Chain 1: 2400 -7655.032 0.012 0.011
Chain 1: 2500 -7591.041 0.008 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002919 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.19 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86033.954 1.000 1.000
Chain 1: 200 -13367.048 3.218 5.436
Chain 1: 300 -9761.086 2.269 1.000
Chain 1: 400 -10611.932 1.721 1.000
Chain 1: 500 -8665.494 1.422 0.369
Chain 1: 600 -8362.918 1.191 0.369
Chain 1: 700 -8373.191 1.021 0.225
Chain 1: 800 -8634.074 0.897 0.225
Chain 1: 900 -8540.466 0.799 0.080
Chain 1: 1000 -8364.324 0.721 0.080
Chain 1: 1100 -8582.676 0.624 0.036 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8288.042 0.083 0.036
Chain 1: 1300 -8475.831 0.049 0.030
Chain 1: 1400 -8477.493 0.041 0.025
Chain 1: 1500 -8336.910 0.020 0.022
Chain 1: 1600 -8448.680 0.018 0.021
Chain 1: 1700 -8535.512 0.019 0.021
Chain 1: 1800 -8129.687 0.021 0.021
Chain 1: 1900 -8226.110 0.021 0.021
Chain 1: 2000 -8198.254 0.019 0.017
Chain 1: 2100 -8318.762 0.018 0.014
Chain 1: 2200 -8129.375 0.017 0.014
Chain 1: 2300 -8265.883 0.016 0.014
Chain 1: 2400 -8273.032 0.016 0.014
Chain 1: 2500 -8239.422 0.015 0.013
Chain 1: 2600 -8237.394 0.013 0.012
Chain 1: 2700 -8151.458 0.014 0.012
Chain 1: 2800 -8116.653 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003333 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.33 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8386555.561 1.000 1.000
Chain 1: 200 -1581368.727 2.652 4.303
Chain 1: 300 -889727.915 2.027 1.000
Chain 1: 400 -456734.884 1.757 1.000
Chain 1: 500 -357387.436 1.461 0.948
Chain 1: 600 -232440.729 1.307 0.948
Chain 1: 700 -118941.516 1.257 0.948
Chain 1: 800 -86187.026 1.147 0.948
Chain 1: 900 -66567.232 1.053 0.777
Chain 1: 1000 -51387.137 0.977 0.777
Chain 1: 1100 -38884.026 0.909 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38063.645 0.481 0.380
Chain 1: 1300 -26041.077 0.449 0.380
Chain 1: 1400 -25761.153 0.356 0.322
Chain 1: 1500 -22353.576 0.343 0.322
Chain 1: 1600 -21571.527 0.293 0.295
Chain 1: 1700 -20447.992 0.203 0.295
Chain 1: 1800 -20392.935 0.165 0.152
Chain 1: 1900 -20718.903 0.137 0.055
Chain 1: 2000 -19231.802 0.116 0.055
Chain 1: 2100 -19470.039 0.085 0.036
Chain 1: 2200 -19696.121 0.084 0.036
Chain 1: 2300 -19313.707 0.039 0.020
Chain 1: 2400 -19085.920 0.039 0.020
Chain 1: 2500 -18887.771 0.025 0.016
Chain 1: 2600 -18518.248 0.024 0.016
Chain 1: 2700 -18475.374 0.018 0.012
Chain 1: 2800 -18192.235 0.020 0.016
Chain 1: 2900 -18473.373 0.020 0.015
Chain 1: 3000 -18459.594 0.012 0.012
Chain 1: 3100 -18544.545 0.011 0.012
Chain 1: 3200 -18235.381 0.012 0.015
Chain 1: 3300 -18440.011 0.011 0.012
Chain 1: 3400 -17915.135 0.013 0.015
Chain 1: 3500 -18526.667 0.015 0.016
Chain 1: 3600 -17833.828 0.017 0.016
Chain 1: 3700 -18220.233 0.019 0.017
Chain 1: 3800 -17180.651 0.023 0.021
Chain 1: 3900 -17176.831 0.022 0.021
Chain 1: 4000 -17294.127 0.022 0.021
Chain 1: 4100 -17207.887 0.022 0.021
Chain 1: 4200 -17024.370 0.022 0.021
Chain 1: 4300 -17162.635 0.021 0.021
Chain 1: 4400 -17119.587 0.019 0.011
Chain 1: 4500 -17022.162 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001299 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.99 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49187.350 1.000 1.000
Chain 1: 200 -20897.009 1.177 1.354
Chain 1: 300 -15583.302 0.898 1.000
Chain 1: 400 -14573.276 0.691 1.000
Chain 1: 500 -12545.329 0.585 0.341
Chain 1: 600 -22086.121 0.560 0.432
Chain 1: 700 -15785.203 0.537 0.399
Chain 1: 800 -14328.320 0.482 0.399
Chain 1: 900 -14398.924 0.429 0.341
Chain 1: 1000 -13119.951 0.396 0.341
Chain 1: 1100 -16182.836 0.315 0.189
Chain 1: 1200 -11259.839 0.223 0.189
Chain 1: 1300 -12487.074 0.199 0.162
Chain 1: 1400 -10114.958 0.216 0.189
Chain 1: 1500 -10892.570 0.207 0.189
Chain 1: 1600 -12704.313 0.178 0.143
Chain 1: 1700 -12220.928 0.142 0.102
Chain 1: 1800 -30074.551 0.191 0.143
Chain 1: 1900 -10851.588 0.368 0.189
Chain 1: 2000 -18780.031 0.400 0.235
Chain 1: 2100 -9494.008 0.479 0.422
Chain 1: 2200 -9665.711 0.437 0.235
Chain 1: 2300 -9311.275 0.431 0.235
Chain 1: 2400 -10212.365 0.416 0.143
Chain 1: 2500 -9445.335 0.417 0.143
Chain 1: 2600 -9572.121 0.404 0.088
Chain 1: 2700 -9309.280 0.403 0.088
Chain 1: 2800 -11236.707 0.361 0.088
Chain 1: 2900 -9772.643 0.199 0.088
Chain 1: 3000 -8915.062 0.166 0.088
Chain 1: 3100 -9984.142 0.079 0.088
Chain 1: 3200 -10298.582 0.080 0.088
Chain 1: 3300 -9049.271 0.090 0.096
Chain 1: 3400 -9054.671 0.082 0.096
Chain 1: 3500 -9615.144 0.079 0.096
Chain 1: 3600 -10485.778 0.086 0.096
Chain 1: 3700 -8698.877 0.104 0.107
Chain 1: 3800 -10294.605 0.102 0.107
Chain 1: 3900 -10200.196 0.088 0.096
Chain 1: 4000 -10843.089 0.085 0.083
Chain 1: 4100 -11669.458 0.081 0.071
Chain 1: 4200 -16151.035 0.106 0.083
Chain 1: 4300 -8838.874 0.175 0.083
Chain 1: 4400 -9455.179 0.181 0.083
Chain 1: 4500 -9984.744 0.181 0.083
Chain 1: 4600 -9728.217 0.175 0.071
Chain 1: 4700 -9794.349 0.155 0.065
Chain 1: 4800 -8529.599 0.154 0.065
Chain 1: 4900 -13724.004 0.191 0.071
Chain 1: 5000 -9394.531 0.231 0.148
Chain 1: 5100 -8602.174 0.234 0.148
Chain 1: 5200 -9002.801 0.210 0.092
Chain 1: 5300 -13516.178 0.161 0.092
Chain 1: 5400 -8732.511 0.209 0.148
Chain 1: 5500 -12008.974 0.231 0.273
Chain 1: 5600 -8617.556 0.268 0.334
Chain 1: 5700 -13046.381 0.301 0.339
Chain 1: 5800 -8740.125 0.336 0.378
Chain 1: 5900 -8362.372 0.302 0.339
Chain 1: 6000 -8694.257 0.260 0.334
Chain 1: 6100 -13161.800 0.285 0.339
Chain 1: 6200 -8663.565 0.332 0.339
Chain 1: 6300 -9120.440 0.304 0.339
Chain 1: 6400 -9847.192 0.256 0.339
Chain 1: 6500 -8828.366 0.241 0.339
Chain 1: 6600 -8572.532 0.204 0.115
Chain 1: 6700 -10674.939 0.190 0.115
Chain 1: 6800 -9395.614 0.154 0.115
Chain 1: 6900 -10959.339 0.164 0.136
Chain 1: 7000 -9353.252 0.178 0.143
Chain 1: 7100 -11176.362 0.160 0.143
Chain 1: 7200 -9065.679 0.131 0.143
Chain 1: 7300 -8322.671 0.135 0.143
Chain 1: 7400 -8562.669 0.131 0.143
Chain 1: 7500 -11609.585 0.145 0.163
Chain 1: 7600 -8707.088 0.176 0.172
Chain 1: 7700 -8328.141 0.161 0.163
Chain 1: 7800 -12592.129 0.181 0.172
Chain 1: 7900 -8403.217 0.216 0.233
Chain 1: 8000 -9642.646 0.212 0.233
Chain 1: 8100 -11750.684 0.214 0.233
Chain 1: 8200 -8137.551 0.235 0.262
Chain 1: 8300 -8111.268 0.226 0.262
Chain 1: 8400 -9574.912 0.239 0.262
Chain 1: 8500 -8179.573 0.229 0.179
Chain 1: 8600 -8065.399 0.198 0.171
Chain 1: 8700 -8184.701 0.194 0.171
Chain 1: 8800 -10014.513 0.179 0.171
Chain 1: 8900 -12369.708 0.148 0.171
Chain 1: 9000 -9382.445 0.167 0.179
Chain 1: 9100 -8216.765 0.163 0.171
Chain 1: 9200 -8291.016 0.120 0.153
Chain 1: 9300 -8888.636 0.126 0.153
Chain 1: 9400 -8207.031 0.119 0.142
Chain 1: 9500 -8514.101 0.106 0.083
Chain 1: 9600 -8426.582 0.105 0.083
Chain 1: 9700 -8189.905 0.107 0.083
Chain 1: 9800 -10351.820 0.109 0.083
Chain 1: 9900 -8305.669 0.115 0.083
Chain 1: 10000 -8482.340 0.085 0.067
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001382 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.82 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57509.286 1.000 1.000
Chain 1: 200 -17563.127 1.637 2.274
Chain 1: 300 -8783.343 1.425 1.000
Chain 1: 400 -8121.362 1.089 1.000
Chain 1: 500 -8237.075 0.874 1.000
Chain 1: 600 -8102.126 0.731 1.000
Chain 1: 700 -7750.359 0.633 0.082
Chain 1: 800 -8284.648 0.562 0.082
Chain 1: 900 -7985.269 0.504 0.064
Chain 1: 1000 -7836.132 0.455 0.064
Chain 1: 1100 -7660.433 0.358 0.045
Chain 1: 1200 -7675.161 0.130 0.037
Chain 1: 1300 -7802.967 0.032 0.023
Chain 1: 1400 -7797.079 0.024 0.019
Chain 1: 1500 -7491.512 0.027 0.023
Chain 1: 1600 -7691.432 0.028 0.026
Chain 1: 1700 -7479.997 0.026 0.026
Chain 1: 1800 -7578.280 0.021 0.023
Chain 1: 1900 -7640.724 0.018 0.019
Chain 1: 2000 -7509.018 0.018 0.018
Chain 1: 2100 -7533.257 0.016 0.016
Chain 1: 2200 -7666.656 0.017 0.017
Chain 1: 2300 -7531.304 0.017 0.018
Chain 1: 2400 -7579.520 0.018 0.018
Chain 1: 2500 -7519.489 0.015 0.017
Chain 1: 2600 -7478.056 0.013 0.013
Chain 1: 2700 -7476.949 0.010 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.006463 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 64.63 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86104.011 1.000 1.000
Chain 1: 200 -13657.719 3.152 5.304
Chain 1: 300 -9946.046 2.226 1.000
Chain 1: 400 -11216.138 1.698 1.000
Chain 1: 500 -8955.944 1.409 0.373
Chain 1: 600 -8328.463 1.186 0.373
Chain 1: 700 -8509.844 1.020 0.252
Chain 1: 800 -8632.196 0.894 0.252
Chain 1: 900 -8677.519 0.795 0.113
Chain 1: 1000 -8774.627 0.717 0.113
Chain 1: 1100 -8550.660 0.620 0.075 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8320.043 0.092 0.028
Chain 1: 1300 -8630.952 0.058 0.028
Chain 1: 1400 -8613.064 0.047 0.026
Chain 1: 1500 -8466.423 0.024 0.021
Chain 1: 1600 -8578.146 0.017 0.017
Chain 1: 1700 -8649.873 0.016 0.014
Chain 1: 1800 -8214.137 0.020 0.017
Chain 1: 1900 -8319.000 0.021 0.017
Chain 1: 2000 -8294.750 0.020 0.017
Chain 1: 2100 -8250.968 0.018 0.013
Chain 1: 2200 -8237.286 0.015 0.013
Chain 1: 2300 -8375.092 0.013 0.013
Chain 1: 2400 -8219.625 0.015 0.013
Chain 1: 2500 -8289.242 0.014 0.013
Chain 1: 2600 -8207.054 0.014 0.010
Chain 1: 2700 -8238.946 0.013 0.010
Chain 1: 2800 -8198.873 0.009 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003351 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.51 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8389381.000 1.000 1.000
Chain 1: 200 -1585824.869 2.645 4.290
Chain 1: 300 -891577.271 2.023 1.000
Chain 1: 400 -457578.337 1.754 1.000
Chain 1: 500 -357832.920 1.459 0.948
Chain 1: 600 -232784.405 1.306 0.948
Chain 1: 700 -119218.438 1.255 0.948
Chain 1: 800 -86445.470 1.146 0.948
Chain 1: 900 -66845.227 1.051 0.779
Chain 1: 1000 -51690.827 0.975 0.779
Chain 1: 1100 -39195.234 0.907 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38382.650 0.480 0.379
Chain 1: 1300 -26361.109 0.448 0.379
Chain 1: 1400 -26084.486 0.354 0.319
Chain 1: 1500 -22675.962 0.341 0.319
Chain 1: 1600 -21894.097 0.291 0.293
Chain 1: 1700 -20770.306 0.201 0.293
Chain 1: 1800 -20715.324 0.164 0.150
Chain 1: 1900 -21041.875 0.136 0.054
Chain 1: 2000 -19553.154 0.114 0.054
Chain 1: 2100 -19791.768 0.083 0.036
Chain 1: 2200 -20018.172 0.082 0.036
Chain 1: 2300 -19635.286 0.039 0.019
Chain 1: 2400 -19407.278 0.039 0.019
Chain 1: 2500 -19209.025 0.025 0.016
Chain 1: 2600 -18839.173 0.023 0.016
Chain 1: 2700 -18796.078 0.018 0.012
Chain 1: 2800 -18512.649 0.019 0.015
Chain 1: 2900 -18794.029 0.019 0.015
Chain 1: 3000 -18780.325 0.012 0.012
Chain 1: 3100 -18865.350 0.011 0.012
Chain 1: 3200 -18555.833 0.012 0.015
Chain 1: 3300 -18760.691 0.011 0.012
Chain 1: 3400 -18235.168 0.012 0.015
Chain 1: 3500 -18847.655 0.015 0.015
Chain 1: 3600 -18153.543 0.016 0.015
Chain 1: 3700 -18540.921 0.018 0.017
Chain 1: 3800 -17499.330 0.023 0.021
Chain 1: 3900 -17495.400 0.021 0.021
Chain 1: 4000 -17612.756 0.022 0.021
Chain 1: 4100 -17526.434 0.022 0.021
Chain 1: 4200 -17342.406 0.021 0.021
Chain 1: 4300 -17481.042 0.021 0.021
Chain 1: 4400 -17437.649 0.018 0.011
Chain 1: 4500 -17340.108 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001308 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.08 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12383.898 1.000 1.000
Chain 1: 200 -9334.449 0.663 1.000
Chain 1: 300 -8158.500 0.490 0.327
Chain 1: 400 -8255.381 0.371 0.327
Chain 1: 500 -8224.377 0.297 0.144
Chain 1: 600 -8041.764 0.252 0.144
Chain 1: 700 -7995.713 0.216 0.023
Chain 1: 800 -7995.642 0.189 0.023
Chain 1: 900 -7936.858 0.169 0.012
Chain 1: 1000 -8023.896 0.153 0.012
Chain 1: 1100 -8133.447 0.055 0.012
Chain 1: 1200 -8006.754 0.024 0.012
Chain 1: 1300 -7935.466 0.010 0.011
Chain 1: 1400 -7954.084 0.009 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002267 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 22.67 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56950.047 1.000 1.000
Chain 1: 200 -17512.207 1.626 2.252
Chain 1: 300 -8700.575 1.422 1.013
Chain 1: 400 -8275.202 1.079 1.013
Chain 1: 500 -8315.377 0.864 1.000
Chain 1: 600 -8745.228 0.728 1.000
Chain 1: 700 -7741.946 0.643 0.130
Chain 1: 800 -8200.575 0.569 0.130
Chain 1: 900 -7969.831 0.509 0.056
Chain 1: 1000 -7864.558 0.460 0.056
Chain 1: 1100 -7796.277 0.361 0.051
Chain 1: 1200 -7754.712 0.136 0.049
Chain 1: 1300 -7707.049 0.035 0.029
Chain 1: 1400 -7820.614 0.032 0.015
Chain 1: 1500 -7528.460 0.035 0.029
Chain 1: 1600 -7592.674 0.031 0.015
Chain 1: 1700 -7476.283 0.020 0.015
Chain 1: 1800 -7569.539 0.015 0.013
Chain 1: 1900 -7564.186 0.012 0.012
Chain 1: 2000 -7558.706 0.011 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003589 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.89 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86619.183 1.000 1.000
Chain 1: 200 -13540.719 3.198 5.397
Chain 1: 300 -9913.719 2.254 1.000
Chain 1: 400 -10794.950 1.711 1.000
Chain 1: 500 -8882.536 1.412 0.366
Chain 1: 600 -8466.753 1.185 0.366
Chain 1: 700 -8576.136 1.017 0.215
Chain 1: 800 -9173.988 0.898 0.215
Chain 1: 900 -8687.441 0.805 0.082
Chain 1: 1000 -8493.349 0.727 0.082
Chain 1: 1100 -8736.327 0.629 0.065 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8388.358 0.094 0.056
Chain 1: 1300 -8612.823 0.060 0.049
Chain 1: 1400 -8612.195 0.052 0.041
Chain 1: 1500 -8502.722 0.031 0.028
Chain 1: 1600 -8607.333 0.028 0.026
Chain 1: 1700 -8695.314 0.027 0.026
Chain 1: 1800 -8287.368 0.026 0.026
Chain 1: 1900 -8383.974 0.021 0.023
Chain 1: 2000 -8356.219 0.019 0.013
Chain 1: 2100 -8476.944 0.018 0.013
Chain 1: 2200 -8293.893 0.016 0.013
Chain 1: 2300 -8423.657 0.015 0.013
Chain 1: 2400 -8433.678 0.015 0.013
Chain 1: 2500 -8395.970 0.014 0.012
Chain 1: 2600 -8394.703 0.013 0.012
Chain 1: 2700 -8309.544 0.013 0.012
Chain 1: 2800 -8274.414 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003174 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.74 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8396749.102 1.000 1.000
Chain 1: 200 -1585967.923 2.647 4.294
Chain 1: 300 -892393.465 2.024 1.000
Chain 1: 400 -458374.296 1.755 1.000
Chain 1: 500 -358720.324 1.459 0.947
Chain 1: 600 -233546.408 1.305 0.947
Chain 1: 700 -119522.977 1.255 0.947
Chain 1: 800 -86646.785 1.146 0.947
Chain 1: 900 -66948.867 1.051 0.777
Chain 1: 1000 -51705.208 0.975 0.777
Chain 1: 1100 -39144.744 0.908 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38315.554 0.480 0.379
Chain 1: 1300 -26239.573 0.449 0.379
Chain 1: 1400 -25954.655 0.355 0.321
Chain 1: 1500 -22533.289 0.342 0.321
Chain 1: 1600 -21746.671 0.292 0.295
Chain 1: 1700 -20617.051 0.202 0.294
Chain 1: 1800 -20560.203 0.165 0.152
Chain 1: 1900 -20886.172 0.137 0.055
Chain 1: 2000 -19395.667 0.115 0.055
Chain 1: 2100 -19634.191 0.084 0.036
Chain 1: 2200 -19860.798 0.083 0.036
Chain 1: 2300 -19477.900 0.039 0.020
Chain 1: 2400 -19250.013 0.039 0.020
Chain 1: 2500 -19052.102 0.025 0.016
Chain 1: 2600 -18682.424 0.024 0.016
Chain 1: 2700 -18639.399 0.018 0.012
Chain 1: 2800 -18356.354 0.020 0.015
Chain 1: 2900 -18637.588 0.019 0.015
Chain 1: 3000 -18623.783 0.012 0.012
Chain 1: 3100 -18708.738 0.011 0.012
Chain 1: 3200 -18399.523 0.012 0.015
Chain 1: 3300 -18604.157 0.011 0.012
Chain 1: 3400 -18079.273 0.013 0.015
Chain 1: 3500 -18690.906 0.015 0.015
Chain 1: 3600 -17997.951 0.017 0.015
Chain 1: 3700 -18384.498 0.018 0.017
Chain 1: 3800 -17344.768 0.023 0.021
Chain 1: 3900 -17340.936 0.021 0.021
Chain 1: 4000 -17458.225 0.022 0.021
Chain 1: 4100 -17372.019 0.022 0.021
Chain 1: 4200 -17188.375 0.021 0.021
Chain 1: 4300 -17326.682 0.021 0.021
Chain 1: 4400 -17283.620 0.019 0.011
Chain 1: 4500 -17186.174 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001122 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 11.22 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49166.557 1.000 1.000
Chain 1: 200 -19052.586 1.290 1.581
Chain 1: 300 -16327.065 0.916 1.000
Chain 1: 400 -23755.488 0.765 1.000
Chain 1: 500 -13722.578 0.758 0.731
Chain 1: 600 -12297.064 0.651 0.731
Chain 1: 700 -16256.964 0.593 0.313
Chain 1: 800 -13010.225 0.550 0.313
Chain 1: 900 -15252.118 0.505 0.250
Chain 1: 1000 -11572.561 0.487 0.313
Chain 1: 1100 -13790.422 0.403 0.250
Chain 1: 1200 -10521.504 0.276 0.250
Chain 1: 1300 -12917.569 0.277 0.250
Chain 1: 1400 -19557.186 0.280 0.250
Chain 1: 1500 -10392.679 0.295 0.250
Chain 1: 1600 -10774.665 0.287 0.250
Chain 1: 1700 -11551.382 0.270 0.250
Chain 1: 1800 -11165.350 0.248 0.185
Chain 1: 1900 -10603.963 0.239 0.185
Chain 1: 2000 -9859.586 0.214 0.161
Chain 1: 2100 -9336.339 0.204 0.075
Chain 1: 2200 -12582.021 0.199 0.075
Chain 1: 2300 -9486.767 0.213 0.075
Chain 1: 2400 -8940.266 0.185 0.067
Chain 1: 2500 -9630.076 0.104 0.067
Chain 1: 2600 -9467.438 0.102 0.067
Chain 1: 2700 -11911.289 0.116 0.072
Chain 1: 2800 -8990.134 0.145 0.075
Chain 1: 2900 -18326.362 0.191 0.205
Chain 1: 3000 -9287.227 0.280 0.258
Chain 1: 3100 -9891.761 0.281 0.258
Chain 1: 3200 -15465.160 0.291 0.325
Chain 1: 3300 -9879.184 0.315 0.325
Chain 1: 3400 -8968.765 0.319 0.325
Chain 1: 3500 -9921.654 0.321 0.325
Chain 1: 3600 -14170.914 0.350 0.325
Chain 1: 3700 -9056.477 0.386 0.360
Chain 1: 3800 -10966.202 0.371 0.360
Chain 1: 3900 -13510.599 0.338 0.300
Chain 1: 4000 -9182.960 0.288 0.300
Chain 1: 4100 -8846.563 0.286 0.300
Chain 1: 4200 -8876.930 0.250 0.188
Chain 1: 4300 -9740.424 0.203 0.174
Chain 1: 4400 -9461.795 0.195 0.174
Chain 1: 4500 -10170.129 0.193 0.174
Chain 1: 4600 -8772.836 0.179 0.159
Chain 1: 4700 -12695.623 0.153 0.159
Chain 1: 4800 -12323.566 0.139 0.089
Chain 1: 4900 -9167.791 0.154 0.089
Chain 1: 5000 -18453.528 0.158 0.089
Chain 1: 5100 -8888.239 0.261 0.159
Chain 1: 5200 -9880.665 0.271 0.159
Chain 1: 5300 -12594.225 0.284 0.215
Chain 1: 5400 -8727.371 0.325 0.309
Chain 1: 5500 -11716.929 0.344 0.309
Chain 1: 5600 -13100.202 0.338 0.309
Chain 1: 5700 -8963.269 0.354 0.344
Chain 1: 5800 -10530.101 0.365 0.344
Chain 1: 5900 -13417.193 0.352 0.255
Chain 1: 6000 -8668.090 0.357 0.255
Chain 1: 6100 -9288.155 0.256 0.215
Chain 1: 6200 -8985.905 0.249 0.215
Chain 1: 6300 -10644.421 0.243 0.215
Chain 1: 6400 -10590.982 0.200 0.156
Chain 1: 6500 -8775.462 0.195 0.156
Chain 1: 6600 -8515.798 0.187 0.156
Chain 1: 6700 -10244.864 0.158 0.156
Chain 1: 6800 -8735.688 0.160 0.169
Chain 1: 6900 -12123.665 0.167 0.169
Chain 1: 7000 -9865.395 0.135 0.169
Chain 1: 7100 -8214.200 0.148 0.173
Chain 1: 7200 -8590.831 0.149 0.173
Chain 1: 7300 -9486.264 0.143 0.173
Chain 1: 7400 -8918.319 0.149 0.173
Chain 1: 7500 -8338.715 0.135 0.169
Chain 1: 7600 -8794.662 0.137 0.169
Chain 1: 7700 -8592.474 0.123 0.094
Chain 1: 7800 -8486.262 0.107 0.070
Chain 1: 7900 -8594.866 0.080 0.064
Chain 1: 8000 -10291.647 0.074 0.064
Chain 1: 8100 -9052.112 0.067 0.064
Chain 1: 8200 -8114.771 0.075 0.070
Chain 1: 8300 -10251.748 0.086 0.070
Chain 1: 8400 -12386.342 0.097 0.116
Chain 1: 8500 -10137.854 0.112 0.137
Chain 1: 8600 -14619.663 0.138 0.165
Chain 1: 8700 -9090.410 0.196 0.172
Chain 1: 8800 -8513.316 0.202 0.172
Chain 1: 8900 -10745.605 0.221 0.208
Chain 1: 9000 -10974.395 0.207 0.208
Chain 1: 9100 -8666.177 0.220 0.208
Chain 1: 9200 -8773.995 0.209 0.208
Chain 1: 9300 -8605.793 0.190 0.208
Chain 1: 9400 -10303.222 0.190 0.208
Chain 1: 9500 -8212.782 0.193 0.208
Chain 1: 9600 -8383.351 0.164 0.165
Chain 1: 9700 -11664.847 0.132 0.165
Chain 1: 9800 -8360.912 0.164 0.208
Chain 1: 9900 -9866.451 0.159 0.165
Chain 1: 10000 -8469.738 0.173 0.165
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001389 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.89 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -62994.751 1.000 1.000
Chain 1: 200 -18103.606 1.740 2.480
Chain 1: 300 -8788.505 1.513 1.060
Chain 1: 400 -8608.885 1.140 1.060
Chain 1: 500 -7974.700 0.928 1.000
Chain 1: 600 -8821.832 0.789 1.000
Chain 1: 700 -7917.223 0.693 0.114
Chain 1: 800 -7863.217 0.607 0.114
Chain 1: 900 -7998.103 0.542 0.096
Chain 1: 1000 -7754.008 0.491 0.096
Chain 1: 1100 -7765.810 0.391 0.080
Chain 1: 1200 -7607.153 0.145 0.031
Chain 1: 1300 -7636.027 0.039 0.021
Chain 1: 1400 -7905.289 0.041 0.031
Chain 1: 1500 -7666.644 0.036 0.031
Chain 1: 1600 -7835.639 0.028 0.022
Chain 1: 1700 -7558.169 0.020 0.022
Chain 1: 1800 -7635.765 0.021 0.022
Chain 1: 1900 -7623.912 0.019 0.022
Chain 1: 2000 -7654.494 0.017 0.021
Chain 1: 2100 -7633.425 0.017 0.021
Chain 1: 2200 -7747.063 0.016 0.015
Chain 1: 2300 -7656.097 0.017 0.015
Chain 1: 2400 -7691.449 0.014 0.012
Chain 1: 2500 -7638.393 0.011 0.010
Chain 1: 2600 -7586.035 0.010 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00289 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 28.9 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86304.934 1.000 1.000
Chain 1: 200 -13494.760 3.198 5.395
Chain 1: 300 -9899.059 2.253 1.000
Chain 1: 400 -10677.273 1.708 1.000
Chain 1: 500 -8866.093 1.407 0.363
Chain 1: 600 -8396.554 1.182 0.363
Chain 1: 700 -8475.290 1.014 0.204
Chain 1: 800 -9202.687 0.898 0.204
Chain 1: 900 -8699.739 0.804 0.079
Chain 1: 1000 -8522.189 0.726 0.079
Chain 1: 1100 -8796.411 0.629 0.073 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8450.098 0.094 0.058
Chain 1: 1300 -8622.887 0.059 0.056
Chain 1: 1400 -8628.457 0.052 0.041
Chain 1: 1500 -8488.764 0.033 0.031
Chain 1: 1600 -8599.843 0.029 0.021
Chain 1: 1700 -8686.402 0.029 0.021
Chain 1: 1800 -8285.070 0.026 0.021
Chain 1: 1900 -8383.628 0.021 0.020
Chain 1: 2000 -8355.095 0.020 0.016
Chain 1: 2100 -8474.881 0.018 0.014
Chain 1: 2200 -8265.542 0.016 0.014
Chain 1: 2300 -8416.455 0.016 0.014
Chain 1: 2400 -8295.099 0.017 0.015
Chain 1: 2500 -8359.012 0.017 0.014
Chain 1: 2600 -8381.547 0.016 0.014
Chain 1: 2700 -8300.299 0.016 0.014
Chain 1: 2800 -8273.649 0.011 0.012
Chain 1: 2900 -8329.076 0.011 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003363 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.63 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8414262.358 1.000 1.000
Chain 1: 200 -1583582.632 2.657 4.313
Chain 1: 300 -889939.484 2.031 1.000
Chain 1: 400 -457238.761 1.760 1.000
Chain 1: 500 -357547.450 1.464 0.946
Chain 1: 600 -232491.779 1.309 0.946
Chain 1: 700 -118958.671 1.259 0.946
Chain 1: 800 -86255.039 1.149 0.946
Chain 1: 900 -66636.239 1.054 0.779
Chain 1: 1000 -51466.636 0.978 0.779
Chain 1: 1100 -38981.336 0.910 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38159.456 0.481 0.379
Chain 1: 1300 -26151.903 0.449 0.379
Chain 1: 1400 -25873.795 0.355 0.320
Chain 1: 1500 -22471.257 0.342 0.320
Chain 1: 1600 -21691.027 0.292 0.295
Chain 1: 1700 -20568.902 0.202 0.294
Chain 1: 1800 -20514.142 0.165 0.151
Chain 1: 1900 -20840.046 0.137 0.055
Chain 1: 2000 -19354.167 0.115 0.055
Chain 1: 2100 -19592.173 0.084 0.036
Chain 1: 2200 -19818.257 0.083 0.036
Chain 1: 2300 -19435.889 0.039 0.020
Chain 1: 2400 -19208.131 0.039 0.020
Chain 1: 2500 -19010.137 0.025 0.016
Chain 1: 2600 -18640.563 0.023 0.016
Chain 1: 2700 -18597.661 0.018 0.012
Chain 1: 2800 -18314.647 0.020 0.015
Chain 1: 2900 -18595.698 0.019 0.015
Chain 1: 3000 -18581.881 0.012 0.012
Chain 1: 3100 -18666.854 0.011 0.012
Chain 1: 3200 -18357.705 0.012 0.015
Chain 1: 3300 -18562.315 0.011 0.012
Chain 1: 3400 -18037.561 0.013 0.015
Chain 1: 3500 -18648.923 0.015 0.015
Chain 1: 3600 -17956.239 0.017 0.015
Chain 1: 3700 -18342.550 0.019 0.017
Chain 1: 3800 -17303.272 0.023 0.021
Chain 1: 3900 -17299.451 0.021 0.021
Chain 1: 4000 -17416.740 0.022 0.021
Chain 1: 4100 -17330.567 0.022 0.021
Chain 1: 4200 -17147.052 0.022 0.021
Chain 1: 4300 -17285.276 0.021 0.021
Chain 1: 4400 -17242.265 0.019 0.011
Chain 1: 4500 -17144.851 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001435 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.35 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48412.098 1.000 1.000
Chain 1: 200 -15548.831 1.557 2.114
Chain 1: 300 -21143.201 1.126 1.000
Chain 1: 400 -12798.028 1.008 1.000
Chain 1: 500 -15105.147 0.837 0.652
Chain 1: 600 -13699.119 0.714 0.652
Chain 1: 700 -12745.912 0.623 0.265
Chain 1: 800 -10669.963 0.569 0.265
Chain 1: 900 -10535.977 0.508 0.195
Chain 1: 1000 -13012.724 0.476 0.195
Chain 1: 1100 -24307.931 0.422 0.195
Chain 1: 1200 -11260.337 0.327 0.195
Chain 1: 1300 -10685.399 0.306 0.190
Chain 1: 1400 -9737.006 0.250 0.153
Chain 1: 1500 -10295.144 0.240 0.103
Chain 1: 1600 -9848.837 0.235 0.097
Chain 1: 1700 -9680.316 0.229 0.097
Chain 1: 1800 -14586.842 0.243 0.097
Chain 1: 1900 -10510.893 0.281 0.190
Chain 1: 2000 -13749.419 0.285 0.236
Chain 1: 2100 -11144.403 0.262 0.234
Chain 1: 2200 -10113.196 0.156 0.102
Chain 1: 2300 -11553.995 0.163 0.125
Chain 1: 2400 -8964.284 0.183 0.234
Chain 1: 2500 -13952.568 0.213 0.236
Chain 1: 2600 -13296.765 0.213 0.236
Chain 1: 2700 -9265.543 0.255 0.289
Chain 1: 2800 -8658.948 0.228 0.236
Chain 1: 2900 -9007.579 0.194 0.234
Chain 1: 3000 -15186.482 0.211 0.234
Chain 1: 3100 -9105.893 0.254 0.289
Chain 1: 3200 -10126.997 0.254 0.289
Chain 1: 3300 -15494.233 0.276 0.346
Chain 1: 3400 -11985.898 0.277 0.346
Chain 1: 3500 -9108.117 0.272 0.316
Chain 1: 3600 -9584.696 0.272 0.316
Chain 1: 3700 -16815.984 0.272 0.316
Chain 1: 3800 -8836.957 0.355 0.346
Chain 1: 3900 -11833.204 0.377 0.346
Chain 1: 4000 -8765.315 0.371 0.346
Chain 1: 4100 -8517.299 0.307 0.316
Chain 1: 4200 -9007.725 0.302 0.316
Chain 1: 4300 -8436.274 0.275 0.293
Chain 1: 4400 -8269.944 0.247 0.253
Chain 1: 4500 -8942.272 0.223 0.075
Chain 1: 4600 -8240.221 0.227 0.085
Chain 1: 4700 -9833.604 0.200 0.085
Chain 1: 4800 -8597.795 0.124 0.085
Chain 1: 4900 -9326.784 0.107 0.078
Chain 1: 5000 -12892.407 0.099 0.078
Chain 1: 5100 -8323.562 0.151 0.085
Chain 1: 5200 -9147.855 0.155 0.090
Chain 1: 5300 -11714.641 0.170 0.144
Chain 1: 5400 -12892.588 0.177 0.144
Chain 1: 5500 -8489.872 0.221 0.162
Chain 1: 5600 -8221.936 0.216 0.162
Chain 1: 5700 -12266.323 0.233 0.219
Chain 1: 5800 -8289.994 0.266 0.277
Chain 1: 5900 -8728.701 0.264 0.277
Chain 1: 6000 -9434.626 0.244 0.219
Chain 1: 6100 -9717.380 0.192 0.091
Chain 1: 6200 -8430.772 0.198 0.153
Chain 1: 6300 -13100.151 0.212 0.153
Chain 1: 6400 -12983.570 0.203 0.153
Chain 1: 6500 -9378.345 0.190 0.153
Chain 1: 6600 -10993.067 0.201 0.153
Chain 1: 6700 -10005.616 0.178 0.147
Chain 1: 6800 -9697.630 0.133 0.099
Chain 1: 6900 -12105.205 0.148 0.147
Chain 1: 7000 -8366.224 0.185 0.153
Chain 1: 7100 -8633.937 0.186 0.153
Chain 1: 7200 -8209.007 0.176 0.147
Chain 1: 7300 -9981.457 0.158 0.147
Chain 1: 7400 -7915.311 0.183 0.178
Chain 1: 7500 -9790.710 0.164 0.178
Chain 1: 7600 -8551.687 0.163 0.178
Chain 1: 7700 -8941.007 0.158 0.178
Chain 1: 7800 -8623.246 0.158 0.178
Chain 1: 7900 -8998.206 0.143 0.145
Chain 1: 8000 -10891.733 0.115 0.145
Chain 1: 8100 -11136.191 0.114 0.145
Chain 1: 8200 -10331.753 0.117 0.145
Chain 1: 8300 -7914.707 0.130 0.145
Chain 1: 8400 -10098.914 0.125 0.145
Chain 1: 8500 -7951.994 0.133 0.145
Chain 1: 8600 -9737.301 0.137 0.174
Chain 1: 8700 -8002.661 0.154 0.183
Chain 1: 8800 -8124.438 0.152 0.183
Chain 1: 8900 -10856.879 0.173 0.216
Chain 1: 9000 -11164.121 0.159 0.216
Chain 1: 9100 -8136.579 0.194 0.217
Chain 1: 9200 -11542.557 0.215 0.252
Chain 1: 9300 -8281.580 0.224 0.252
Chain 1: 9400 -11460.856 0.230 0.270
Chain 1: 9500 -7937.358 0.248 0.277
Chain 1: 9600 -9530.804 0.246 0.277
Chain 1: 9700 -7887.937 0.245 0.277
Chain 1: 9800 -8179.235 0.247 0.277
Chain 1: 9900 -9018.166 0.231 0.277
Chain 1: 10000 -7856.386 0.243 0.277
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001518 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.18 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61348.740 1.000 1.000
Chain 1: 200 -17331.604 1.770 2.540
Chain 1: 300 -8555.626 1.522 1.026
Chain 1: 400 -8068.497 1.156 1.026
Chain 1: 500 -8202.757 0.928 1.000
Chain 1: 600 -8010.361 0.778 1.000
Chain 1: 700 -8008.089 0.667 0.060
Chain 1: 800 -7889.085 0.585 0.060
Chain 1: 900 -7709.759 0.523 0.024
Chain 1: 1000 -7639.639 0.471 0.024
Chain 1: 1100 -7544.589 0.373 0.023
Chain 1: 1200 -7591.091 0.119 0.016
Chain 1: 1300 -7483.290 0.018 0.015
Chain 1: 1400 -7767.413 0.016 0.015
Chain 1: 1500 -7473.541 0.018 0.015
Chain 1: 1600 -7379.026 0.017 0.014
Chain 1: 1700 -7386.799 0.017 0.014
Chain 1: 1800 -7428.939 0.016 0.013
Chain 1: 1900 -7397.313 0.014 0.013
Chain 1: 2000 -7461.197 0.014 0.013
Chain 1: 2100 -7503.270 0.013 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002913 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.13 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85435.385 1.000 1.000
Chain 1: 200 -12994.297 3.287 5.575
Chain 1: 300 -9509.280 2.314 1.000
Chain 1: 400 -10240.064 1.753 1.000
Chain 1: 500 -8365.887 1.447 0.366
Chain 1: 600 -8502.940 1.209 0.366
Chain 1: 700 -8179.049 1.042 0.224
Chain 1: 800 -8553.743 0.917 0.224
Chain 1: 900 -8419.394 0.817 0.071
Chain 1: 1000 -8157.998 0.738 0.071
Chain 1: 1100 -8439.604 0.642 0.044 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8196.244 0.087 0.040
Chain 1: 1300 -8156.475 0.051 0.033
Chain 1: 1400 -8160.825 0.044 0.032
Chain 1: 1500 -8178.432 0.022 0.030
Chain 1: 1600 -8177.687 0.020 0.030
Chain 1: 1700 -8124.356 0.017 0.016
Chain 1: 1800 -8002.127 0.014 0.015
Chain 1: 1900 -8113.542 0.014 0.014
Chain 1: 2000 -8077.526 0.011 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003514 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.14 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8395271.396 1.000 1.000
Chain 1: 200 -1583203.283 2.651 4.303
Chain 1: 300 -890226.533 2.027 1.000
Chain 1: 400 -456782.048 1.758 1.000
Chain 1: 500 -357244.532 1.462 0.949
Chain 1: 600 -232343.030 1.308 0.949
Chain 1: 700 -118644.952 1.258 0.949
Chain 1: 800 -85858.606 1.148 0.949
Chain 1: 900 -66208.051 1.054 0.778
Chain 1: 1000 -50998.543 0.978 0.778
Chain 1: 1100 -38477.545 0.911 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37646.309 0.483 0.382
Chain 1: 1300 -25622.568 0.452 0.382
Chain 1: 1400 -25338.507 0.358 0.325
Chain 1: 1500 -21931.705 0.346 0.325
Chain 1: 1600 -21148.505 0.296 0.298
Chain 1: 1700 -20025.810 0.205 0.297
Chain 1: 1800 -19970.189 0.167 0.155
Chain 1: 1900 -20295.384 0.139 0.056
Chain 1: 2000 -18810.138 0.117 0.056
Chain 1: 2100 -19048.199 0.086 0.037
Chain 1: 2200 -19273.760 0.085 0.037
Chain 1: 2300 -18892.032 0.040 0.020
Chain 1: 2400 -18664.521 0.040 0.020
Chain 1: 2500 -18466.456 0.026 0.016
Chain 1: 2600 -18097.708 0.024 0.016
Chain 1: 2700 -18054.988 0.019 0.012
Chain 1: 2800 -17772.236 0.020 0.016
Chain 1: 2900 -18053.047 0.020 0.016
Chain 1: 3000 -18039.295 0.012 0.012
Chain 1: 3100 -18124.127 0.011 0.012
Chain 1: 3200 -17815.474 0.012 0.016
Chain 1: 3300 -18019.684 0.011 0.012
Chain 1: 3400 -17495.754 0.013 0.016
Chain 1: 3500 -18105.869 0.015 0.016
Chain 1: 3600 -17414.890 0.017 0.016
Chain 1: 3700 -17799.945 0.019 0.017
Chain 1: 3800 -16763.236 0.024 0.022
Chain 1: 3900 -16759.481 0.022 0.022
Chain 1: 4000 -16876.777 0.023 0.022
Chain 1: 4100 -16790.703 0.023 0.022
Chain 1: 4200 -16607.755 0.022 0.022
Chain 1: 4300 -16745.588 0.022 0.022
Chain 1: 4400 -16703.060 0.019 0.011
Chain 1: 4500 -16605.720 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001232 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.32 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48670.975 1.000 1.000
Chain 1: 200 -20135.743 1.209 1.417
Chain 1: 300 -14610.227 0.932 1.000
Chain 1: 400 -14901.987 0.704 1.000
Chain 1: 500 -12607.729 0.599 0.378
Chain 1: 600 -16839.033 0.541 0.378
Chain 1: 700 -17795.727 0.472 0.251
Chain 1: 800 -12723.818 0.463 0.378
Chain 1: 900 -12646.663 0.412 0.251
Chain 1: 1000 -11006.105 0.386 0.251
Chain 1: 1100 -10922.072 0.286 0.182
Chain 1: 1200 -11107.241 0.146 0.149
Chain 1: 1300 -11510.047 0.112 0.054
Chain 1: 1400 -9737.929 0.128 0.149
Chain 1: 1500 -11765.610 0.127 0.149
Chain 1: 1600 -9857.562 0.121 0.149
Chain 1: 1700 -9880.018 0.116 0.149
Chain 1: 1800 -12299.813 0.096 0.149
Chain 1: 1900 -14844.186 0.113 0.171
Chain 1: 2000 -11990.581 0.122 0.172
Chain 1: 2100 -10647.529 0.133 0.172
Chain 1: 2200 -12093.032 0.144 0.172
Chain 1: 2300 -13564.117 0.151 0.172
Chain 1: 2400 -11432.552 0.151 0.172
Chain 1: 2500 -10023.591 0.148 0.171
Chain 1: 2600 -9435.081 0.135 0.141
Chain 1: 2700 -10830.269 0.148 0.141
Chain 1: 2800 -11044.784 0.130 0.129
Chain 1: 2900 -9674.834 0.127 0.129
Chain 1: 3000 -8967.296 0.111 0.126
Chain 1: 3100 -9242.958 0.102 0.120
Chain 1: 3200 -9050.873 0.092 0.108
Chain 1: 3300 -9714.897 0.088 0.079
Chain 1: 3400 -9173.156 0.075 0.068
Chain 1: 3500 -9332.438 0.063 0.062
Chain 1: 3600 -9135.380 0.059 0.059
Chain 1: 3700 -10608.213 0.060 0.059
Chain 1: 3800 -16879.100 0.095 0.068
Chain 1: 3900 -10016.611 0.149 0.068
Chain 1: 4000 -11555.318 0.155 0.068
Chain 1: 4100 -9114.938 0.178 0.133
Chain 1: 4200 -12456.360 0.203 0.139
Chain 1: 4300 -10165.713 0.219 0.225
Chain 1: 4400 -14499.614 0.243 0.268
Chain 1: 4500 -8533.006 0.311 0.268
Chain 1: 4600 -11925.693 0.337 0.284
Chain 1: 4700 -8941.602 0.357 0.299
Chain 1: 4800 -8782.899 0.321 0.284
Chain 1: 4900 -12633.012 0.283 0.284
Chain 1: 5000 -14638.731 0.284 0.284
Chain 1: 5100 -10174.112 0.301 0.299
Chain 1: 5200 -8729.570 0.291 0.299
Chain 1: 5300 -12512.135 0.298 0.302
Chain 1: 5400 -16974.083 0.295 0.302
Chain 1: 5500 -12543.295 0.260 0.302
Chain 1: 5600 -9378.703 0.265 0.305
Chain 1: 5700 -12951.397 0.260 0.302
Chain 1: 5800 -13815.573 0.264 0.302
Chain 1: 5900 -9179.056 0.284 0.302
Chain 1: 6000 -11611.180 0.291 0.302
Chain 1: 6100 -9318.373 0.272 0.276
Chain 1: 6200 -8947.446 0.260 0.276
Chain 1: 6300 -8883.942 0.230 0.263
Chain 1: 6400 -11929.820 0.229 0.255
Chain 1: 6500 -8544.297 0.234 0.255
Chain 1: 6600 -9315.184 0.208 0.246
Chain 1: 6700 -8866.387 0.186 0.209
Chain 1: 6800 -11360.719 0.201 0.220
Chain 1: 6900 -9129.822 0.175 0.220
Chain 1: 7000 -12390.133 0.181 0.244
Chain 1: 7100 -10873.693 0.170 0.220
Chain 1: 7200 -8293.275 0.197 0.244
Chain 1: 7300 -9437.369 0.208 0.244
Chain 1: 7400 -8554.903 0.193 0.220
Chain 1: 7500 -8270.683 0.157 0.139
Chain 1: 7600 -8532.600 0.152 0.139
Chain 1: 7700 -8645.469 0.148 0.139
Chain 1: 7800 -8946.546 0.129 0.121
Chain 1: 7900 -9267.186 0.108 0.103
Chain 1: 8000 -10012.921 0.090 0.074
Chain 1: 8100 -8562.966 0.093 0.074
Chain 1: 8200 -9137.194 0.068 0.063
Chain 1: 8300 -8863.958 0.059 0.035
Chain 1: 8400 -11654.751 0.072 0.035
Chain 1: 8500 -8310.292 0.109 0.063
Chain 1: 8600 -9355.821 0.117 0.074
Chain 1: 8700 -8375.294 0.128 0.112
Chain 1: 8800 -8407.219 0.125 0.112
Chain 1: 8900 -10375.158 0.140 0.117
Chain 1: 9000 -9779.620 0.139 0.117
Chain 1: 9100 -9902.057 0.123 0.112
Chain 1: 9200 -8417.592 0.134 0.117
Chain 1: 9300 -11835.925 0.160 0.176
Chain 1: 9400 -9164.878 0.165 0.176
Chain 1: 9500 -11847.373 0.148 0.176
Chain 1: 9600 -9564.379 0.161 0.190
Chain 1: 9700 -11468.791 0.165 0.190
Chain 1: 9800 -8430.412 0.201 0.226
Chain 1: 9900 -9105.803 0.190 0.226
Chain 1: 10000 -8997.945 0.185 0.226
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001397 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.97 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57253.794 1.000 1.000
Chain 1: 200 -17516.815 1.634 2.269
Chain 1: 300 -8820.210 1.418 1.000
Chain 1: 400 -8484.048 1.074 1.000
Chain 1: 500 -8230.510 0.865 0.986
Chain 1: 600 -8623.653 0.728 0.986
Chain 1: 700 -8104.642 0.634 0.064
Chain 1: 800 -8253.193 0.557 0.064
Chain 1: 900 -8086.859 0.497 0.046
Chain 1: 1000 -8349.544 0.450 0.046
Chain 1: 1100 -7893.048 0.356 0.046
Chain 1: 1200 -7818.879 0.130 0.040
Chain 1: 1300 -7840.359 0.032 0.031
Chain 1: 1400 -8013.101 0.030 0.031
Chain 1: 1500 -7726.713 0.031 0.031
Chain 1: 1600 -7929.227 0.029 0.026
Chain 1: 1700 -7640.383 0.026 0.026
Chain 1: 1800 -7717.030 0.025 0.026
Chain 1: 1900 -7640.803 0.024 0.026
Chain 1: 2000 -7736.865 0.022 0.022
Chain 1: 2100 -7751.706 0.017 0.012
Chain 1: 2200 -7829.928 0.017 0.012
Chain 1: 2300 -7719.492 0.018 0.014
Chain 1: 2400 -7774.641 0.017 0.012
Chain 1: 2500 -7653.611 0.014 0.012
Chain 1: 2600 -7673.370 0.012 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003285 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.85 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86065.728 1.000 1.000
Chain 1: 200 -13509.100 3.185 5.371
Chain 1: 300 -9924.391 2.244 1.000
Chain 1: 400 -10719.797 1.702 1.000
Chain 1: 500 -8890.326 1.402 0.361
Chain 1: 600 -8593.903 1.174 0.361
Chain 1: 700 -8481.039 1.009 0.206
Chain 1: 800 -9275.942 0.893 0.206
Chain 1: 900 -8723.636 0.801 0.086
Chain 1: 1000 -8501.710 0.724 0.086
Chain 1: 1100 -8737.268 0.626 0.074 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8414.067 0.093 0.063
Chain 1: 1300 -8628.287 0.059 0.038
Chain 1: 1400 -8626.109 0.052 0.034
Chain 1: 1500 -8520.045 0.033 0.027
Chain 1: 1600 -8623.665 0.030 0.026
Chain 1: 1700 -8712.304 0.030 0.026
Chain 1: 1800 -8309.059 0.026 0.026
Chain 1: 1900 -8408.249 0.021 0.025
Chain 1: 2000 -8379.608 0.019 0.012
Chain 1: 2100 -8499.410 0.018 0.012
Chain 1: 2200 -8290.034 0.016 0.012
Chain 1: 2300 -8440.973 0.016 0.012
Chain 1: 2400 -8319.648 0.017 0.014
Chain 1: 2500 -8383.548 0.017 0.014
Chain 1: 2600 -8406.071 0.016 0.014
Chain 1: 2700 -8324.813 0.016 0.014
Chain 1: 2800 -8298.177 0.011 0.012
Chain 1: 2900 -8353.583 0.011 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00518 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 51.8 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8407723.101 1.000 1.000
Chain 1: 200 -1586992.629 2.649 4.298
Chain 1: 300 -891987.327 2.026 1.000
Chain 1: 400 -457958.510 1.756 1.000
Chain 1: 500 -357944.520 1.461 0.948
Chain 1: 600 -232795.103 1.307 0.948
Chain 1: 700 -119120.508 1.257 0.948
Chain 1: 800 -86336.377 1.147 0.948
Chain 1: 900 -66708.813 1.052 0.779
Chain 1: 1000 -51527.999 0.976 0.779
Chain 1: 1100 -39024.360 0.909 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38202.930 0.481 0.380
Chain 1: 1300 -26184.160 0.449 0.380
Chain 1: 1400 -25904.642 0.355 0.320
Chain 1: 1500 -22498.067 0.342 0.320
Chain 1: 1600 -21716.300 0.292 0.295
Chain 1: 1700 -20593.273 0.202 0.294
Chain 1: 1800 -20538.066 0.165 0.151
Chain 1: 1900 -20863.972 0.137 0.055
Chain 1: 2000 -19377.273 0.115 0.055
Chain 1: 2100 -19615.525 0.084 0.036
Chain 1: 2200 -19841.517 0.083 0.036
Chain 1: 2300 -19459.233 0.039 0.020
Chain 1: 2400 -19231.443 0.039 0.020
Chain 1: 2500 -19033.340 0.025 0.016
Chain 1: 2600 -18663.888 0.023 0.016
Chain 1: 2700 -18621.021 0.018 0.012
Chain 1: 2800 -18337.888 0.020 0.015
Chain 1: 2900 -18619.034 0.019 0.015
Chain 1: 3000 -18605.289 0.012 0.012
Chain 1: 3100 -18690.193 0.011 0.012
Chain 1: 3200 -18381.096 0.012 0.015
Chain 1: 3300 -18585.691 0.011 0.012
Chain 1: 3400 -18060.885 0.013 0.015
Chain 1: 3500 -18672.261 0.015 0.015
Chain 1: 3600 -17979.686 0.017 0.015
Chain 1: 3700 -18365.867 0.018 0.017
Chain 1: 3800 -17326.652 0.023 0.021
Chain 1: 3900 -17322.835 0.021 0.021
Chain 1: 4000 -17440.160 0.022 0.021
Chain 1: 4100 -17353.902 0.022 0.021
Chain 1: 4200 -17170.450 0.021 0.021
Chain 1: 4300 -17308.658 0.021 0.021
Chain 1: 4400 -17265.678 0.019 0.011
Chain 1: 4500 -17168.244 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001424 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.24 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12319.013 1.000 1.000
Chain 1: 200 -9238.692 0.667 1.000
Chain 1: 300 -8048.416 0.494 0.333
Chain 1: 400 -8215.664 0.375 0.333
Chain 1: 500 -8082.909 0.304 0.148
Chain 1: 600 -8006.227 0.255 0.148
Chain 1: 700 -7918.430 0.220 0.020
Chain 1: 800 -7960.834 0.193 0.020
Chain 1: 900 -8080.876 0.173 0.016
Chain 1: 1000 -7979.211 0.157 0.016
Chain 1: 1100 -8028.832 0.058 0.015
Chain 1: 1200 -7927.512 0.026 0.013
Chain 1: 1300 -7891.409 0.011 0.013
Chain 1: 1400 -7908.341 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001795 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 17.95 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61857.114 1.000 1.000
Chain 1: 200 -17831.226 1.735 2.469
Chain 1: 300 -8825.003 1.497 1.021
Chain 1: 400 -9342.078 1.136 1.021
Chain 1: 500 -8445.504 0.930 1.000
Chain 1: 600 -8629.644 0.779 1.000
Chain 1: 700 -8321.462 0.673 0.106
Chain 1: 800 -8182.456 0.591 0.106
Chain 1: 900 -7999.921 0.528 0.055
Chain 1: 1000 -7989.540 0.475 0.055
Chain 1: 1100 -7668.200 0.379 0.042
Chain 1: 1200 -7605.346 0.133 0.037
Chain 1: 1300 -7825.480 0.034 0.028
Chain 1: 1400 -7675.618 0.030 0.023
Chain 1: 1500 -7586.704 0.021 0.021
Chain 1: 1600 -7584.086 0.019 0.020
Chain 1: 1700 -7544.664 0.016 0.017
Chain 1: 1800 -7583.432 0.014 0.012
Chain 1: 1900 -7598.367 0.012 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003653 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 36.53 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86006.522 1.000 1.000
Chain 1: 200 -13450.599 3.197 5.394
Chain 1: 300 -9854.640 2.253 1.000
Chain 1: 400 -10766.027 1.711 1.000
Chain 1: 500 -8729.624 1.415 0.365
Chain 1: 600 -8334.227 1.187 0.365
Chain 1: 700 -8726.471 1.024 0.233
Chain 1: 800 -9211.040 0.903 0.233
Chain 1: 900 -8658.605 0.810 0.085
Chain 1: 1000 -8445.911 0.731 0.085
Chain 1: 1100 -8721.246 0.634 0.064 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8409.479 0.099 0.053
Chain 1: 1300 -8558.195 0.064 0.047
Chain 1: 1400 -8564.234 0.055 0.045
Chain 1: 1500 -8433.634 0.034 0.037
Chain 1: 1600 -8543.485 0.030 0.032
Chain 1: 1700 -8630.485 0.027 0.025
Chain 1: 1800 -8226.231 0.026 0.025
Chain 1: 1900 -8323.903 0.021 0.017
Chain 1: 2000 -8295.677 0.019 0.015
Chain 1: 2100 -8415.626 0.017 0.014
Chain 1: 2200 -8209.337 0.016 0.014
Chain 1: 2300 -8359.159 0.016 0.014
Chain 1: 2400 -8364.627 0.016 0.014
Chain 1: 2500 -8337.823 0.015 0.013
Chain 1: 2600 -8337.228 0.014 0.012
Chain 1: 2700 -8248.287 0.014 0.012
Chain 1: 2800 -8214.792 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003612 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 36.12 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8386319.949 1.000 1.000
Chain 1: 200 -1584751.312 2.646 4.292
Chain 1: 300 -891348.989 2.023 1.000
Chain 1: 400 -457792.845 1.754 1.000
Chain 1: 500 -358331.389 1.459 0.947
Chain 1: 600 -233201.453 1.305 0.947
Chain 1: 700 -119303.904 1.255 0.947
Chain 1: 800 -86487.266 1.146 0.947
Chain 1: 900 -66803.582 1.051 0.778
Chain 1: 1000 -51579.912 0.975 0.778
Chain 1: 1100 -39035.898 0.908 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38207.018 0.481 0.379
Chain 1: 1300 -26145.785 0.449 0.379
Chain 1: 1400 -25861.885 0.355 0.321
Chain 1: 1500 -22444.781 0.343 0.321
Chain 1: 1600 -21659.509 0.293 0.295
Chain 1: 1700 -20531.610 0.203 0.295
Chain 1: 1800 -20475.164 0.165 0.152
Chain 1: 1900 -20801.079 0.137 0.055
Chain 1: 2000 -19311.633 0.115 0.055
Chain 1: 2100 -19550.064 0.085 0.036
Chain 1: 2200 -19776.542 0.083 0.036
Chain 1: 2300 -19393.767 0.039 0.020
Chain 1: 2400 -19165.909 0.039 0.020
Chain 1: 2500 -18968.007 0.025 0.016
Chain 1: 2600 -18598.441 0.024 0.016
Chain 1: 2700 -18555.426 0.018 0.012
Chain 1: 2800 -18272.437 0.020 0.015
Chain 1: 2900 -18553.601 0.020 0.015
Chain 1: 3000 -18539.762 0.012 0.012
Chain 1: 3100 -18624.739 0.011 0.012
Chain 1: 3200 -18315.587 0.012 0.015
Chain 1: 3300 -18520.160 0.011 0.012
Chain 1: 3400 -17995.404 0.013 0.015
Chain 1: 3500 -18606.886 0.015 0.015
Chain 1: 3600 -17914.063 0.017 0.015
Chain 1: 3700 -18300.523 0.019 0.017
Chain 1: 3800 -17261.071 0.023 0.021
Chain 1: 3900 -17257.238 0.022 0.021
Chain 1: 4000 -17374.517 0.022 0.021
Chain 1: 4100 -17288.362 0.022 0.021
Chain 1: 4200 -17104.750 0.022 0.021
Chain 1: 4300 -17243.031 0.021 0.021
Chain 1: 4400 -17200.004 0.019 0.011
Chain 1: 4500 -17102.562 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002588 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.88 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49185.420 1.000 1.000
Chain 1: 200 -18359.123 1.340 1.679
Chain 1: 300 -18881.537 0.902 1.000
Chain 1: 400 -14111.662 0.761 1.000
Chain 1: 500 -20204.698 0.669 0.338
Chain 1: 600 -15611.026 0.607 0.338
Chain 1: 700 -15891.141 0.523 0.302
Chain 1: 800 -15964.636 0.458 0.302
Chain 1: 900 -13211.778 0.430 0.294
Chain 1: 1000 -18743.294 0.417 0.295
Chain 1: 1100 -11997.480 0.373 0.295
Chain 1: 1200 -13073.212 0.213 0.294
Chain 1: 1300 -10654.637 0.233 0.294
Chain 1: 1400 -24930.711 0.257 0.294
Chain 1: 1500 -13477.192 0.311 0.294
Chain 1: 1600 -10088.031 0.316 0.295
Chain 1: 1700 -20056.938 0.364 0.336
Chain 1: 1800 -11529.052 0.437 0.497
Chain 1: 1900 -10436.340 0.427 0.497
Chain 1: 2000 -10792.436 0.400 0.497
Chain 1: 2100 -9677.109 0.356 0.336
Chain 1: 2200 -11503.183 0.363 0.336
Chain 1: 2300 -11827.559 0.343 0.336
Chain 1: 2400 -9379.467 0.312 0.261
Chain 1: 2500 -9486.446 0.228 0.159
Chain 1: 2600 -9353.233 0.196 0.115
Chain 1: 2700 -9431.006 0.147 0.105
Chain 1: 2800 -10038.616 0.079 0.061
Chain 1: 2900 -9250.838 0.077 0.061
Chain 1: 3000 -9662.711 0.078 0.061
Chain 1: 3100 -9242.740 0.071 0.045
Chain 1: 3200 -9378.673 0.057 0.043
Chain 1: 3300 -9762.939 0.058 0.043
Chain 1: 3400 -14884.475 0.067 0.043
Chain 1: 3500 -9735.262 0.118 0.045
Chain 1: 3600 -9632.211 0.118 0.045
Chain 1: 3700 -9954.465 0.120 0.045
Chain 1: 3800 -8802.766 0.127 0.045
Chain 1: 3900 -12936.991 0.151 0.045
Chain 1: 4000 -15923.490 0.165 0.131
Chain 1: 4100 -8825.569 0.241 0.188
Chain 1: 4200 -9601.908 0.248 0.188
Chain 1: 4300 -9966.981 0.248 0.188
Chain 1: 4400 -11455.701 0.226 0.131
Chain 1: 4500 -8796.157 0.204 0.131
Chain 1: 4600 -11334.118 0.225 0.188
Chain 1: 4700 -8864.242 0.249 0.224
Chain 1: 4800 -8915.209 0.237 0.224
Chain 1: 4900 -9227.412 0.208 0.188
Chain 1: 5000 -9122.719 0.191 0.130
Chain 1: 5100 -8788.271 0.114 0.081
Chain 1: 5200 -14030.867 0.143 0.130
Chain 1: 5300 -13183.684 0.146 0.130
Chain 1: 5400 -8562.830 0.187 0.224
Chain 1: 5500 -15739.474 0.203 0.224
Chain 1: 5600 -13114.144 0.200 0.200
Chain 1: 5700 -13510.256 0.175 0.064
Chain 1: 5800 -13848.003 0.177 0.064
Chain 1: 5900 -8920.769 0.229 0.200
Chain 1: 6000 -11629.332 0.251 0.233
Chain 1: 6100 -9132.526 0.275 0.273
Chain 1: 6200 -9028.074 0.238 0.233
Chain 1: 6300 -8202.829 0.242 0.233
Chain 1: 6400 -9416.292 0.201 0.200
Chain 1: 6500 -10149.690 0.163 0.129
Chain 1: 6600 -10458.568 0.146 0.101
Chain 1: 6700 -8416.266 0.167 0.129
Chain 1: 6800 -8374.554 0.165 0.129
Chain 1: 6900 -10706.965 0.131 0.129
Chain 1: 7000 -8735.561 0.131 0.129
Chain 1: 7100 -8616.093 0.105 0.101
Chain 1: 7200 -8814.251 0.106 0.101
Chain 1: 7300 -11360.314 0.118 0.129
Chain 1: 7400 -13337.975 0.120 0.148
Chain 1: 7500 -11379.885 0.130 0.172
Chain 1: 7600 -10111.605 0.140 0.172
Chain 1: 7700 -10928.292 0.123 0.148
Chain 1: 7800 -8473.755 0.151 0.172
Chain 1: 7900 -8570.487 0.131 0.148
Chain 1: 8000 -9105.618 0.114 0.125
Chain 1: 8100 -9540.392 0.117 0.125
Chain 1: 8200 -12271.060 0.137 0.148
Chain 1: 8300 -8602.563 0.157 0.148
Chain 1: 8400 -8510.566 0.144 0.125
Chain 1: 8500 -10847.426 0.148 0.125
Chain 1: 8600 -9311.936 0.152 0.165
Chain 1: 8700 -12020.843 0.167 0.215
Chain 1: 8800 -8241.400 0.184 0.215
Chain 1: 8900 -8732.551 0.188 0.215
Chain 1: 9000 -8241.331 0.189 0.215
Chain 1: 9100 -10682.357 0.207 0.223
Chain 1: 9200 -10835.623 0.186 0.215
Chain 1: 9300 -9257.871 0.160 0.170
Chain 1: 9400 -8764.052 0.165 0.170
Chain 1: 9500 -10417.947 0.159 0.165
Chain 1: 9600 -10495.326 0.144 0.159
Chain 1: 9700 -8419.615 0.146 0.159
Chain 1: 9800 -12488.785 0.132 0.159
Chain 1: 9900 -8261.617 0.178 0.170
Chain 1: 10000 -10933.919 0.196 0.229
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001385 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.85 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58578.092 1.000 1.000
Chain 1: 200 -18029.651 1.624 2.249
Chain 1: 300 -8798.145 1.433 1.049
Chain 1: 400 -8134.575 1.095 1.049
Chain 1: 500 -8350.404 0.881 1.000
Chain 1: 600 -8284.243 0.736 1.000
Chain 1: 700 -8077.628 0.634 0.082
Chain 1: 800 -8097.755 0.555 0.082
Chain 1: 900 -8025.754 0.495 0.026
Chain 1: 1000 -7788.908 0.448 0.030
Chain 1: 1100 -7929.637 0.350 0.026
Chain 1: 1200 -7742.079 0.127 0.026
Chain 1: 1300 -7795.319 0.023 0.024
Chain 1: 1400 -7893.458 0.016 0.018
Chain 1: 1500 -7603.075 0.017 0.018
Chain 1: 1600 -7782.918 0.019 0.023
Chain 1: 1700 -7604.217 0.019 0.023
Chain 1: 1800 -7606.691 0.019 0.023
Chain 1: 1900 -7708.222 0.019 0.023
Chain 1: 2000 -7627.951 0.017 0.018
Chain 1: 2100 -7508.648 0.017 0.016
Chain 1: 2200 -7964.350 0.020 0.016
Chain 1: 2300 -7584.244 0.024 0.023
Chain 1: 2400 -7636.332 0.024 0.023
Chain 1: 2500 -7558.568 0.021 0.016
Chain 1: 2600 -7512.997 0.019 0.013
Chain 1: 2700 -7516.167 0.017 0.011
Chain 1: 2800 -7486.917 0.017 0.011
Chain 1: 2900 -7398.346 0.017 0.011
Chain 1: 3000 -7537.364 0.018 0.012
Chain 1: 3100 -7524.090 0.017 0.010
Chain 1: 3200 -7722.217 0.014 0.010
Chain 1: 3300 -7446.221 0.012 0.010
Chain 1: 3400 -7666.221 0.014 0.012
Chain 1: 3500 -7430.944 0.017 0.018
Chain 1: 3600 -7497.067 0.017 0.018
Chain 1: 3700 -7445.780 0.017 0.018
Chain 1: 3800 -7444.651 0.017 0.018
Chain 1: 3900 -7412.026 0.016 0.018
Chain 1: 4000 -7406.218 0.015 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003709 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 37.09 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85981.179 1.000 1.000
Chain 1: 200 -13720.651 3.133 5.267
Chain 1: 300 -9971.922 2.214 1.000
Chain 1: 400 -11682.457 1.697 1.000
Chain 1: 500 -8595.492 1.430 0.376
Chain 1: 600 -8352.337 1.196 0.376
Chain 1: 700 -8482.516 1.027 0.359
Chain 1: 800 -8628.517 0.901 0.359
Chain 1: 900 -8922.004 0.805 0.146
Chain 1: 1000 -8727.139 0.726 0.146
Chain 1: 1100 -8567.860 0.628 0.033 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8322.184 0.105 0.030
Chain 1: 1300 -8620.901 0.070 0.030
Chain 1: 1400 -8568.664 0.056 0.029
Chain 1: 1500 -8468.102 0.022 0.022
Chain 1: 1600 -8581.803 0.020 0.019
Chain 1: 1700 -8635.626 0.019 0.019
Chain 1: 1800 -8190.738 0.023 0.022
Chain 1: 1900 -8296.205 0.021 0.019
Chain 1: 2000 -8278.989 0.019 0.013
Chain 1: 2100 -8417.438 0.019 0.013
Chain 1: 2200 -8190.434 0.019 0.013
Chain 1: 2300 -8293.088 0.016 0.013
Chain 1: 2400 -8361.147 0.017 0.013
Chain 1: 2500 -8303.654 0.016 0.013
Chain 1: 2600 -8320.314 0.015 0.012
Chain 1: 2700 -8226.333 0.015 0.012
Chain 1: 2800 -8170.713 0.011 0.011
Chain 1: 2900 -8271.249 0.011 0.011
Chain 1: 3000 -8114.653 0.012 0.012
Chain 1: 3100 -8256.115 0.012 0.012
Chain 1: 3200 -8125.037 0.011 0.012
Chain 1: 3300 -8355.436 0.013 0.012
Chain 1: 3400 -8361.109 0.012 0.012
Chain 1: 3500 -8231.588 0.013 0.016
Chain 1: 3600 -8082.276 0.015 0.016
Chain 1: 3700 -8229.383 0.015 0.017
Chain 1: 3800 -8084.919 0.016 0.018
Chain 1: 3900 -8016.707 0.016 0.018
Chain 1: 4000 -8127.910 0.015 0.017
Chain 1: 4100 -8091.985 0.014 0.016
Chain 1: 4200 -8077.835 0.013 0.016
Chain 1: 4300 -8111.290 0.010 0.014
Chain 1: 4400 -8068.218 0.011 0.014
Chain 1: 4500 -8166.227 0.010 0.012
Chain 1: 4600 -8057.745 0.010 0.012 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003412 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.12 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8390836.533 1.000 1.000
Chain 1: 200 -1584124.140 2.648 4.297
Chain 1: 300 -891981.840 2.024 1.000
Chain 1: 400 -458621.207 1.754 1.000
Chain 1: 500 -358935.387 1.459 0.945
Chain 1: 600 -233732.251 1.305 0.945
Chain 1: 700 -119735.658 1.255 0.945
Chain 1: 800 -86884.580 1.145 0.945
Chain 1: 900 -67187.365 1.050 0.776
Chain 1: 1000 -51957.952 0.975 0.776
Chain 1: 1100 -39399.959 0.907 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38579.686 0.479 0.378
Chain 1: 1300 -26490.115 0.447 0.378
Chain 1: 1400 -26208.131 0.354 0.319
Chain 1: 1500 -22782.901 0.341 0.319
Chain 1: 1600 -21996.673 0.291 0.293
Chain 1: 1700 -20864.159 0.201 0.293
Chain 1: 1800 -20807.462 0.164 0.150
Chain 1: 1900 -21134.196 0.136 0.054
Chain 1: 2000 -19641.013 0.114 0.054
Chain 1: 2100 -19879.599 0.083 0.036
Chain 1: 2200 -20107.018 0.083 0.036
Chain 1: 2300 -19723.277 0.039 0.019
Chain 1: 2400 -19495.100 0.039 0.019
Chain 1: 2500 -19297.292 0.025 0.015
Chain 1: 2600 -18926.650 0.023 0.015
Chain 1: 2700 -18883.398 0.018 0.012
Chain 1: 2800 -18600.000 0.019 0.015
Chain 1: 2900 -18881.684 0.019 0.015
Chain 1: 3000 -18867.760 0.012 0.012
Chain 1: 3100 -18952.830 0.011 0.012
Chain 1: 3200 -18643.060 0.012 0.015
Chain 1: 3300 -18848.175 0.011 0.012
Chain 1: 3400 -18322.307 0.012 0.015
Chain 1: 3500 -18935.392 0.015 0.015
Chain 1: 3600 -18240.607 0.016 0.015
Chain 1: 3700 -18628.509 0.018 0.017
Chain 1: 3800 -17585.903 0.023 0.021
Chain 1: 3900 -17582.038 0.021 0.021
Chain 1: 4000 -17699.327 0.022 0.021
Chain 1: 4100 -17612.942 0.022 0.021
Chain 1: 4200 -17428.722 0.021 0.021
Chain 1: 4300 -17567.420 0.021 0.021
Chain 1: 4400 -17523.832 0.018 0.011
Chain 1: 4500 -17426.346 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001568 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.68 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12230.190 1.000 1.000
Chain 1: 200 -9237.742 0.662 1.000
Chain 1: 300 -7956.781 0.495 0.324
Chain 1: 400 -8176.774 0.378 0.324
Chain 1: 500 -8039.064 0.306 0.161
Chain 1: 600 -7894.552 0.258 0.161
Chain 1: 700 -7818.435 0.222 0.027
Chain 1: 800 -7846.246 0.195 0.027
Chain 1: 900 -7874.745 0.174 0.018
Chain 1: 1000 -7890.560 0.157 0.018
Chain 1: 1100 -7881.846 0.057 0.017
Chain 1: 1200 -7856.480 0.025 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00159 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.9 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61654.246 1.000 1.000
Chain 1: 200 -17753.648 1.736 2.473
Chain 1: 300 -8731.139 1.502 1.033
Chain 1: 400 -9248.909 1.141 1.033
Chain 1: 500 -8159.181 0.939 1.000
Chain 1: 600 -8261.902 0.785 1.000
Chain 1: 700 -7708.860 0.683 0.134
Chain 1: 800 -7924.134 0.601 0.134
Chain 1: 900 -7932.845 0.534 0.072
Chain 1: 1000 -7708.069 0.484 0.072
Chain 1: 1100 -7561.363 0.386 0.056
Chain 1: 1200 -7696.336 0.140 0.029
Chain 1: 1300 -7535.695 0.039 0.027
Chain 1: 1400 -7603.779 0.034 0.021
Chain 1: 1500 -7490.978 0.022 0.019
Chain 1: 1600 -7694.773 0.024 0.021
Chain 1: 1700 -7423.115 0.020 0.021
Chain 1: 1800 -7509.118 0.019 0.019
Chain 1: 1900 -7512.782 0.019 0.019
Chain 1: 2000 -7565.190 0.016 0.018
Chain 1: 2100 -7469.474 0.016 0.015
Chain 1: 2200 -7594.691 0.016 0.015
Chain 1: 2300 -7494.208 0.015 0.013
Chain 1: 2400 -7531.503 0.014 0.013
Chain 1: 2500 -7457.587 0.014 0.013
Chain 1: 2600 -7403.398 0.012 0.011
Chain 1: 2700 -7397.161 0.008 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003214 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.14 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86322.908 1.000 1.000
Chain 1: 200 -13386.698 3.224 5.448
Chain 1: 300 -9776.427 2.273 1.000
Chain 1: 400 -10618.279 1.724 1.000
Chain 1: 500 -8725.974 1.423 0.369
Chain 1: 600 -8270.842 1.195 0.369
Chain 1: 700 -8420.643 1.027 0.217
Chain 1: 800 -8849.175 0.904 0.217
Chain 1: 900 -8616.668 0.807 0.079
Chain 1: 1000 -8384.161 0.729 0.079
Chain 1: 1100 -8629.081 0.632 0.055 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8269.944 0.091 0.048
Chain 1: 1300 -8479.894 0.057 0.043
Chain 1: 1400 -8477.769 0.049 0.028
Chain 1: 1500 -8368.671 0.029 0.028
Chain 1: 1600 -8472.907 0.024 0.027
Chain 1: 1700 -8561.663 0.024 0.027
Chain 1: 1800 -8157.336 0.024 0.027
Chain 1: 1900 -8255.966 0.022 0.025
Chain 1: 2000 -8227.457 0.020 0.013
Chain 1: 2100 -8347.292 0.018 0.013
Chain 1: 2200 -8151.875 0.016 0.013
Chain 1: 2300 -8290.812 0.016 0.013
Chain 1: 2400 -8166.263 0.017 0.014
Chain 1: 2500 -8231.126 0.017 0.014
Chain 1: 2600 -8254.471 0.016 0.014
Chain 1: 2700 -8172.975 0.016 0.014
Chain 1: 2800 -8145.667 0.011 0.012
Chain 1: 2900 -8201.078 0.010 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003086 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.86 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8392866.851 1.000 1.000
Chain 1: 200 -1582213.011 2.652 4.305
Chain 1: 300 -891647.241 2.026 1.000
Chain 1: 400 -458395.000 1.756 1.000
Chain 1: 500 -358994.663 1.460 0.945
Chain 1: 600 -233843.724 1.306 0.945
Chain 1: 700 -119582.561 1.256 0.945
Chain 1: 800 -86679.948 1.146 0.945
Chain 1: 900 -66923.009 1.052 0.774
Chain 1: 1000 -51637.357 0.976 0.774
Chain 1: 1100 -39045.659 0.909 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38209.396 0.480 0.380
Chain 1: 1300 -26099.811 0.449 0.380
Chain 1: 1400 -25811.279 0.356 0.322
Chain 1: 1500 -22382.123 0.343 0.322
Chain 1: 1600 -21593.584 0.294 0.296
Chain 1: 1700 -20459.610 0.204 0.295
Chain 1: 1800 -20401.715 0.166 0.153
Chain 1: 1900 -20727.621 0.138 0.055
Chain 1: 2000 -19235.293 0.116 0.055
Chain 1: 2100 -19473.768 0.085 0.037
Chain 1: 2200 -19700.803 0.084 0.037
Chain 1: 2300 -19317.546 0.040 0.020
Chain 1: 2400 -19089.640 0.040 0.020
Chain 1: 2500 -18892.032 0.025 0.016
Chain 1: 2600 -18522.126 0.024 0.016
Chain 1: 2700 -18479.006 0.018 0.012
Chain 1: 2800 -18196.162 0.020 0.016
Chain 1: 2900 -18477.382 0.020 0.015
Chain 1: 3000 -18463.465 0.012 0.012
Chain 1: 3100 -18548.477 0.011 0.012
Chain 1: 3200 -18239.212 0.012 0.015
Chain 1: 3300 -18443.875 0.011 0.012
Chain 1: 3400 -17919.090 0.013 0.015
Chain 1: 3500 -18530.666 0.015 0.016
Chain 1: 3600 -17837.725 0.017 0.016
Chain 1: 3700 -18224.342 0.019 0.017
Chain 1: 3800 -17184.738 0.023 0.021
Chain 1: 3900 -17180.940 0.022 0.021
Chain 1: 4000 -17298.182 0.022 0.021
Chain 1: 4100 -17212.037 0.022 0.021
Chain 1: 4200 -17028.384 0.022 0.021
Chain 1: 4300 -17166.654 0.021 0.021
Chain 1: 4400 -17123.616 0.019 0.011
Chain 1: 4500 -17026.191 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00134 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.4 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13112.522 1.000 1.000
Chain 1: 200 -9884.409 0.663 1.000
Chain 1: 300 -8520.045 0.496 0.327
Chain 1: 400 -8721.507 0.377 0.327
Chain 1: 500 -8659.757 0.303 0.160
Chain 1: 600 -8453.007 0.257 0.160
Chain 1: 700 -8348.842 0.222 0.024
Chain 1: 800 -8379.402 0.195 0.024
Chain 1: 900 -8463.479 0.174 0.023
Chain 1: 1000 -8417.270 0.157 0.023
Chain 1: 1100 -8436.725 0.058 0.012
Chain 1: 1200 -8353.429 0.026 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001651 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 16.51 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -62480.971 1.000 1.000
Chain 1: 200 -18639.965 1.676 2.352
Chain 1: 300 -9293.740 1.453 1.006
Chain 1: 400 -9227.899 1.091 1.006
Chain 1: 500 -8512.137 0.890 1.000
Chain 1: 600 -9390.553 0.757 1.000
Chain 1: 700 -8614.647 0.662 0.094
Chain 1: 800 -7880.244 0.591 0.094
Chain 1: 900 -8074.310 0.528 0.093
Chain 1: 1000 -8025.542 0.476 0.093
Chain 1: 1100 -7936.158 0.377 0.090
Chain 1: 1200 -8067.769 0.143 0.084
Chain 1: 1300 -8019.912 0.043 0.024
Chain 1: 1400 -7771.149 0.046 0.032
Chain 1: 1500 -7664.960 0.039 0.024
Chain 1: 1600 -7975.343 0.033 0.024
Chain 1: 1700 -7732.562 0.027 0.024
Chain 1: 1800 -7661.922 0.019 0.016
Chain 1: 1900 -7657.317 0.017 0.014
Chain 1: 2000 -7759.676 0.017 0.014
Chain 1: 2100 -7695.182 0.017 0.014
Chain 1: 2200 -7896.080 0.018 0.014
Chain 1: 2300 -7734.700 0.019 0.021
Chain 1: 2400 -7644.398 0.017 0.014
Chain 1: 2500 -7704.937 0.017 0.013
Chain 1: 2600 -7630.105 0.014 0.012
Chain 1: 2700 -7623.633 0.011 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.007439 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 74.39 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87080.815 1.000 1.000
Chain 1: 200 -14248.498 3.056 5.112
Chain 1: 300 -10494.071 2.156 1.000
Chain 1: 400 -12151.912 1.651 1.000
Chain 1: 500 -9042.422 1.390 0.358
Chain 1: 600 -9491.216 1.166 0.358
Chain 1: 700 -8811.713 1.011 0.344
Chain 1: 800 -10037.628 0.900 0.344
Chain 1: 900 -9345.123 0.808 0.136
Chain 1: 1000 -9091.651 0.730 0.136
Chain 1: 1100 -9260.407 0.632 0.122 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8790.838 0.126 0.077
Chain 1: 1300 -9117.422 0.094 0.074
Chain 1: 1400 -9036.244 0.081 0.053
Chain 1: 1500 -9001.166 0.047 0.047
Chain 1: 1600 -9074.150 0.043 0.036
Chain 1: 1700 -9130.463 0.036 0.028
Chain 1: 1800 -8681.906 0.029 0.028
Chain 1: 1900 -8788.031 0.023 0.018
Chain 1: 2000 -8782.564 0.020 0.012
Chain 1: 2100 -8947.768 0.020 0.012
Chain 1: 2200 -8682.655 0.018 0.012
Chain 1: 2300 -8870.989 0.016 0.012
Chain 1: 2400 -8682.621 0.017 0.018
Chain 1: 2500 -8760.101 0.018 0.018
Chain 1: 2600 -8680.662 0.018 0.018
Chain 1: 2700 -8704.763 0.018 0.018
Chain 1: 2800 -8658.309 0.013 0.012
Chain 1: 2900 -8764.695 0.013 0.012
Chain 1: 3000 -8715.815 0.014 0.012
Chain 1: 3100 -8649.040 0.013 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003006 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.06 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8424436.880 1.000 1.000
Chain 1: 200 -1585500.169 2.657 4.313
Chain 1: 300 -890828.956 2.031 1.000
Chain 1: 400 -458274.332 1.759 1.000
Chain 1: 500 -358488.467 1.463 0.944
Chain 1: 600 -233448.854 1.309 0.944
Chain 1: 700 -119801.319 1.257 0.944
Chain 1: 800 -87080.740 1.147 0.944
Chain 1: 900 -67457.750 1.052 0.780
Chain 1: 1000 -52296.390 0.976 0.780
Chain 1: 1100 -39807.090 0.907 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38992.645 0.478 0.376
Chain 1: 1300 -26967.602 0.444 0.376
Chain 1: 1400 -26692.012 0.351 0.314
Chain 1: 1500 -23284.392 0.338 0.314
Chain 1: 1600 -22503.529 0.288 0.291
Chain 1: 1700 -21378.685 0.198 0.290
Chain 1: 1800 -21323.773 0.161 0.146
Chain 1: 1900 -21650.578 0.133 0.053
Chain 1: 2000 -20161.460 0.112 0.053
Chain 1: 2100 -20399.763 0.081 0.035
Chain 1: 2200 -20626.605 0.080 0.035
Chain 1: 2300 -20243.326 0.038 0.019
Chain 1: 2400 -20015.213 0.038 0.019
Chain 1: 2500 -19817.176 0.024 0.015
Chain 1: 2600 -19446.661 0.023 0.015
Chain 1: 2700 -19403.487 0.018 0.012
Chain 1: 2800 -19120.039 0.019 0.015
Chain 1: 2900 -19401.558 0.019 0.015
Chain 1: 3000 -19387.706 0.011 0.012
Chain 1: 3100 -19472.785 0.011 0.011
Chain 1: 3200 -19163.015 0.011 0.015
Chain 1: 3300 -19368.106 0.010 0.011
Chain 1: 3400 -18842.212 0.012 0.015
Chain 1: 3500 -19455.253 0.014 0.015
Chain 1: 3600 -18760.405 0.016 0.015
Chain 1: 3700 -19148.319 0.018 0.016
Chain 1: 3800 -18105.607 0.022 0.020
Chain 1: 3900 -18101.682 0.021 0.020
Chain 1: 4000 -18219.006 0.021 0.020
Chain 1: 4100 -18132.635 0.021 0.020
Chain 1: 4200 -17948.371 0.021 0.020
Chain 1: 4300 -18087.126 0.020 0.020
Chain 1: 4400 -18043.496 0.018 0.010
Chain 1: 4500 -17945.969 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001356 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.56 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49074.605 1.000 1.000
Chain 1: 200 -18915.902 1.297 1.594
Chain 1: 300 -20161.255 0.885 1.000
Chain 1: 400 -16078.893 0.728 1.000
Chain 1: 500 -20667.820 0.626 0.254
Chain 1: 600 -12815.739 0.624 0.613
Chain 1: 700 -15736.540 0.561 0.254
Chain 1: 800 -14313.137 0.504 0.254
Chain 1: 900 -12498.457 0.464 0.222
Chain 1: 1000 -14958.633 0.434 0.222
Chain 1: 1100 -10514.553 0.376 0.222
Chain 1: 1200 -18593.116 0.260 0.222
Chain 1: 1300 -15813.478 0.272 0.222
Chain 1: 1400 -14335.263 0.257 0.186
Chain 1: 1500 -14525.066 0.236 0.176
Chain 1: 1600 -11248.357 0.204 0.176
Chain 1: 1700 -12788.405 0.197 0.164
Chain 1: 1800 -11347.181 0.200 0.164
Chain 1: 1900 -10035.066 0.198 0.164
Chain 1: 2000 -9861.171 0.184 0.131
Chain 1: 2100 -22216.395 0.197 0.131
Chain 1: 2200 -12048.251 0.238 0.131
Chain 1: 2300 -25602.025 0.273 0.131
Chain 1: 2400 -9839.678 0.423 0.291
Chain 1: 2500 -11245.931 0.434 0.291
Chain 1: 2600 -9424.851 0.425 0.193
Chain 1: 2700 -9189.404 0.415 0.193
Chain 1: 2800 -11127.665 0.420 0.193
Chain 1: 2900 -17000.721 0.441 0.345
Chain 1: 3000 -9021.866 0.528 0.529 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 3100 -9420.091 0.477 0.345
Chain 1: 3200 -10559.189 0.403 0.193
Chain 1: 3300 -14358.132 0.376 0.193
Chain 1: 3400 -9251.198 0.271 0.193
Chain 1: 3500 -10008.619 0.267 0.193
Chain 1: 3600 -9856.349 0.249 0.174
Chain 1: 3700 -19271.246 0.295 0.265
Chain 1: 3800 -12491.436 0.332 0.345
Chain 1: 3900 -9869.997 0.324 0.266
Chain 1: 4000 -10648.112 0.243 0.265
Chain 1: 4100 -12834.239 0.256 0.265
Chain 1: 4200 -10773.036 0.264 0.265
Chain 1: 4300 -10536.814 0.240 0.191
Chain 1: 4400 -9601.079 0.194 0.170
Chain 1: 4500 -13009.093 0.213 0.191
Chain 1: 4600 -13879.235 0.218 0.191
Chain 1: 4700 -12768.678 0.177 0.170
Chain 1: 4800 -9063.030 0.164 0.170
Chain 1: 4900 -12474.483 0.165 0.170
Chain 1: 5000 -13724.919 0.167 0.170
Chain 1: 5100 -9146.501 0.200 0.191
Chain 1: 5200 -10969.262 0.197 0.166
Chain 1: 5300 -11789.460 0.202 0.166
Chain 1: 5400 -16428.231 0.220 0.262
Chain 1: 5500 -11794.023 0.233 0.273
Chain 1: 5600 -9342.189 0.253 0.273
Chain 1: 5700 -8781.914 0.251 0.273
Chain 1: 5800 -9774.467 0.220 0.262
Chain 1: 5900 -14723.017 0.227 0.262
Chain 1: 6000 -12736.266 0.233 0.262
Chain 1: 6100 -8920.058 0.226 0.262
Chain 1: 6200 -8534.791 0.214 0.262
Chain 1: 6300 -8893.520 0.211 0.262
Chain 1: 6400 -8440.068 0.188 0.156
Chain 1: 6500 -12617.182 0.182 0.156
Chain 1: 6600 -9809.222 0.184 0.156
Chain 1: 6700 -10999.278 0.189 0.156
Chain 1: 6800 -8594.299 0.206 0.280
Chain 1: 6900 -8885.325 0.176 0.156
Chain 1: 7000 -8978.602 0.162 0.108
Chain 1: 7100 -9188.951 0.121 0.054
Chain 1: 7200 -10199.569 0.126 0.099
Chain 1: 7300 -8281.064 0.146 0.108
Chain 1: 7400 -8541.377 0.143 0.108
Chain 1: 7500 -9040.203 0.116 0.099
Chain 1: 7600 -10595.994 0.102 0.099
Chain 1: 7700 -9233.200 0.106 0.099
Chain 1: 7800 -8714.246 0.084 0.060
Chain 1: 7900 -9322.701 0.087 0.065
Chain 1: 8000 -8497.937 0.096 0.097
Chain 1: 8100 -8621.079 0.095 0.097
Chain 1: 8200 -8710.127 0.086 0.065
Chain 1: 8300 -8287.681 0.068 0.060
Chain 1: 8400 -8677.975 0.069 0.060
Chain 1: 8500 -8216.005 0.069 0.060
Chain 1: 8600 -8436.655 0.057 0.056
Chain 1: 8700 -8723.114 0.046 0.051
Chain 1: 8800 -8280.794 0.045 0.051
Chain 1: 8900 -11920.755 0.069 0.051
Chain 1: 9000 -8722.258 0.096 0.051
Chain 1: 9100 -8916.844 0.097 0.051
Chain 1: 9200 -8278.130 0.104 0.053
Chain 1: 9300 -8223.219 0.099 0.053
Chain 1: 9400 -8488.802 0.098 0.053
Chain 1: 9500 -10083.563 0.108 0.053
Chain 1: 9600 -10292.399 0.107 0.053
Chain 1: 9700 -9168.842 0.116 0.077
Chain 1: 9800 -8563.180 0.118 0.077
Chain 1: 9900 -11377.203 0.112 0.077
Chain 1: 10000 -9572.359 0.094 0.077
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001437 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.37 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58563.441 1.000 1.000
Chain 1: 200 -18003.070 1.626 2.253
Chain 1: 300 -8798.739 1.433 1.046
Chain 1: 400 -8215.280 1.093 1.046
Chain 1: 500 -8237.904 0.875 1.000
Chain 1: 600 -8902.460 0.741 1.000
Chain 1: 700 -8422.667 0.643 0.075
Chain 1: 800 -8222.578 0.566 0.075
Chain 1: 900 -8078.641 0.505 0.071
Chain 1: 1000 -7799.459 0.458 0.071
Chain 1: 1100 -8002.549 0.361 0.057
Chain 1: 1200 -7858.622 0.137 0.036
Chain 1: 1300 -7613.954 0.036 0.032
Chain 1: 1400 -7905.737 0.033 0.032
Chain 1: 1500 -7592.995 0.036 0.036
Chain 1: 1600 -7754.445 0.031 0.032
Chain 1: 1700 -7683.891 0.026 0.025
Chain 1: 1800 -7665.308 0.024 0.025
Chain 1: 1900 -7578.556 0.023 0.025
Chain 1: 2000 -7643.328 0.021 0.021
Chain 1: 2100 -7583.273 0.019 0.018
Chain 1: 2200 -7739.968 0.019 0.020
Chain 1: 2300 -7566.487 0.018 0.020
Chain 1: 2400 -7632.228 0.015 0.011
Chain 1: 2500 -7641.906 0.011 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004281 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 42.81 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86190.966 1.000 1.000
Chain 1: 200 -13716.159 3.142 5.284
Chain 1: 300 -10018.455 2.218 1.000
Chain 1: 400 -11170.528 1.689 1.000
Chain 1: 500 -9028.226 1.399 0.369
Chain 1: 600 -8392.816 1.178 0.369
Chain 1: 700 -8531.413 1.012 0.237
Chain 1: 800 -8965.349 0.892 0.237
Chain 1: 900 -8830.991 0.794 0.103
Chain 1: 1000 -8683.545 0.717 0.103
Chain 1: 1100 -8650.774 0.617 0.076 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8401.862 0.092 0.048
Chain 1: 1300 -8673.284 0.058 0.031
Chain 1: 1400 -8651.368 0.048 0.030
Chain 1: 1500 -8547.956 0.025 0.017
Chain 1: 1600 -8655.051 0.019 0.016
Chain 1: 1700 -8726.484 0.018 0.015
Chain 1: 1800 -8293.053 0.018 0.015
Chain 1: 1900 -8397.433 0.018 0.012
Chain 1: 2000 -8372.785 0.017 0.012
Chain 1: 2100 -8510.697 0.018 0.012
Chain 1: 2200 -8303.819 0.018 0.012
Chain 1: 2300 -8442.253 0.016 0.012
Chain 1: 2400 -8301.294 0.017 0.016
Chain 1: 2500 -8371.111 0.017 0.016
Chain 1: 2600 -8284.275 0.017 0.016
Chain 1: 2700 -8317.551 0.016 0.016
Chain 1: 2800 -8278.804 0.012 0.012
Chain 1: 2900 -8370.819 0.012 0.011
Chain 1: 3000 -8197.148 0.013 0.016
Chain 1: 3100 -8360.902 0.014 0.016
Chain 1: 3200 -8233.751 0.013 0.015
Chain 1: 3300 -8242.882 0.011 0.011
Chain 1: 3400 -8394.454 0.011 0.011
Chain 1: 3500 -8385.756 0.011 0.011
Chain 1: 3600 -8191.014 0.012 0.015
Chain 1: 3700 -8334.151 0.013 0.017
Chain 1: 3800 -8197.863 0.014 0.017
Chain 1: 3900 -8133.080 0.014 0.017
Chain 1: 4000 -8207.342 0.013 0.017
Chain 1: 4100 -8198.403 0.011 0.015
Chain 1: 4200 -8186.952 0.010 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003401 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.01 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8398252.217 1.000 1.000
Chain 1: 200 -1585752.488 2.648 4.296
Chain 1: 300 -891212.943 2.025 1.000
Chain 1: 400 -457932.084 1.755 1.000
Chain 1: 500 -358230.244 1.460 0.946
Chain 1: 600 -233126.186 1.306 0.946
Chain 1: 700 -119429.769 1.256 0.946
Chain 1: 800 -86634.703 1.146 0.946
Chain 1: 900 -66997.153 1.051 0.779
Chain 1: 1000 -51809.572 0.975 0.779
Chain 1: 1100 -39293.700 0.907 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38473.519 0.480 0.379
Chain 1: 1300 -26435.054 0.447 0.379
Chain 1: 1400 -26155.345 0.354 0.319
Chain 1: 1500 -22743.440 0.341 0.319
Chain 1: 1600 -21960.392 0.291 0.293
Chain 1: 1700 -20834.743 0.201 0.293
Chain 1: 1800 -20779.214 0.163 0.150
Chain 1: 1900 -21105.668 0.136 0.054
Chain 1: 2000 -19616.208 0.114 0.054
Chain 1: 2100 -19854.757 0.083 0.036
Chain 1: 2200 -20081.384 0.082 0.036
Chain 1: 2300 -19698.339 0.039 0.019
Chain 1: 2400 -19470.311 0.039 0.019
Chain 1: 2500 -19272.202 0.025 0.015
Chain 1: 2600 -18902.250 0.023 0.015
Chain 1: 2700 -18859.087 0.018 0.012
Chain 1: 2800 -18575.766 0.019 0.015
Chain 1: 2900 -18857.142 0.019 0.015
Chain 1: 3000 -18843.352 0.012 0.012
Chain 1: 3100 -18928.408 0.011 0.012
Chain 1: 3200 -18618.899 0.012 0.015
Chain 1: 3300 -18823.738 0.011 0.012
Chain 1: 3400 -18298.277 0.012 0.015
Chain 1: 3500 -18910.741 0.015 0.015
Chain 1: 3600 -18216.626 0.016 0.015
Chain 1: 3700 -18604.021 0.018 0.017
Chain 1: 3800 -17562.517 0.023 0.021
Chain 1: 3900 -17558.603 0.021 0.021
Chain 1: 4000 -17675.932 0.022 0.021
Chain 1: 4100 -17589.651 0.022 0.021
Chain 1: 4200 -17405.592 0.021 0.021
Chain 1: 4300 -17544.229 0.021 0.021
Chain 1: 4400 -17500.839 0.018 0.011
Chain 1: 4500 -17403.299 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001273 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.73 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12412.139 1.000 1.000
Chain 1: 200 -9438.431 0.658 1.000
Chain 1: 300 -8273.389 0.485 0.315
Chain 1: 400 -8454.626 0.369 0.315
Chain 1: 500 -8300.830 0.299 0.141
Chain 1: 600 -8175.848 0.252 0.141
Chain 1: 700 -8111.411 0.217 0.021
Chain 1: 800 -8118.926 0.190 0.021
Chain 1: 900 -8135.430 0.169 0.019
Chain 1: 1000 -8171.453 0.153 0.019
Chain 1: 1100 -8235.263 0.053 0.015
Chain 1: 1200 -8122.431 0.023 0.014
Chain 1: 1300 -8106.648 0.009 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001425 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.25 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61814.033 1.000 1.000
Chain 1: 200 -17879.878 1.729 2.457
Chain 1: 300 -8860.229 1.492 1.018
Chain 1: 400 -9463.806 1.135 1.018
Chain 1: 500 -8004.298 0.944 1.000
Chain 1: 600 -8729.240 0.801 1.000
Chain 1: 700 -7788.976 0.704 0.182
Chain 1: 800 -7755.478 0.616 0.182
Chain 1: 900 -8025.820 0.551 0.121
Chain 1: 1000 -7683.180 0.501 0.121
Chain 1: 1100 -7712.312 0.401 0.083
Chain 1: 1200 -7708.562 0.155 0.064
Chain 1: 1300 -7758.453 0.054 0.045
Chain 1: 1400 -7656.659 0.049 0.034
Chain 1: 1500 -7636.295 0.031 0.013
Chain 1: 1600 -7777.988 0.025 0.013
Chain 1: 1700 -7549.515 0.016 0.013
Chain 1: 1800 -7650.773 0.017 0.013
Chain 1: 1900 -7588.105 0.014 0.013
Chain 1: 2000 -7632.563 0.010 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004008 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 40.08 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86149.671 1.000 1.000
Chain 1: 200 -13529.264 3.184 5.368
Chain 1: 300 -9986.756 2.241 1.000
Chain 1: 400 -10823.306 1.700 1.000
Chain 1: 500 -8892.818 1.403 0.355
Chain 1: 600 -8538.267 1.176 0.355
Chain 1: 700 -8957.313 1.015 0.217
Chain 1: 800 -8816.986 0.890 0.217
Chain 1: 900 -8861.779 0.792 0.077
Chain 1: 1000 -8728.425 0.714 0.077
Chain 1: 1100 -8922.682 0.616 0.047 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8641.947 0.083 0.042
Chain 1: 1300 -8689.966 0.048 0.032
Chain 1: 1400 -8759.374 0.041 0.022
Chain 1: 1500 -8616.014 0.021 0.017
Chain 1: 1600 -8719.488 0.018 0.016
Chain 1: 1700 -8803.693 0.014 0.015
Chain 1: 1800 -8419.570 0.017 0.015
Chain 1: 1900 -8521.585 0.018 0.015
Chain 1: 2000 -8491.342 0.017 0.012
Chain 1: 2100 -8625.363 0.016 0.012
Chain 1: 2200 -8410.121 0.015 0.012
Chain 1: 2300 -8551.325 0.016 0.016
Chain 1: 2400 -8562.585 0.016 0.016
Chain 1: 2500 -8530.872 0.015 0.012
Chain 1: 2600 -8529.113 0.013 0.012
Chain 1: 2700 -8438.241 0.013 0.012
Chain 1: 2800 -8416.039 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003665 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 36.65 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8377493.232 1.000 1.000
Chain 1: 200 -1579135.217 2.653 4.305
Chain 1: 300 -890623.990 2.026 1.000
Chain 1: 400 -458042.576 1.756 1.000
Chain 1: 500 -358785.328 1.460 0.944
Chain 1: 600 -233733.567 1.306 0.944
Chain 1: 700 -119603.894 1.255 0.944
Chain 1: 800 -86731.911 1.146 0.944
Chain 1: 900 -66997.802 1.051 0.773
Chain 1: 1000 -51729.343 0.976 0.773
Chain 1: 1100 -39150.637 0.908 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38315.613 0.480 0.379
Chain 1: 1300 -26223.296 0.448 0.379
Chain 1: 1400 -25935.065 0.355 0.321
Chain 1: 1500 -22510.702 0.343 0.321
Chain 1: 1600 -21722.841 0.293 0.295
Chain 1: 1700 -20591.319 0.203 0.295
Chain 1: 1800 -20533.857 0.165 0.152
Chain 1: 1900 -20859.479 0.137 0.055
Chain 1: 2000 -19368.979 0.115 0.055
Chain 1: 2100 -19607.291 0.084 0.036
Chain 1: 2200 -19833.917 0.083 0.036
Chain 1: 2300 -19451.120 0.039 0.020
Chain 1: 2400 -19223.374 0.039 0.020
Chain 1: 2500 -19025.673 0.025 0.016
Chain 1: 2600 -18656.160 0.024 0.016
Chain 1: 2700 -18613.187 0.018 0.012
Chain 1: 2800 -18330.456 0.020 0.015
Chain 1: 2900 -18611.531 0.020 0.015
Chain 1: 3000 -18597.640 0.012 0.012
Chain 1: 3100 -18682.575 0.011 0.012
Chain 1: 3200 -18373.561 0.012 0.015
Chain 1: 3300 -18578.045 0.011 0.012
Chain 1: 3400 -18053.651 0.013 0.015
Chain 1: 3500 -18664.598 0.015 0.015
Chain 1: 3600 -17972.517 0.017 0.015
Chain 1: 3700 -18358.488 0.018 0.017
Chain 1: 3800 -17320.152 0.023 0.021
Chain 1: 3900 -17316.397 0.021 0.021
Chain 1: 4000 -17433.643 0.022 0.021
Chain 1: 4100 -17347.544 0.022 0.021
Chain 1: 4200 -17164.202 0.021 0.021
Chain 1: 4300 -17302.262 0.021 0.021
Chain 1: 4400 -17259.444 0.019 0.011
Chain 1: 4500 -17162.086 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00146 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.6 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -11956.415 1.000 1.000
Chain 1: 200 -8858.715 0.675 1.000
Chain 1: 300 -7843.768 0.493 0.350
Chain 1: 400 -7931.860 0.373 0.350
Chain 1: 500 -7620.268 0.306 0.129
Chain 1: 600 -7643.956 0.256 0.129
Chain 1: 700 -7600.507 0.220 0.041
Chain 1: 800 -7690.480 0.194 0.041
Chain 1: 900 -7649.027 0.173 0.012
Chain 1: 1000 -7652.575 0.156 0.012
Chain 1: 1100 -7700.658 0.056 0.011
Chain 1: 1200 -7633.100 0.022 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001366 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.66 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57371.577 1.000 1.000
Chain 1: 200 -17079.360 1.680 2.359
Chain 1: 300 -8360.496 1.467 1.043
Chain 1: 400 -7914.132 1.115 1.043
Chain 1: 500 -8144.591 0.897 1.000
Chain 1: 600 -8532.622 0.755 1.000
Chain 1: 700 -8035.301 0.656 0.062
Chain 1: 800 -7942.007 0.576 0.062
Chain 1: 900 -7752.815 0.514 0.056
Chain 1: 1000 -7689.182 0.464 0.056
Chain 1: 1100 -7581.039 0.365 0.045
Chain 1: 1200 -7455.947 0.131 0.028
Chain 1: 1300 -7450.502 0.027 0.024
Chain 1: 1400 -7691.961 0.024 0.024
Chain 1: 1500 -7459.782 0.025 0.024
Chain 1: 1600 -7456.963 0.020 0.017
Chain 1: 1700 -7337.484 0.016 0.016
Chain 1: 1800 -7400.866 0.015 0.016
Chain 1: 1900 -7418.364 0.013 0.014
Chain 1: 2000 -7424.728 0.012 0.014
Chain 1: 2100 -7449.133 0.011 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002989 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.89 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86351.306 1.000 1.000
Chain 1: 200 -12956.849 3.332 5.665
Chain 1: 300 -9442.562 2.346 1.000
Chain 1: 400 -10264.146 1.779 1.000
Chain 1: 500 -8295.883 1.471 0.372
Chain 1: 600 -8029.954 1.231 0.372
Chain 1: 700 -8323.633 1.060 0.237
Chain 1: 800 -8590.845 0.932 0.237
Chain 1: 900 -8344.998 0.831 0.080
Chain 1: 1000 -8070.653 0.752 0.080
Chain 1: 1100 -8363.789 0.655 0.035 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8112.888 0.092 0.035
Chain 1: 1300 -8038.962 0.056 0.034
Chain 1: 1400 -8040.535 0.048 0.033
Chain 1: 1500 -8075.359 0.024 0.031
Chain 1: 1600 -8080.993 0.021 0.031
Chain 1: 1700 -8018.888 0.018 0.029
Chain 1: 1800 -7899.799 0.017 0.015
Chain 1: 1900 -8014.389 0.015 0.014
Chain 1: 2000 -7975.243 0.012 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003165 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.65 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8439194.815 1.000 1.000
Chain 1: 200 -1591386.659 2.652 4.303
Chain 1: 300 -890848.374 2.030 1.000
Chain 1: 400 -456581.306 1.760 1.000
Chain 1: 500 -356159.953 1.465 0.951
Chain 1: 600 -231333.463 1.310 0.951
Chain 1: 700 -118096.536 1.260 0.951
Chain 1: 800 -85413.662 1.150 0.951
Chain 1: 900 -65871.359 1.056 0.786
Chain 1: 1000 -50747.837 0.980 0.786
Chain 1: 1100 -38306.112 0.912 0.540 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37487.782 0.484 0.383
Chain 1: 1300 -25543.926 0.452 0.383
Chain 1: 1400 -25268.904 0.358 0.325
Chain 1: 1500 -21881.783 0.346 0.325
Chain 1: 1600 -21104.642 0.295 0.298
Chain 1: 1700 -19991.355 0.205 0.297
Chain 1: 1800 -19937.995 0.167 0.155
Chain 1: 1900 -20263.426 0.139 0.056
Chain 1: 2000 -18782.469 0.117 0.056
Chain 1: 2100 -19020.559 0.086 0.037
Chain 1: 2200 -19245.292 0.085 0.037
Chain 1: 2300 -18864.172 0.040 0.020
Chain 1: 2400 -18636.664 0.040 0.020
Chain 1: 2500 -18438.128 0.026 0.016
Chain 1: 2600 -18069.664 0.024 0.016
Chain 1: 2700 -18027.036 0.019 0.013
Chain 1: 2800 -17743.968 0.020 0.016
Chain 1: 2900 -18024.749 0.020 0.016
Chain 1: 3000 -18011.155 0.012 0.013
Chain 1: 3100 -18095.969 0.011 0.012
Chain 1: 3200 -17787.297 0.012 0.016
Chain 1: 3300 -17991.513 0.011 0.012
Chain 1: 3400 -17467.362 0.013 0.016
Chain 1: 3500 -18077.654 0.015 0.016
Chain 1: 3600 -17386.390 0.017 0.016
Chain 1: 3700 -17771.572 0.019 0.017
Chain 1: 3800 -16734.319 0.024 0.022
Chain 1: 3900 -16730.450 0.022 0.022
Chain 1: 4000 -16847.844 0.023 0.022
Chain 1: 4100 -16761.693 0.023 0.022
Chain 1: 4200 -16578.621 0.022 0.022
Chain 1: 4300 -16716.605 0.022 0.022
Chain 1: 4400 -16673.993 0.019 0.011
Chain 1: 4500 -16576.549 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001402 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.02 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49231.557 1.000 1.000
Chain 1: 200 -20832.210 1.182 1.363
Chain 1: 300 -16114.681 0.885 1.000
Chain 1: 400 -18513.583 0.696 1.000
Chain 1: 500 -19452.348 0.567 0.293
Chain 1: 600 -12674.444 0.561 0.535
Chain 1: 700 -15129.483 0.504 0.293
Chain 1: 800 -15606.907 0.445 0.293
Chain 1: 900 -13628.553 0.412 0.162
Chain 1: 1000 -10853.358 0.396 0.256
Chain 1: 1100 -16484.464 0.330 0.256
Chain 1: 1200 -10523.158 0.251 0.256
Chain 1: 1300 -10147.530 0.225 0.162
Chain 1: 1400 -14910.074 0.244 0.256
Chain 1: 1500 -10908.920 0.276 0.319
Chain 1: 1600 -10400.640 0.227 0.256
Chain 1: 1700 -11532.606 0.221 0.256
Chain 1: 1800 -10532.040 0.227 0.256
Chain 1: 1900 -10835.298 0.216 0.256
Chain 1: 2000 -13018.244 0.207 0.168
Chain 1: 2100 -10754.770 0.194 0.168
Chain 1: 2200 -9613.681 0.149 0.119
Chain 1: 2300 -18322.817 0.193 0.168
Chain 1: 2400 -11603.422 0.219 0.168
Chain 1: 2500 -11647.434 0.183 0.119
Chain 1: 2600 -9525.911 0.200 0.168
Chain 1: 2700 -9570.635 0.191 0.168
Chain 1: 2800 -9425.013 0.183 0.168
Chain 1: 2900 -14488.826 0.215 0.210
Chain 1: 3000 -16475.162 0.210 0.210
Chain 1: 3100 -9025.017 0.272 0.223
Chain 1: 3200 -11692.609 0.282 0.228
Chain 1: 3300 -15490.597 0.259 0.228
Chain 1: 3400 -9718.467 0.261 0.228
Chain 1: 3500 -9829.591 0.262 0.228
Chain 1: 3600 -10562.754 0.246 0.228
Chain 1: 3700 -17306.904 0.285 0.245
Chain 1: 3800 -8821.816 0.380 0.349
Chain 1: 3900 -13261.901 0.378 0.335
Chain 1: 4000 -9044.061 0.413 0.390
Chain 1: 4100 -9210.991 0.332 0.335
Chain 1: 4200 -9936.654 0.316 0.335
Chain 1: 4300 -12883.218 0.315 0.335
Chain 1: 4400 -12081.763 0.262 0.229
Chain 1: 4500 -10472.371 0.276 0.229
Chain 1: 4600 -8753.925 0.289 0.229
Chain 1: 4700 -9197.938 0.255 0.196
Chain 1: 4800 -13901.267 0.192 0.196
Chain 1: 4900 -9233.008 0.209 0.196
Chain 1: 5000 -10964.235 0.179 0.158
Chain 1: 5100 -10657.790 0.180 0.158
Chain 1: 5200 -9217.591 0.188 0.158
Chain 1: 5300 -12068.600 0.189 0.158
Chain 1: 5400 -8539.575 0.223 0.196
Chain 1: 5500 -11742.403 0.235 0.236
Chain 1: 5600 -11480.852 0.218 0.236
Chain 1: 5700 -9771.042 0.231 0.236
Chain 1: 5800 -8578.273 0.211 0.175
Chain 1: 5900 -12469.385 0.191 0.175
Chain 1: 6000 -8550.142 0.221 0.236
Chain 1: 6100 -8498.320 0.219 0.236
Chain 1: 6200 -8685.893 0.206 0.236
Chain 1: 6300 -12701.616 0.214 0.273
Chain 1: 6400 -9212.824 0.210 0.273
Chain 1: 6500 -9056.849 0.185 0.175
Chain 1: 6600 -8526.202 0.189 0.175
Chain 1: 6700 -11760.430 0.199 0.275
Chain 1: 6800 -8627.315 0.221 0.312
Chain 1: 6900 -8376.107 0.193 0.275
Chain 1: 7000 -8561.274 0.149 0.062
Chain 1: 7100 -8409.101 0.150 0.062
Chain 1: 7200 -8884.525 0.154 0.062
Chain 1: 7300 -8846.408 0.122 0.054
Chain 1: 7400 -8831.791 0.085 0.030
Chain 1: 7500 -12644.904 0.113 0.054
Chain 1: 7600 -8788.112 0.151 0.054
Chain 1: 7700 -12267.864 0.152 0.054
Chain 1: 7800 -9698.508 0.142 0.054
Chain 1: 7900 -9475.812 0.141 0.054
Chain 1: 8000 -8438.041 0.151 0.123
Chain 1: 8100 -8395.199 0.150 0.123
Chain 1: 8200 -9226.498 0.154 0.123
Chain 1: 8300 -8445.660 0.162 0.123
Chain 1: 8400 -8411.768 0.163 0.123
Chain 1: 8500 -8737.358 0.136 0.092
Chain 1: 8600 -11004.653 0.113 0.092
Chain 1: 8700 -9013.578 0.107 0.092
Chain 1: 8800 -10958.674 0.098 0.092
Chain 1: 8900 -11922.767 0.104 0.092
Chain 1: 9000 -8448.642 0.133 0.092
Chain 1: 9100 -8503.972 0.133 0.092
Chain 1: 9200 -8898.391 0.128 0.092
Chain 1: 9300 -10062.078 0.130 0.116
Chain 1: 9400 -8399.564 0.150 0.177
Chain 1: 9500 -8468.978 0.147 0.177
Chain 1: 9600 -10461.192 0.145 0.177
Chain 1: 9700 -11316.605 0.131 0.116
Chain 1: 9800 -9305.729 0.135 0.116
Chain 1: 9900 -8265.508 0.139 0.126
Chain 1: 10000 -9391.513 0.110 0.120
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001423 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.23 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58272.200 1.000 1.000
Chain 1: 200 -17817.194 1.635 2.271
Chain 1: 300 -8800.349 1.432 1.025
Chain 1: 400 -8211.850 1.092 1.025
Chain 1: 500 -8723.169 0.885 1.000
Chain 1: 600 -8514.297 0.742 1.000
Chain 1: 700 -8052.522 0.644 0.072
Chain 1: 800 -8355.882 0.568 0.072
Chain 1: 900 -7924.999 0.511 0.059
Chain 1: 1000 -8012.010 0.461 0.059
Chain 1: 1100 -7769.171 0.364 0.057
Chain 1: 1200 -7701.493 0.138 0.054
Chain 1: 1300 -7767.802 0.036 0.036
Chain 1: 1400 -7920.133 0.031 0.031
Chain 1: 1500 -7625.219 0.029 0.031
Chain 1: 1600 -7572.119 0.027 0.031
Chain 1: 1700 -7685.645 0.023 0.019
Chain 1: 1800 -7700.967 0.020 0.015
Chain 1: 1900 -7696.986 0.014 0.011
Chain 1: 2000 -7651.010 0.014 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003683 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 36.83 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86823.303 1.000 1.000
Chain 1: 200 -13664.731 3.177 5.354
Chain 1: 300 -10001.645 2.240 1.000
Chain 1: 400 -10966.523 1.702 1.000
Chain 1: 500 -8989.777 1.406 0.366
Chain 1: 600 -8518.093 1.181 0.366
Chain 1: 700 -8507.938 1.012 0.220
Chain 1: 800 -8693.234 0.888 0.220
Chain 1: 900 -8701.004 0.790 0.088
Chain 1: 1000 -8707.171 0.711 0.088
Chain 1: 1100 -8825.701 0.612 0.055 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8322.739 0.083 0.055
Chain 1: 1300 -8753.182 0.051 0.049
Chain 1: 1400 -8704.400 0.043 0.021
Chain 1: 1500 -8543.757 0.023 0.019
Chain 1: 1600 -8663.678 0.019 0.014
Chain 1: 1700 -8732.330 0.019 0.014
Chain 1: 1800 -8307.906 0.022 0.014
Chain 1: 1900 -8409.466 0.023 0.014
Chain 1: 2000 -8384.094 0.024 0.014
Chain 1: 2100 -8510.517 0.024 0.015
Chain 1: 2200 -8311.015 0.020 0.015
Chain 1: 2300 -8404.472 0.016 0.014
Chain 1: 2400 -8472.873 0.016 0.014
Chain 1: 2500 -8419.123 0.015 0.012
Chain 1: 2600 -8421.011 0.014 0.011
Chain 1: 2700 -8337.492 0.014 0.011
Chain 1: 2800 -8296.668 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003384 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.84 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8424777.068 1.000 1.000
Chain 1: 200 -1588603.705 2.652 4.303
Chain 1: 300 -890497.251 2.029 1.000
Chain 1: 400 -457183.172 1.759 1.000
Chain 1: 500 -357391.558 1.463 0.948
Chain 1: 600 -232462.152 1.309 0.948
Chain 1: 700 -119038.721 1.258 0.948
Chain 1: 800 -86333.288 1.148 0.948
Chain 1: 900 -66750.017 1.053 0.784
Chain 1: 1000 -51609.491 0.977 0.784
Chain 1: 1100 -39140.365 0.909 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38323.550 0.481 0.379
Chain 1: 1300 -26334.525 0.448 0.379
Chain 1: 1400 -26058.526 0.354 0.319
Chain 1: 1500 -22659.609 0.341 0.319
Chain 1: 1600 -21880.387 0.291 0.293
Chain 1: 1700 -20760.624 0.201 0.293
Chain 1: 1800 -20706.292 0.163 0.150
Chain 1: 1900 -21032.617 0.136 0.054
Chain 1: 2000 -19546.789 0.114 0.054
Chain 1: 2100 -19785.097 0.083 0.036
Chain 1: 2200 -20011.061 0.082 0.036
Chain 1: 2300 -19628.611 0.039 0.019
Chain 1: 2400 -19400.725 0.039 0.019
Chain 1: 2500 -19202.433 0.025 0.016
Chain 1: 2600 -18832.861 0.023 0.016
Chain 1: 2700 -18789.857 0.018 0.012
Chain 1: 2800 -18506.550 0.019 0.015
Chain 1: 2900 -18787.755 0.019 0.015
Chain 1: 3000 -18774.014 0.012 0.012
Chain 1: 3100 -18859.019 0.011 0.012
Chain 1: 3200 -18549.689 0.012 0.015
Chain 1: 3300 -18754.401 0.011 0.012
Chain 1: 3400 -18229.212 0.012 0.015
Chain 1: 3500 -18841.175 0.015 0.015
Chain 1: 3600 -18147.673 0.016 0.015
Chain 1: 3700 -18534.593 0.018 0.017
Chain 1: 3800 -17493.966 0.023 0.021
Chain 1: 3900 -17490.030 0.021 0.021
Chain 1: 4000 -17607.394 0.022 0.021
Chain 1: 4100 -17521.151 0.022 0.021
Chain 1: 4200 -17337.278 0.021 0.021
Chain 1: 4300 -17475.800 0.021 0.021
Chain 1: 4400 -17432.581 0.018 0.011
Chain 1: 4500 -17335.033 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003979 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 39.79 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13360.520 1.000 1.000
Chain 1: 200 -10013.792 0.667 1.000
Chain 1: 300 -8515.016 0.503 0.334
Chain 1: 400 -8835.081 0.387 0.334
Chain 1: 500 -8526.659 0.317 0.176
Chain 1: 600 -8513.979 0.264 0.176
Chain 1: 700 -8397.537 0.228 0.036
Chain 1: 800 -8495.965 0.201 0.036
Chain 1: 900 -8335.087 0.181 0.036
Chain 1: 1000 -8566.929 0.166 0.036
Chain 1: 1100 -8523.968 0.066 0.027
Chain 1: 1200 -8425.796 0.034 0.019
Chain 1: 1300 -8381.534 0.017 0.014
Chain 1: 1400 -8401.779 0.013 0.012
Chain 1: 1500 -8494.663 0.011 0.012
Chain 1: 1600 -8411.355 0.012 0.012
Chain 1: 1700 -8376.993 0.011 0.011
Chain 1: 1800 -8349.213 0.010 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001571 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.71 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -50935.851 1.000 1.000
Chain 1: 200 -17025.791 1.496 1.992
Chain 1: 300 -9234.003 1.279 1.000
Chain 1: 400 -8465.446 0.982 1.000
Chain 1: 500 -8226.414 0.791 0.844
Chain 1: 600 -8377.079 0.662 0.844
Chain 1: 700 -8553.030 0.571 0.091
Chain 1: 800 -8651.705 0.501 0.091
Chain 1: 900 -8124.768 0.452 0.065
Chain 1: 1000 -7968.116 0.409 0.065
Chain 1: 1100 -8037.094 0.310 0.029
Chain 1: 1200 -7959.619 0.112 0.021
Chain 1: 1300 -8170.275 0.030 0.021
Chain 1: 1400 -8010.744 0.023 0.020
Chain 1: 1500 -7686.381 0.024 0.020
Chain 1: 1600 -7912.983 0.025 0.021
Chain 1: 1700 -7940.428 0.023 0.020
Chain 1: 1800 -7790.314 0.024 0.020
Chain 1: 1900 -7695.942 0.019 0.020
Chain 1: 2000 -7874.556 0.019 0.020
Chain 1: 2100 -7685.345 0.021 0.023
Chain 1: 2200 -8022.863 0.024 0.025
Chain 1: 2300 -7810.612 0.024 0.025
Chain 1: 2400 -7738.949 0.023 0.025
Chain 1: 2500 -7730.343 0.019 0.023
Chain 1: 2600 -7671.856 0.017 0.019
Chain 1: 2700 -7664.540 0.017 0.019
Chain 1: 2800 -7554.884 0.016 0.015
Chain 1: 2900 -7512.221 0.016 0.015
Chain 1: 3000 -7687.110 0.016 0.015
Chain 1: 3100 -7650.722 0.014 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003144 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.44 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86553.678 1.000 1.000
Chain 1: 200 -14377.870 3.010 5.020
Chain 1: 300 -10591.747 2.126 1.000
Chain 1: 400 -12412.446 1.631 1.000
Chain 1: 500 -9078.901 1.378 0.367
Chain 1: 600 -9084.172 1.149 0.367
Chain 1: 700 -9122.524 0.985 0.357
Chain 1: 800 -9259.585 0.864 0.357
Chain 1: 900 -9440.281 0.770 0.147
Chain 1: 1000 -9134.259 0.696 0.147
Chain 1: 1100 -9353.836 0.599 0.034 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8922.512 0.102 0.034
Chain 1: 1300 -9210.774 0.069 0.031
Chain 1: 1400 -9051.261 0.056 0.023
Chain 1: 1500 -9068.578 0.019 0.019
Chain 1: 1600 -9150.261 0.020 0.019
Chain 1: 1700 -9207.827 0.021 0.019
Chain 1: 1800 -8752.970 0.024 0.023
Chain 1: 1900 -8866.823 0.024 0.023
Chain 1: 2000 -8880.782 0.020 0.018
Chain 1: 2100 -8806.986 0.019 0.013
Chain 1: 2200 -8781.303 0.014 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003428 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.28 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8377613.337 1.000 1.000
Chain 1: 200 -1577323.517 2.656 4.311
Chain 1: 300 -890565.356 2.027 1.000
Chain 1: 400 -458110.491 1.757 1.000
Chain 1: 500 -359126.492 1.460 0.944
Chain 1: 600 -234307.653 1.306 0.944
Chain 1: 700 -120412.684 1.254 0.944
Chain 1: 800 -87575.954 1.144 0.944
Chain 1: 900 -67876.573 1.050 0.771
Chain 1: 1000 -52645.756 0.974 0.771
Chain 1: 1100 -40082.646 0.905 0.533 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39263.611 0.476 0.375
Chain 1: 1300 -27161.585 0.443 0.375
Chain 1: 1400 -26879.758 0.350 0.313
Chain 1: 1500 -23450.758 0.337 0.313
Chain 1: 1600 -22663.852 0.287 0.290
Chain 1: 1700 -21529.493 0.198 0.289
Chain 1: 1800 -21472.450 0.161 0.146
Chain 1: 1900 -21799.416 0.133 0.053
Chain 1: 2000 -20304.879 0.112 0.053
Chain 1: 2100 -20543.609 0.081 0.035
Chain 1: 2200 -20771.253 0.080 0.035
Chain 1: 2300 -20387.233 0.038 0.019
Chain 1: 2400 -20158.930 0.038 0.019
Chain 1: 2500 -19961.168 0.024 0.015
Chain 1: 2600 -19590.282 0.023 0.015
Chain 1: 2700 -19547.000 0.018 0.012
Chain 1: 2800 -19263.544 0.019 0.015
Chain 1: 2900 -19545.294 0.019 0.014
Chain 1: 3000 -19531.389 0.011 0.012
Chain 1: 3100 -19616.473 0.011 0.011
Chain 1: 3200 -19306.578 0.011 0.014
Chain 1: 3300 -19511.792 0.010 0.011
Chain 1: 3400 -18985.716 0.012 0.014
Chain 1: 3500 -19599.140 0.014 0.015
Chain 1: 3600 -18903.882 0.016 0.015
Chain 1: 3700 -19292.152 0.018 0.016
Chain 1: 3800 -18248.833 0.022 0.020
Chain 1: 3900 -18244.945 0.020 0.020
Chain 1: 4000 -18362.232 0.021 0.020
Chain 1: 4100 -18275.801 0.021 0.020
Chain 1: 4200 -18091.435 0.021 0.020
Chain 1: 4300 -18230.258 0.020 0.020
Chain 1: 4400 -18186.540 0.018 0.010
Chain 1: 4500 -18088.998 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001438 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.38 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48735.000 1.000 1.000
Chain 1: 200 -18915.507 1.288 1.576
Chain 1: 300 -12737.526 1.020 1.000
Chain 1: 400 -14161.905 0.791 1.000
Chain 1: 500 -18816.962 0.682 0.485
Chain 1: 600 -12631.917 0.650 0.490
Chain 1: 700 -13029.840 0.561 0.485
Chain 1: 800 -10791.752 0.517 0.485
Chain 1: 900 -10968.079 0.461 0.247
Chain 1: 1000 -13549.938 0.434 0.247
Chain 1: 1100 -10416.202 0.364 0.247
Chain 1: 1200 -10558.560 0.208 0.207
Chain 1: 1300 -11264.289 0.166 0.191
Chain 1: 1400 -16561.206 0.188 0.207
Chain 1: 1500 -11097.615 0.212 0.207
Chain 1: 1600 -13182.692 0.179 0.191
Chain 1: 1700 -21845.457 0.216 0.207
Chain 1: 1800 -10741.624 0.298 0.301
Chain 1: 1900 -9240.594 0.313 0.301
Chain 1: 2000 -9612.677 0.298 0.301
Chain 1: 2100 -10297.317 0.274 0.162
Chain 1: 2200 -9295.421 0.284 0.162
Chain 1: 2300 -12644.256 0.304 0.265
Chain 1: 2400 -9000.840 0.313 0.265
Chain 1: 2500 -15930.372 0.307 0.265
Chain 1: 2600 -9925.798 0.352 0.397
Chain 1: 2700 -9836.421 0.313 0.265
Chain 1: 2800 -15440.300 0.246 0.265
Chain 1: 2900 -9170.001 0.298 0.363
Chain 1: 3000 -10047.876 0.303 0.363
Chain 1: 3100 -8924.979 0.309 0.363
Chain 1: 3200 -8912.151 0.298 0.363
Chain 1: 3300 -8817.410 0.273 0.363
Chain 1: 3400 -8970.216 0.234 0.126
Chain 1: 3500 -9608.431 0.197 0.087
Chain 1: 3600 -9124.134 0.142 0.066
Chain 1: 3700 -16524.384 0.186 0.087
Chain 1: 3800 -15721.728 0.154 0.066
Chain 1: 3900 -9203.976 0.157 0.066
Chain 1: 4000 -11492.911 0.168 0.066
Chain 1: 4100 -9784.938 0.173 0.066
Chain 1: 4200 -14465.988 0.205 0.175
Chain 1: 4300 -12562.365 0.219 0.175
Chain 1: 4400 -8722.532 0.262 0.199
Chain 1: 4500 -9238.250 0.260 0.199
Chain 1: 4600 -8337.127 0.266 0.199
Chain 1: 4700 -9991.842 0.238 0.175
Chain 1: 4800 -8422.658 0.251 0.186
Chain 1: 4900 -8443.034 0.181 0.175
Chain 1: 5000 -10432.491 0.180 0.175
Chain 1: 5100 -8940.869 0.179 0.167
Chain 1: 5200 -9104.727 0.149 0.166
Chain 1: 5300 -12046.594 0.158 0.167
Chain 1: 5400 -10499.554 0.129 0.166
Chain 1: 5500 -10144.859 0.126 0.166
Chain 1: 5600 -8622.028 0.133 0.167
Chain 1: 5700 -9273.933 0.124 0.167
Chain 1: 5800 -8283.203 0.117 0.147
Chain 1: 5900 -8784.171 0.123 0.147
Chain 1: 6000 -9241.092 0.108 0.120
Chain 1: 6100 -8575.096 0.100 0.078
Chain 1: 6200 -8230.791 0.102 0.078
Chain 1: 6300 -8500.842 0.081 0.070
Chain 1: 6400 -9310.095 0.075 0.070
Chain 1: 6500 -8333.274 0.083 0.078
Chain 1: 6600 -8299.105 0.066 0.070
Chain 1: 6700 -8896.538 0.065 0.067
Chain 1: 6800 -9298.974 0.058 0.057
Chain 1: 6900 -9491.327 0.054 0.049
Chain 1: 7000 -10950.613 0.062 0.067
Chain 1: 7100 -8030.838 0.091 0.067
Chain 1: 7200 -12236.728 0.121 0.087
Chain 1: 7300 -10602.399 0.133 0.117
Chain 1: 7400 -8448.236 0.150 0.133
Chain 1: 7500 -9216.038 0.147 0.133
Chain 1: 7600 -8325.492 0.157 0.133
Chain 1: 7700 -8008.232 0.154 0.133
Chain 1: 7800 -8429.736 0.155 0.133
Chain 1: 7900 -8080.401 0.157 0.133
Chain 1: 8000 -12156.791 0.177 0.154
Chain 1: 8100 -8366.743 0.186 0.154
Chain 1: 8200 -10001.638 0.168 0.154
Chain 1: 8300 -8036.615 0.177 0.163
Chain 1: 8400 -7975.716 0.153 0.107
Chain 1: 8500 -8001.952 0.145 0.107
Chain 1: 8600 -10850.953 0.160 0.163
Chain 1: 8700 -10094.530 0.164 0.163
Chain 1: 8800 -8072.076 0.184 0.245
Chain 1: 8900 -11080.755 0.207 0.251
Chain 1: 9000 -8120.839 0.210 0.251
Chain 1: 9100 -8765.675 0.172 0.245
Chain 1: 9200 -8131.469 0.163 0.245
Chain 1: 9300 -8143.596 0.139 0.078
Chain 1: 9400 -9971.526 0.156 0.183
Chain 1: 9500 -7991.865 0.181 0.248
Chain 1: 9600 -8055.609 0.155 0.183
Chain 1: 9700 -9862.483 0.166 0.183
Chain 1: 9800 -10270.329 0.145 0.183
Chain 1: 9900 -8731.462 0.136 0.176
Chain 1: 10000 -8835.010 0.100 0.078
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001384 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.84 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -62940.519 1.000 1.000
Chain 1: 200 -17848.613 1.763 2.526
Chain 1: 300 -8621.136 1.532 1.070
Chain 1: 400 -8207.845 1.162 1.070
Chain 1: 500 -8198.183 0.930 1.000
Chain 1: 600 -8733.572 0.785 1.000
Chain 1: 700 -7874.830 0.688 0.109
Chain 1: 800 -7997.233 0.604 0.109
Chain 1: 900 -7896.406 0.539 0.061
Chain 1: 1000 -7581.033 0.489 0.061
Chain 1: 1100 -7607.087 0.389 0.050
Chain 1: 1200 -7599.416 0.137 0.042
Chain 1: 1300 -7689.635 0.031 0.015
Chain 1: 1400 -7625.376 0.027 0.013
Chain 1: 1500 -7599.771 0.027 0.013
Chain 1: 1600 -7513.281 0.022 0.012
Chain 1: 1700 -7491.246 0.011 0.012
Chain 1: 1800 -7535.855 0.010 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003439 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.39 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86240.975 1.000 1.000
Chain 1: 200 -13140.632 3.281 5.563
Chain 1: 300 -9580.591 2.312 1.000
Chain 1: 400 -10420.852 1.754 1.000
Chain 1: 500 -8490.680 1.448 0.372
Chain 1: 600 -8113.935 1.215 0.372
Chain 1: 700 -8267.308 1.044 0.227
Chain 1: 800 -8661.831 0.919 0.227
Chain 1: 900 -8446.352 0.820 0.081
Chain 1: 1000 -8168.142 0.741 0.081
Chain 1: 1100 -8321.917 0.643 0.046 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8084.468 0.090 0.046
Chain 1: 1300 -8306.148 0.055 0.034
Chain 1: 1400 -8294.477 0.047 0.029
Chain 1: 1500 -8200.923 0.026 0.027
Chain 1: 1600 -8297.446 0.022 0.026
Chain 1: 1700 -8385.716 0.021 0.026
Chain 1: 1800 -7994.838 0.022 0.026
Chain 1: 1900 -8097.312 0.021 0.018
Chain 1: 2000 -8067.286 0.017 0.013
Chain 1: 2100 -8196.091 0.017 0.013
Chain 1: 2200 -7982.601 0.017 0.013
Chain 1: 2300 -8126.138 0.016 0.013
Chain 1: 2400 -8140.318 0.016 0.013
Chain 1: 2500 -8107.379 0.015 0.013
Chain 1: 2600 -8108.574 0.014 0.013
Chain 1: 2700 -8015.961 0.014 0.013
Chain 1: 2800 -7990.327 0.010 0.012 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002939 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.39 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8417971.474 1.000 1.000
Chain 1: 200 -1583602.107 2.658 4.316
Chain 1: 300 -889595.476 2.032 1.000
Chain 1: 400 -457117.188 1.760 1.000
Chain 1: 500 -357342.161 1.464 0.946
Chain 1: 600 -232405.851 1.310 0.946
Chain 1: 700 -118713.222 1.259 0.946
Chain 1: 800 -85979.958 1.150 0.946
Chain 1: 900 -66335.619 1.055 0.780
Chain 1: 1000 -51141.532 0.979 0.780
Chain 1: 1100 -38637.071 0.911 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37811.474 0.482 0.381
Chain 1: 1300 -25790.672 0.451 0.381
Chain 1: 1400 -25510.478 0.357 0.324
Chain 1: 1500 -22104.946 0.345 0.324
Chain 1: 1600 -21323.241 0.294 0.297
Chain 1: 1700 -20199.859 0.204 0.296
Chain 1: 1800 -20144.634 0.166 0.154
Chain 1: 1900 -20470.317 0.138 0.056
Chain 1: 2000 -18984.242 0.117 0.056
Chain 1: 2100 -19222.219 0.085 0.037
Chain 1: 2200 -19448.241 0.084 0.037
Chain 1: 2300 -19066.031 0.040 0.020
Chain 1: 2400 -18838.327 0.040 0.020
Chain 1: 2500 -18640.404 0.026 0.016
Chain 1: 2600 -18270.934 0.024 0.016
Chain 1: 2700 -18228.085 0.019 0.012
Chain 1: 2800 -17945.155 0.020 0.016
Chain 1: 2900 -18226.223 0.020 0.015
Chain 1: 3000 -18212.357 0.012 0.012
Chain 1: 3100 -18297.274 0.011 0.012
Chain 1: 3200 -17988.254 0.012 0.015
Chain 1: 3300 -18192.786 0.011 0.012
Chain 1: 3400 -17668.257 0.013 0.015
Chain 1: 3500 -18279.241 0.015 0.016
Chain 1: 3600 -17587.123 0.017 0.016
Chain 1: 3700 -17973.021 0.019 0.017
Chain 1: 3800 -16934.526 0.023 0.021
Chain 1: 3900 -16930.739 0.022 0.021
Chain 1: 4000 -17048.039 0.023 0.021
Chain 1: 4100 -16961.857 0.023 0.021
Chain 1: 4200 -16778.533 0.022 0.021
Chain 1: 4300 -16916.605 0.022 0.021
Chain 1: 4400 -16873.741 0.019 0.011
Chain 1: 4500 -16776.363 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001296 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.96 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48830.719 1.000 1.000
Chain 1: 200 -20109.854 1.214 1.428
Chain 1: 300 -20237.589 0.812 1.000
Chain 1: 400 -17336.391 0.650 1.000
Chain 1: 500 -19521.626 0.543 0.167
Chain 1: 600 -18462.418 0.462 0.167
Chain 1: 700 -15146.106 0.427 0.167
Chain 1: 800 -10876.809 0.423 0.219
Chain 1: 900 -11245.021 0.379 0.167
Chain 1: 1000 -14049.329 0.361 0.200
Chain 1: 1100 -12258.908 0.276 0.167
Chain 1: 1200 -11304.830 0.142 0.146
Chain 1: 1300 -14958.695 0.166 0.167
Chain 1: 1400 -11703.345 0.177 0.200
Chain 1: 1500 -10003.865 0.182 0.200
Chain 1: 1600 -14093.090 0.206 0.219
Chain 1: 1700 -11544.122 0.206 0.221
Chain 1: 1800 -11422.091 0.168 0.200
Chain 1: 1900 -14242.046 0.184 0.200
Chain 1: 2000 -15699.364 0.174 0.198
Chain 1: 2100 -15012.029 0.163 0.198
Chain 1: 2200 -9502.912 0.213 0.221
Chain 1: 2300 -10117.445 0.195 0.198
Chain 1: 2400 -14360.433 0.196 0.198
Chain 1: 2500 -9584.913 0.229 0.221
Chain 1: 2600 -15931.164 0.240 0.221
Chain 1: 2700 -8717.270 0.301 0.295
Chain 1: 2800 -10012.744 0.313 0.295
Chain 1: 2900 -9209.278 0.302 0.295
Chain 1: 3000 -8880.379 0.296 0.295
Chain 1: 3100 -15371.953 0.334 0.398
Chain 1: 3200 -8870.137 0.349 0.398
Chain 1: 3300 -9567.361 0.350 0.398
Chain 1: 3400 -10295.783 0.328 0.398
Chain 1: 3500 -11310.539 0.287 0.129
Chain 1: 3600 -9247.879 0.269 0.129
Chain 1: 3700 -8614.554 0.194 0.090
Chain 1: 3800 -9901.364 0.194 0.090
Chain 1: 3900 -9192.554 0.193 0.090
Chain 1: 4000 -8621.791 0.196 0.090
Chain 1: 4100 -8909.509 0.157 0.077
Chain 1: 4200 -10590.239 0.099 0.077
Chain 1: 4300 -8590.077 0.115 0.090
Chain 1: 4400 -12835.225 0.141 0.130
Chain 1: 4500 -8637.747 0.181 0.159
Chain 1: 4600 -9516.888 0.168 0.130
Chain 1: 4700 -8428.042 0.174 0.130
Chain 1: 4800 -8645.281 0.163 0.129
Chain 1: 4900 -9072.919 0.160 0.129
Chain 1: 5000 -14344.793 0.190 0.159
Chain 1: 5100 -16334.527 0.199 0.159
Chain 1: 5200 -15696.299 0.187 0.129
Chain 1: 5300 -9444.036 0.230 0.129
Chain 1: 5400 -8508.164 0.208 0.122
Chain 1: 5500 -8534.214 0.160 0.110
Chain 1: 5600 -13370.515 0.187 0.122
Chain 1: 5700 -9473.852 0.215 0.122
Chain 1: 5800 -9034.404 0.217 0.122
Chain 1: 5900 -9183.133 0.214 0.122
Chain 1: 6000 -8782.023 0.182 0.110
Chain 1: 6100 -11066.512 0.191 0.110
Chain 1: 6200 -8223.930 0.221 0.206
Chain 1: 6300 -11663.943 0.184 0.206
Chain 1: 6400 -8334.980 0.213 0.295
Chain 1: 6500 -8671.467 0.217 0.295
Chain 1: 6600 -8345.999 0.185 0.206
Chain 1: 6700 -8603.935 0.146 0.049
Chain 1: 6800 -8243.556 0.146 0.046
Chain 1: 6900 -8173.692 0.145 0.046
Chain 1: 7000 -8638.352 0.146 0.054
Chain 1: 7100 -8542.330 0.127 0.044
Chain 1: 7200 -8187.272 0.096 0.043
Chain 1: 7300 -8952.965 0.075 0.043
Chain 1: 7400 -8181.786 0.045 0.043
Chain 1: 7500 -10039.491 0.059 0.044
Chain 1: 7600 -8650.749 0.072 0.054
Chain 1: 7700 -8482.754 0.071 0.054
Chain 1: 7800 -8318.830 0.068 0.054
Chain 1: 7900 -10080.389 0.085 0.086
Chain 1: 8000 -10546.789 0.084 0.086
Chain 1: 8100 -8389.664 0.108 0.094
Chain 1: 8200 -7957.224 0.110 0.094
Chain 1: 8300 -10707.808 0.127 0.161
Chain 1: 8400 -11183.304 0.121 0.161
Chain 1: 8500 -8010.075 0.143 0.161
Chain 1: 8600 -9790.377 0.145 0.175
Chain 1: 8700 -10065.632 0.145 0.175
Chain 1: 8800 -8022.659 0.169 0.182
Chain 1: 8900 -9536.674 0.167 0.182
Chain 1: 9000 -8269.644 0.178 0.182
Chain 1: 9100 -8632.002 0.157 0.159
Chain 1: 9200 -8917.891 0.155 0.159
Chain 1: 9300 -8821.601 0.130 0.153
Chain 1: 9400 -8454.820 0.130 0.153
Chain 1: 9500 -8361.039 0.092 0.043
Chain 1: 9600 -8139.157 0.076 0.042
Chain 1: 9700 -8118.698 0.074 0.042
Chain 1: 9800 -10286.744 0.069 0.042
Chain 1: 9900 -10890.842 0.059 0.042
Chain 1: 10000 -8358.093 0.074 0.042
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001455 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.55 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61438.434 1.000 1.000
Chain 1: 200 -17563.367 1.749 2.498
Chain 1: 300 -8739.702 1.503 1.010
Chain 1: 400 -9313.176 1.142 1.010
Chain 1: 500 -8386.523 0.936 1.000
Chain 1: 600 -9044.141 0.792 1.000
Chain 1: 700 -7777.192 0.702 0.163
Chain 1: 800 -8114.129 0.620 0.163
Chain 1: 900 -7799.225 0.555 0.110
Chain 1: 1000 -7867.011 0.501 0.110
Chain 1: 1100 -7645.194 0.403 0.073
Chain 1: 1200 -7560.425 0.155 0.062
Chain 1: 1300 -7783.094 0.057 0.042
Chain 1: 1400 -7693.171 0.052 0.040
Chain 1: 1500 -7625.875 0.042 0.029
Chain 1: 1600 -7559.357 0.035 0.029
Chain 1: 1700 -7544.791 0.019 0.012
Chain 1: 1800 -7610.775 0.016 0.011
Chain 1: 1900 -7502.623 0.013 0.011
Chain 1: 2000 -7589.090 0.013 0.011
Chain 1: 2100 -7678.416 0.012 0.011
Chain 1: 2200 -7697.275 0.011 0.011
Chain 1: 2300 -7591.122 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003403 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.03 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85896.369 1.000 1.000
Chain 1: 200 -13238.284 3.244 5.488
Chain 1: 300 -9695.276 2.285 1.000
Chain 1: 400 -10541.539 1.734 1.000
Chain 1: 500 -8588.841 1.432 0.365
Chain 1: 600 -8239.345 1.201 0.365
Chain 1: 700 -8337.766 1.031 0.227
Chain 1: 800 -8955.457 0.911 0.227
Chain 1: 900 -8474.306 0.816 0.080
Chain 1: 1000 -8387.403 0.735 0.080
Chain 1: 1100 -8590.277 0.638 0.069 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8299.231 0.092 0.057
Chain 1: 1300 -8241.261 0.056 0.042
Chain 1: 1400 -8304.856 0.049 0.035
Chain 1: 1500 -8287.954 0.027 0.024
Chain 1: 1600 -8286.905 0.022 0.012
Chain 1: 1700 -8218.003 0.022 0.010
Chain 1: 1800 -8099.823 0.017 0.010
Chain 1: 1900 -8218.238 0.012 0.010
Chain 1: 2000 -8178.092 0.012 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003461 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.61 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8410658.041 1.000 1.000
Chain 1: 200 -1582865.650 2.657 4.314
Chain 1: 300 -889389.795 2.031 1.000
Chain 1: 400 -456405.840 1.760 1.000
Chain 1: 500 -356751.462 1.464 0.949
Chain 1: 600 -232007.068 1.310 0.949
Chain 1: 700 -118622.957 1.259 0.949
Chain 1: 800 -85919.453 1.149 0.949
Chain 1: 900 -66329.182 1.055 0.780
Chain 1: 1000 -51170.615 0.979 0.780
Chain 1: 1100 -38693.817 0.911 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37873.040 0.482 0.381
Chain 1: 1300 -25879.527 0.450 0.381
Chain 1: 1400 -25602.146 0.356 0.322
Chain 1: 1500 -22202.066 0.344 0.322
Chain 1: 1600 -21422.146 0.294 0.296
Chain 1: 1700 -20302.064 0.204 0.295
Chain 1: 1800 -20247.607 0.166 0.153
Chain 1: 1900 -20573.270 0.138 0.055
Chain 1: 2000 -19088.649 0.116 0.055
Chain 1: 2100 -19326.742 0.085 0.036
Chain 1: 2200 -19552.334 0.084 0.036
Chain 1: 2300 -19170.446 0.040 0.020
Chain 1: 2400 -18942.772 0.040 0.020
Chain 1: 2500 -18744.600 0.025 0.016
Chain 1: 2600 -18375.450 0.024 0.016
Chain 1: 2700 -18332.703 0.019 0.012
Chain 1: 2800 -18049.676 0.020 0.016
Chain 1: 2900 -18330.615 0.020 0.015
Chain 1: 3000 -18316.910 0.012 0.012
Chain 1: 3100 -18401.791 0.011 0.012
Chain 1: 3200 -18092.853 0.012 0.015
Chain 1: 3300 -18297.305 0.011 0.012
Chain 1: 3400 -17772.827 0.013 0.015
Chain 1: 3500 -18383.735 0.015 0.016
Chain 1: 3600 -17691.672 0.017 0.016
Chain 1: 3700 -18077.467 0.019 0.017
Chain 1: 3800 -17039.115 0.023 0.021
Chain 1: 3900 -17035.298 0.022 0.021
Chain 1: 4000 -17152.602 0.022 0.021
Chain 1: 4100 -17066.421 0.022 0.021
Chain 1: 4200 -16883.161 0.022 0.021
Chain 1: 4300 -17021.246 0.022 0.021
Chain 1: 4400 -16978.407 0.019 0.011
Chain 1: 4500 -16881.000 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001399 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.99 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49556.152 1.000 1.000
Chain 1: 200 -16344.436 1.516 2.032
Chain 1: 300 -21293.539 1.088 1.000
Chain 1: 400 -14781.877 0.926 1.000
Chain 1: 500 -15476.679 0.750 0.441
Chain 1: 600 -14025.760 0.642 0.441
Chain 1: 700 -13728.712 0.554 0.232
Chain 1: 800 -14283.931 0.489 0.232
Chain 1: 900 -16928.228 0.452 0.156
Chain 1: 1000 -23493.615 0.435 0.232
Chain 1: 1100 -10986.717 0.449 0.232
Chain 1: 1200 -11081.415 0.246 0.156
Chain 1: 1300 -20112.532 0.268 0.156
Chain 1: 1400 -16297.183 0.247 0.156
Chain 1: 1500 -18200.500 0.253 0.156
Chain 1: 1600 -10860.828 0.311 0.234
Chain 1: 1700 -22473.391 0.360 0.279
Chain 1: 1800 -16846.539 0.390 0.334
Chain 1: 1900 -10449.303 0.435 0.449
Chain 1: 2000 -21795.363 0.459 0.517 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2100 -10894.155 0.446 0.517 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2200 -14150.361 0.468 0.517 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2300 -10260.144 0.461 0.517 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2400 -10511.103 0.440 0.517 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2500 -10561.578 0.430 0.517 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2600 -10142.822 0.366 0.379
Chain 1: 2700 -16916.517 0.355 0.379
Chain 1: 2800 -11101.923 0.374 0.400
Chain 1: 2900 -14238.862 0.334 0.379
Chain 1: 3000 -9856.163 0.327 0.379
Chain 1: 3100 -10263.603 0.231 0.230
Chain 1: 3200 -12057.596 0.223 0.220
Chain 1: 3300 -9872.789 0.207 0.220
Chain 1: 3400 -9557.900 0.208 0.220
Chain 1: 3500 -10443.662 0.216 0.220
Chain 1: 3600 -9958.458 0.217 0.220
Chain 1: 3700 -10253.298 0.179 0.149
Chain 1: 3800 -14836.548 0.158 0.149
Chain 1: 3900 -10625.542 0.175 0.149
Chain 1: 4000 -11523.516 0.139 0.085
Chain 1: 4100 -11759.928 0.137 0.085
Chain 1: 4200 -12929.637 0.131 0.085
Chain 1: 4300 -9406.797 0.146 0.085
Chain 1: 4400 -9146.240 0.146 0.085
Chain 1: 4500 -9199.610 0.138 0.078
Chain 1: 4600 -9938.848 0.141 0.078
Chain 1: 4700 -10735.380 0.145 0.078
Chain 1: 4800 -9049.048 0.133 0.078
Chain 1: 4900 -9487.038 0.098 0.074
Chain 1: 5000 -11115.383 0.105 0.074
Chain 1: 5100 -11477.298 0.106 0.074
Chain 1: 5200 -17774.395 0.132 0.074
Chain 1: 5300 -13510.731 0.126 0.074
Chain 1: 5400 -9380.981 0.168 0.146
Chain 1: 5500 -8887.920 0.172 0.146
Chain 1: 5600 -8966.554 0.166 0.146
Chain 1: 5700 -13127.668 0.190 0.186
Chain 1: 5800 -9293.978 0.213 0.316
Chain 1: 5900 -13946.552 0.242 0.317
Chain 1: 6000 -9665.304 0.271 0.334
Chain 1: 6100 -8860.751 0.277 0.334
Chain 1: 6200 -9170.309 0.245 0.317
Chain 1: 6300 -12438.418 0.240 0.317
Chain 1: 6400 -13204.380 0.202 0.263
Chain 1: 6500 -9357.198 0.237 0.317
Chain 1: 6600 -10279.371 0.245 0.317
Chain 1: 6700 -8628.261 0.233 0.263
Chain 1: 6800 -10727.996 0.211 0.196
Chain 1: 6900 -15701.670 0.209 0.196
Chain 1: 7000 -10036.609 0.221 0.196
Chain 1: 7100 -11061.188 0.222 0.196
Chain 1: 7200 -10943.625 0.219 0.196
Chain 1: 7300 -11631.866 0.199 0.191
Chain 1: 7400 -8738.198 0.226 0.196
Chain 1: 7500 -12153.651 0.213 0.196
Chain 1: 7600 -9514.511 0.232 0.277
Chain 1: 7700 -9363.340 0.215 0.277
Chain 1: 7800 -13767.847 0.227 0.281
Chain 1: 7900 -8834.347 0.251 0.281
Chain 1: 8000 -12015.063 0.221 0.277
Chain 1: 8100 -9967.512 0.232 0.277
Chain 1: 8200 -8738.996 0.245 0.277
Chain 1: 8300 -8783.226 0.240 0.277
Chain 1: 8400 -9090.492 0.210 0.265
Chain 1: 8500 -8754.267 0.186 0.205
Chain 1: 8600 -12001.100 0.185 0.205
Chain 1: 8700 -10958.562 0.193 0.205
Chain 1: 8800 -8933.802 0.184 0.205
Chain 1: 8900 -11002.990 0.147 0.188
Chain 1: 9000 -13038.561 0.136 0.156
Chain 1: 9100 -9840.717 0.148 0.156
Chain 1: 9200 -9749.376 0.135 0.156
Chain 1: 9300 -8531.492 0.149 0.156
Chain 1: 9400 -10089.165 0.161 0.156
Chain 1: 9500 -9699.345 0.161 0.156
Chain 1: 9600 -8665.342 0.146 0.154
Chain 1: 9700 -8720.601 0.137 0.154
Chain 1: 9800 -8789.703 0.115 0.143
Chain 1: 9900 -9420.655 0.103 0.119
Chain 1: 10000 -8815.619 0.094 0.069
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001431 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.31 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -62320.132 1.000 1.000
Chain 1: 200 -18620.366 1.673 2.347
Chain 1: 300 -9238.036 1.454 1.016
Chain 1: 400 -8891.401 1.100 1.016
Chain 1: 500 -8381.106 0.892 1.000
Chain 1: 600 -9149.908 0.758 1.000
Chain 1: 700 -8332.820 0.663 0.098
Chain 1: 800 -8324.088 0.581 0.098
Chain 1: 900 -7958.619 0.521 0.084
Chain 1: 1000 -7851.812 0.471 0.084
Chain 1: 1100 -7767.696 0.372 0.061
Chain 1: 1200 -7762.349 0.137 0.046
Chain 1: 1300 -7870.597 0.037 0.039
Chain 1: 1400 -7856.348 0.033 0.014
Chain 1: 1500 -7560.493 0.031 0.014
Chain 1: 1600 -7912.683 0.027 0.014
Chain 1: 1700 -7476.113 0.023 0.014
Chain 1: 1800 -7619.190 0.025 0.019
Chain 1: 1900 -7549.279 0.021 0.014
Chain 1: 2000 -7766.327 0.023 0.019
Chain 1: 2100 -7538.994 0.024 0.028
Chain 1: 2200 -7821.285 0.028 0.030
Chain 1: 2300 -7665.908 0.029 0.030
Chain 1: 2400 -7672.065 0.029 0.030
Chain 1: 2500 -7606.068 0.025 0.028
Chain 1: 2600 -7569.021 0.022 0.020
Chain 1: 2700 -7567.273 0.016 0.019
Chain 1: 2800 -7693.800 0.015 0.016
Chain 1: 2900 -7395.017 0.019 0.020
Chain 1: 3000 -7536.498 0.018 0.019
Chain 1: 3100 -7550.080 0.015 0.016
Chain 1: 3200 -7664.247 0.013 0.015
Chain 1: 3300 -7437.774 0.014 0.015
Chain 1: 3400 -7696.066 0.017 0.016
Chain 1: 3500 -7479.549 0.019 0.019
Chain 1: 3600 -7528.497 0.019 0.019
Chain 1: 3700 -7473.068 0.020 0.019
Chain 1: 3800 -7457.397 0.018 0.019
Chain 1: 3900 -7436.065 0.015 0.015
Chain 1: 4000 -7423.504 0.013 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002967 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.67 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86313.575 1.000 1.000
Chain 1: 200 -14258.503 3.027 5.053
Chain 1: 300 -10467.517 2.139 1.000
Chain 1: 400 -12498.935 1.645 1.000
Chain 1: 500 -8902.247 1.396 0.404
Chain 1: 600 -8955.222 1.165 0.404
Chain 1: 700 -8770.609 1.001 0.362
Chain 1: 800 -9105.845 0.881 0.362
Chain 1: 900 -9107.402 0.783 0.163
Chain 1: 1000 -9446.224 0.708 0.163
Chain 1: 1100 -9233.750 0.611 0.037 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8733.491 0.111 0.037
Chain 1: 1300 -9081.533 0.078 0.037
Chain 1: 1400 -9015.660 0.063 0.036
Chain 1: 1500 -8959.767 0.023 0.023
Chain 1: 1600 -9034.625 0.023 0.023
Chain 1: 1700 -9084.703 0.022 0.023
Chain 1: 1800 -8625.510 0.024 0.023
Chain 1: 1900 -8736.984 0.025 0.023
Chain 1: 2000 -8751.687 0.021 0.013
Chain 1: 2100 -8846.986 0.020 0.011
Chain 1: 2200 -8626.945 0.017 0.011
Chain 1: 2300 -8827.268 0.015 0.011
Chain 1: 2400 -8634.091 0.017 0.013
Chain 1: 2500 -8708.078 0.017 0.013
Chain 1: 2600 -8618.463 0.017 0.013
Chain 1: 2700 -8651.614 0.017 0.013
Chain 1: 2800 -8602.381 0.012 0.011
Chain 1: 2900 -8717.009 0.012 0.011
Chain 1: 3000 -8633.104 0.013 0.011
Chain 1: 3100 -8594.710 0.013 0.010
Chain 1: 3200 -8567.030 0.010 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003582 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.82 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8374443.937 1.000 1.000
Chain 1: 200 -1579334.845 2.651 4.303
Chain 1: 300 -891783.283 2.024 1.000
Chain 1: 400 -459054.755 1.754 1.000
Chain 1: 500 -359886.026 1.458 0.943
Chain 1: 600 -234796.275 1.304 0.943
Chain 1: 700 -120544.869 1.253 0.943
Chain 1: 800 -87642.035 1.143 0.943
Chain 1: 900 -67891.232 1.049 0.771
Chain 1: 1000 -52619.297 0.973 0.771
Chain 1: 1100 -40022.018 0.904 0.533 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39198.392 0.476 0.375
Chain 1: 1300 -27059.507 0.444 0.375
Chain 1: 1400 -26774.390 0.351 0.315
Chain 1: 1500 -23336.076 0.338 0.315
Chain 1: 1600 -22546.645 0.288 0.291
Chain 1: 1700 -21407.598 0.199 0.290
Chain 1: 1800 -21349.458 0.161 0.147
Chain 1: 1900 -21676.412 0.134 0.053
Chain 1: 2000 -20179.597 0.112 0.053
Chain 1: 2100 -20418.386 0.082 0.035
Chain 1: 2200 -20646.505 0.081 0.035
Chain 1: 2300 -20262.042 0.038 0.019
Chain 1: 2400 -20033.717 0.038 0.019
Chain 1: 2500 -19836.213 0.024 0.015
Chain 1: 2600 -19465.100 0.023 0.015
Chain 1: 2700 -19421.706 0.018 0.012
Chain 1: 2800 -19138.375 0.019 0.015
Chain 1: 2900 -19420.154 0.019 0.015
Chain 1: 3000 -19406.188 0.011 0.012
Chain 1: 3100 -19491.315 0.011 0.011
Chain 1: 3200 -19181.344 0.011 0.015
Chain 1: 3300 -19386.609 0.010 0.011
Chain 1: 3400 -18860.532 0.012 0.015
Chain 1: 3500 -19474.069 0.014 0.015
Chain 1: 3600 -18778.654 0.016 0.015
Chain 1: 3700 -19167.064 0.018 0.016
Chain 1: 3800 -18123.625 0.022 0.020
Chain 1: 3900 -18119.769 0.021 0.020
Chain 1: 4000 -18236.992 0.021 0.020
Chain 1: 4100 -18150.615 0.021 0.020
Chain 1: 4200 -17966.197 0.021 0.020
Chain 1: 4300 -18105.011 0.020 0.020
Chain 1: 4400 -18061.268 0.018 0.010
Chain 1: 4500 -17963.766 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001342 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.42 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12603.006 1.000 1.000
Chain 1: 200 -9183.158 0.686 1.000
Chain 1: 300 -7961.127 0.509 0.372
Chain 1: 400 -8080.881 0.385 0.372
Chain 1: 500 -8053.771 0.309 0.153
Chain 1: 600 -7933.969 0.260 0.153
Chain 1: 700 -7821.817 0.225 0.015
Chain 1: 800 -7822.077 0.197 0.015
Chain 1: 900 -7730.820 0.176 0.015
Chain 1: 1000 -7945.068 0.161 0.015
Chain 1: 1100 -7982.618 0.062 0.015
Chain 1: 1200 -7869.497 0.026 0.014
Chain 1: 1300 -7797.165 0.011 0.014
Chain 1: 1400 -7813.829 0.010 0.012
Chain 1: 1500 -7903.759 0.011 0.012
Chain 1: 1600 -7836.803 0.010 0.011
Chain 1: 1700 -7794.367 0.009 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.0014 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -59921.249 1.000 1.000
Chain 1: 200 -18206.662 1.646 2.291
Chain 1: 300 -8790.153 1.454 1.071
Chain 1: 400 -8739.692 1.092 1.071
Chain 1: 500 -8450.987 0.880 1.000
Chain 1: 600 -8697.417 0.738 1.000
Chain 1: 700 -8015.701 0.645 0.085
Chain 1: 800 -8574.614 0.573 0.085
Chain 1: 900 -7951.753 0.518 0.078
Chain 1: 1000 -7599.296 0.471 0.078
Chain 1: 1100 -7845.037 0.374 0.065
Chain 1: 1200 -7721.305 0.146 0.046
Chain 1: 1300 -7808.194 0.040 0.034
Chain 1: 1400 -7677.161 0.041 0.034
Chain 1: 1500 -7553.376 0.040 0.031
Chain 1: 1600 -7777.281 0.040 0.031
Chain 1: 1700 -7429.740 0.036 0.031
Chain 1: 1800 -7584.990 0.031 0.029
Chain 1: 1900 -7532.321 0.024 0.020
Chain 1: 2000 -7622.766 0.021 0.017
Chain 1: 2100 -7565.668 0.018 0.016
Chain 1: 2200 -7729.887 0.019 0.017
Chain 1: 2300 -7553.591 0.020 0.020
Chain 1: 2400 -7635.018 0.019 0.020
Chain 1: 2500 -7541.382 0.019 0.020
Chain 1: 2600 -7503.619 0.017 0.012
Chain 1: 2700 -7464.430 0.012 0.012
Chain 1: 2800 -7477.370 0.011 0.011
Chain 1: 2900 -7359.124 0.012 0.012
Chain 1: 3000 -7499.290 0.012 0.012
Chain 1: 3100 -7495.545 0.011 0.012
Chain 1: 3200 -7699.049 0.012 0.012
Chain 1: 3300 -7426.643 0.013 0.012
Chain 1: 3400 -7639.962 0.015 0.016
Chain 1: 3500 -7409.432 0.017 0.019
Chain 1: 3600 -7474.695 0.017 0.019
Chain 1: 3700 -7423.365 0.017 0.019
Chain 1: 3800 -7425.934 0.017 0.019
Chain 1: 3900 -7391.820 0.016 0.019
Chain 1: 4000 -7386.860 0.014 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003773 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 37.73 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86512.094 1.000 1.000
Chain 1: 200 -13634.189 3.173 5.345
Chain 1: 300 -9907.382 2.240 1.000
Chain 1: 400 -11361.683 1.712 1.000
Chain 1: 500 -8703.713 1.431 0.376
Chain 1: 600 -8287.052 1.201 0.376
Chain 1: 700 -8329.444 1.030 0.305
Chain 1: 800 -8585.620 0.905 0.305
Chain 1: 900 -8658.585 0.805 0.128
Chain 1: 1000 -8542.066 0.726 0.128
Chain 1: 1100 -8679.737 0.628 0.050 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8274.782 0.098 0.049
Chain 1: 1300 -8560.491 0.064 0.033
Chain 1: 1400 -8539.062 0.051 0.030
Chain 1: 1500 -8417.491 0.022 0.016
Chain 1: 1600 -8532.703 0.019 0.014
Chain 1: 1700 -8592.491 0.019 0.014
Chain 1: 1800 -8154.648 0.021 0.014
Chain 1: 1900 -8258.718 0.022 0.014
Chain 1: 2000 -8236.878 0.020 0.014
Chain 1: 2100 -8207.138 0.019 0.014
Chain 1: 2200 -8175.921 0.015 0.013
Chain 1: 2300 -8313.343 0.013 0.013
Chain 1: 2400 -8157.040 0.015 0.014
Chain 1: 2500 -8228.110 0.014 0.013
Chain 1: 2600 -8141.383 0.014 0.011
Chain 1: 2700 -8178.196 0.014 0.011
Chain 1: 2800 -8136.421 0.009 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003571 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.71 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8425579.577 1.000 1.000
Chain 1: 200 -1586446.337 2.655 4.311
Chain 1: 300 -891697.516 2.030 1.000
Chain 1: 400 -458191.880 1.759 1.000
Chain 1: 500 -358288.231 1.463 0.946
Chain 1: 600 -233158.520 1.309 0.946
Chain 1: 700 -119374.872 1.258 0.946
Chain 1: 800 -86595.782 1.148 0.946
Chain 1: 900 -66934.598 1.053 0.779
Chain 1: 1000 -51736.881 0.977 0.779
Chain 1: 1100 -39220.642 0.909 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38399.934 0.480 0.379
Chain 1: 1300 -26354.929 0.448 0.379
Chain 1: 1400 -26075.776 0.354 0.319
Chain 1: 1500 -22662.944 0.341 0.319
Chain 1: 1600 -21880.413 0.291 0.294
Chain 1: 1700 -20753.579 0.201 0.294
Chain 1: 1800 -20697.972 0.164 0.151
Chain 1: 1900 -21024.507 0.136 0.054
Chain 1: 2000 -19534.840 0.114 0.054
Chain 1: 2100 -19773.122 0.084 0.036
Chain 1: 2200 -19999.997 0.083 0.036
Chain 1: 2300 -19616.785 0.039 0.020
Chain 1: 2400 -19388.739 0.039 0.020
Chain 1: 2500 -19190.798 0.025 0.016
Chain 1: 2600 -18820.426 0.023 0.016
Chain 1: 2700 -18777.298 0.018 0.012
Chain 1: 2800 -18493.975 0.019 0.015
Chain 1: 2900 -18775.433 0.019 0.015
Chain 1: 3000 -18761.535 0.012 0.012
Chain 1: 3100 -18846.598 0.011 0.012
Chain 1: 3200 -18536.968 0.012 0.015
Chain 1: 3300 -18741.963 0.011 0.012
Chain 1: 3400 -18216.328 0.012 0.015
Chain 1: 3500 -18828.993 0.015 0.015
Chain 1: 3600 -18134.666 0.017 0.015
Chain 1: 3700 -18522.212 0.018 0.017
Chain 1: 3800 -17480.322 0.023 0.021
Chain 1: 3900 -17476.441 0.021 0.021
Chain 1: 4000 -17593.744 0.022 0.021
Chain 1: 4100 -17507.404 0.022 0.021
Chain 1: 4200 -17323.328 0.021 0.021
Chain 1: 4300 -17461.940 0.021 0.021
Chain 1: 4400 -17418.474 0.018 0.011
Chain 1: 4500 -17320.974 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001253 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.53 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13601.860 1.000 1.000
Chain 1: 200 -10003.497 0.680 1.000
Chain 1: 300 -8801.453 0.499 0.360
Chain 1: 400 -8517.813 0.382 0.360
Chain 1: 500 -8212.525 0.313 0.137
Chain 1: 600 -8264.270 0.262 0.137
Chain 1: 700 -8146.578 0.227 0.037
Chain 1: 800 -8128.238 0.199 0.037
Chain 1: 900 -8155.171 0.177 0.033
Chain 1: 1000 -8226.512 0.160 0.033
Chain 1: 1100 -8480.745 0.063 0.030
Chain 1: 1200 -8199.528 0.031 0.030
Chain 1: 1300 -8116.371 0.018 0.014
Chain 1: 1400 -8139.102 0.015 0.010
Chain 1: 1500 -8250.789 0.013 0.010
Chain 1: 1600 -8137.251 0.013 0.014
Chain 1: 1700 -8115.925 0.012 0.010
Chain 1: 1800 -8088.122 0.012 0.010
Chain 1: 1900 -8115.981 0.012 0.010
Chain 1: 2000 -8048.784 0.012 0.010
Chain 1: 2100 -8056.673 0.009 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.0014 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -63974.697 1.000 1.000
Chain 1: 200 -18898.900 1.693 2.385
Chain 1: 300 -9119.474 1.486 1.072
Chain 1: 400 -8101.337 1.146 1.072
Chain 1: 500 -8367.258 0.923 1.000
Chain 1: 600 -9001.580 0.781 1.000
Chain 1: 700 -7859.167 0.690 0.145
Chain 1: 800 -8780.952 0.617 0.145
Chain 1: 900 -8312.794 0.555 0.126
Chain 1: 1000 -7834.964 0.505 0.126
Chain 1: 1100 -7757.673 0.406 0.105
Chain 1: 1200 -7856.117 0.169 0.070
Chain 1: 1300 -7948.296 0.063 0.061
Chain 1: 1400 -8050.247 0.052 0.056
Chain 1: 1500 -7601.095 0.054 0.059
Chain 1: 1600 -7872.064 0.051 0.056
Chain 1: 1700 -7692.587 0.039 0.034
Chain 1: 1800 -7812.673 0.030 0.023
Chain 1: 1900 -7708.656 0.025 0.015
Chain 1: 2000 -7798.811 0.020 0.013
Chain 1: 2100 -7672.960 0.021 0.015
Chain 1: 2200 -7850.044 0.022 0.016
Chain 1: 2300 -7665.391 0.023 0.023
Chain 1: 2400 -7718.472 0.023 0.023
Chain 1: 2500 -7753.503 0.017 0.016
Chain 1: 2600 -7631.662 0.015 0.016
Chain 1: 2700 -7697.619 0.014 0.015
Chain 1: 2800 -7625.563 0.013 0.013
Chain 1: 2900 -7524.747 0.013 0.013
Chain 1: 3000 -7656.706 0.014 0.016
Chain 1: 3100 -7627.474 0.013 0.013
Chain 1: 3200 -7838.162 0.013 0.013
Chain 1: 3300 -7554.506 0.014 0.013
Chain 1: 3400 -7789.747 0.017 0.016
Chain 1: 3500 -7531.488 0.020 0.017
Chain 1: 3600 -7592.162 0.019 0.017
Chain 1: 3700 -7547.518 0.019 0.017
Chain 1: 3800 -7558.433 0.018 0.017
Chain 1: 3900 -7507.864 0.017 0.017
Chain 1: 4000 -7492.308 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002939 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.39 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85997.990 1.000 1.000
Chain 1: 200 -14086.215 3.053 5.105
Chain 1: 300 -10317.434 2.157 1.000
Chain 1: 400 -12154.388 1.655 1.000
Chain 1: 500 -8920.743 1.397 0.365
Chain 1: 600 -9890.423 1.180 0.365
Chain 1: 700 -9340.996 1.020 0.362
Chain 1: 800 -9025.628 0.897 0.362
Chain 1: 900 -8917.988 0.799 0.151
Chain 1: 1000 -9203.820 0.722 0.151
Chain 1: 1100 -8910.049 0.625 0.098 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8599.716 0.118 0.059
Chain 1: 1300 -8964.031 0.086 0.041
Chain 1: 1400 -8886.153 0.072 0.036
Chain 1: 1500 -8788.960 0.036 0.035
Chain 1: 1600 -8846.976 0.027 0.033
Chain 1: 1700 -8936.239 0.022 0.031
Chain 1: 1800 -8478.315 0.024 0.031
Chain 1: 1900 -8599.561 0.025 0.031
Chain 1: 2000 -8603.551 0.021 0.014
Chain 1: 2100 -8539.186 0.019 0.011
Chain 1: 2200 -8516.596 0.016 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002955 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.55 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8411443.720 1.000 1.000
Chain 1: 200 -1585661.894 2.652 4.305
Chain 1: 300 -891954.689 2.027 1.000
Chain 1: 400 -458486.204 1.757 1.000
Chain 1: 500 -358447.623 1.461 0.945
Chain 1: 600 -233527.285 1.307 0.945
Chain 1: 700 -119807.428 1.256 0.945
Chain 1: 800 -87012.559 1.146 0.945
Chain 1: 900 -67377.726 1.051 0.778
Chain 1: 1000 -52199.655 0.975 0.778
Chain 1: 1100 -39688.739 0.907 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38876.317 0.478 0.377
Chain 1: 1300 -26831.779 0.445 0.377
Chain 1: 1400 -26556.165 0.352 0.315
Chain 1: 1500 -23141.219 0.339 0.315
Chain 1: 1600 -22358.621 0.289 0.291
Chain 1: 1700 -21231.268 0.199 0.291
Chain 1: 1800 -21175.929 0.162 0.148
Chain 1: 1900 -21502.761 0.134 0.053
Chain 1: 2000 -20011.998 0.112 0.053
Chain 1: 2100 -20250.623 0.082 0.035
Chain 1: 2200 -20477.518 0.081 0.035
Chain 1: 2300 -20094.190 0.038 0.019
Chain 1: 2400 -19865.993 0.038 0.019
Chain 1: 2500 -19667.942 0.024 0.015
Chain 1: 2600 -19297.429 0.023 0.015
Chain 1: 2700 -19254.275 0.018 0.012
Chain 1: 2800 -18970.666 0.019 0.015
Chain 1: 2900 -19252.325 0.019 0.015
Chain 1: 3000 -19238.562 0.012 0.012
Chain 1: 3100 -19323.581 0.011 0.011
Chain 1: 3200 -19013.806 0.011 0.015
Chain 1: 3300 -19218.933 0.010 0.011
Chain 1: 3400 -18692.915 0.012 0.015
Chain 1: 3500 -19306.112 0.014 0.015
Chain 1: 3600 -18611.163 0.016 0.015
Chain 1: 3700 -18999.099 0.018 0.016
Chain 1: 3800 -17956.199 0.022 0.020
Chain 1: 3900 -17952.269 0.021 0.020
Chain 1: 4000 -18069.614 0.021 0.020
Chain 1: 4100 -17983.147 0.021 0.020
Chain 1: 4200 -17798.904 0.021 0.020
Chain 1: 4300 -17937.672 0.021 0.020
Chain 1: 4400 -17894.023 0.018 0.010
Chain 1: 4500 -17796.483 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001432 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.32 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12546.257 1.000 1.000
Chain 1: 200 -9230.875 0.680 1.000
Chain 1: 300 -8055.331 0.502 0.359
Chain 1: 400 -8236.476 0.382 0.359
Chain 1: 500 -8125.327 0.308 0.146
Chain 1: 600 -7953.480 0.260 0.146
Chain 1: 700 -7903.499 0.224 0.022
Chain 1: 800 -7855.408 0.197 0.022
Chain 1: 900 -7725.703 0.177 0.022
Chain 1: 1000 -7911.020 0.162 0.022
Chain 1: 1100 -7943.622 0.062 0.022
Chain 1: 1200 -7853.195 0.027 0.017
Chain 1: 1300 -7775.297 0.014 0.014
Chain 1: 1400 -7809.558 0.012 0.012
Chain 1: 1500 -7909.276 0.012 0.012
Chain 1: 1600 -7828.572 0.011 0.010
Chain 1: 1700 -7791.665 0.010 0.010
Chain 1: 1800 -7763.043 0.010 0.010
Chain 1: 1900 -7790.093 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001525 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.25 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -62219.936 1.000 1.000
Chain 1: 200 -18210.901 1.708 2.417
Chain 1: 300 -8996.686 1.480 1.024
Chain 1: 400 -9736.113 1.129 1.024
Chain 1: 500 -8015.872 0.946 1.000
Chain 1: 600 -8255.889 0.793 1.000
Chain 1: 700 -8601.732 0.686 0.215
Chain 1: 800 -8002.026 0.609 0.215
Chain 1: 900 -8109.708 0.543 0.076
Chain 1: 1000 -7819.009 0.493 0.076
Chain 1: 1100 -7822.470 0.393 0.075
Chain 1: 1200 -7649.682 0.153 0.040
Chain 1: 1300 -7676.095 0.051 0.037
Chain 1: 1400 -7869.506 0.046 0.029
Chain 1: 1500 -7579.704 0.028 0.029
Chain 1: 1600 -7746.211 0.028 0.025
Chain 1: 1700 -7536.016 0.026 0.025
Chain 1: 1800 -7545.703 0.019 0.023
Chain 1: 1900 -7578.182 0.018 0.023
Chain 1: 2000 -7713.288 0.016 0.021
Chain 1: 2100 -7590.851 0.018 0.021
Chain 1: 2200 -7696.391 0.017 0.018
Chain 1: 2300 -7577.479 0.018 0.018
Chain 1: 2400 -7619.541 0.016 0.016
Chain 1: 2500 -7553.178 0.013 0.016
Chain 1: 2600 -7511.145 0.012 0.014
Chain 1: 2700 -7504.302 0.009 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003122 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.22 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86419.814 1.000 1.000
Chain 1: 200 -13760.333 3.140 5.280
Chain 1: 300 -10002.501 2.219 1.000
Chain 1: 400 -11807.670 1.702 1.000
Chain 1: 500 -8444.394 1.441 0.398
Chain 1: 600 -8840.466 1.209 0.398
Chain 1: 700 -8524.252 1.041 0.376
Chain 1: 800 -9041.764 0.918 0.376
Chain 1: 900 -8766.762 0.820 0.153
Chain 1: 1000 -8956.275 0.740 0.153
Chain 1: 1100 -8755.628 0.642 0.057 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8296.284 0.120 0.055
Chain 1: 1300 -8614.839 0.086 0.045
Chain 1: 1400 -8559.412 0.071 0.037
Chain 1: 1500 -8471.444 0.032 0.037
Chain 1: 1600 -8571.224 0.029 0.031
Chain 1: 1700 -8627.887 0.026 0.023
Chain 1: 1800 -8173.490 0.026 0.023
Chain 1: 1900 -8282.502 0.024 0.021
Chain 1: 2000 -8284.762 0.022 0.013
Chain 1: 2100 -8232.684 0.020 0.012
Chain 1: 2200 -8247.959 0.015 0.010
Chain 1: 2300 -8375.765 0.013 0.010
Chain 1: 2400 -8178.188 0.015 0.012
Chain 1: 2500 -8251.114 0.014 0.012
Chain 1: 2600 -8160.309 0.014 0.011
Chain 1: 2700 -8199.545 0.014 0.011
Chain 1: 2800 -8150.831 0.009 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00299 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.9 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8429990.667 1.000 1.000
Chain 1: 200 -1585170.358 2.659 4.318
Chain 1: 300 -890643.625 2.033 1.000
Chain 1: 400 -457944.973 1.761 1.000
Chain 1: 500 -358167.946 1.464 0.945
Chain 1: 600 -233173.450 1.310 0.945
Chain 1: 700 -119435.086 1.259 0.945
Chain 1: 800 -86728.122 1.148 0.945
Chain 1: 900 -67073.445 1.053 0.780
Chain 1: 1000 -51889.830 0.977 0.780
Chain 1: 1100 -39381.357 0.909 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38565.024 0.479 0.377
Chain 1: 1300 -26509.987 0.447 0.377
Chain 1: 1400 -26233.407 0.353 0.318
Chain 1: 1500 -22818.362 0.340 0.318
Chain 1: 1600 -22036.202 0.290 0.293
Chain 1: 1700 -20906.820 0.201 0.293
Chain 1: 1800 -20851.157 0.163 0.150
Chain 1: 1900 -21178.039 0.135 0.054
Chain 1: 2000 -19686.892 0.114 0.054
Chain 1: 2100 -19925.143 0.083 0.035
Chain 1: 2200 -20152.525 0.082 0.035
Chain 1: 2300 -19768.809 0.039 0.019
Chain 1: 2400 -19540.568 0.039 0.019
Chain 1: 2500 -19342.919 0.025 0.015
Chain 1: 2600 -18971.937 0.023 0.015
Chain 1: 2700 -18928.701 0.018 0.012
Chain 1: 2800 -18645.304 0.019 0.015
Chain 1: 2900 -18926.970 0.019 0.015
Chain 1: 3000 -18913.003 0.012 0.012
Chain 1: 3100 -18998.101 0.011 0.012
Chain 1: 3200 -18688.225 0.012 0.015
Chain 1: 3300 -18893.457 0.011 0.012
Chain 1: 3400 -18367.453 0.012 0.015
Chain 1: 3500 -18980.733 0.015 0.015
Chain 1: 3600 -18285.617 0.016 0.015
Chain 1: 3700 -18673.704 0.018 0.017
Chain 1: 3800 -17630.686 0.023 0.021
Chain 1: 3900 -17626.815 0.021 0.021
Chain 1: 4000 -17744.084 0.022 0.021
Chain 1: 4100 -17657.671 0.022 0.021
Chain 1: 4200 -17473.380 0.021 0.021
Chain 1: 4300 -17612.116 0.021 0.021
Chain 1: 4400 -17568.404 0.018 0.011
Chain 1: 4500 -17470.908 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00132 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.2 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -11825.645 1.000 1.000
Chain 1: 200 -8818.314 0.671 1.000
Chain 1: 300 -7797.939 0.491 0.341
Chain 1: 400 -7880.557 0.371 0.341
Chain 1: 500 -7765.437 0.299 0.131
Chain 1: 600 -7622.007 0.253 0.131
Chain 1: 700 -7566.944 0.218 0.019
Chain 1: 800 -7572.871 0.191 0.019
Chain 1: 900 -7491.398 0.171 0.015
Chain 1: 1000 -7619.756 0.155 0.017
Chain 1: 1100 -7662.219 0.056 0.015
Chain 1: 1200 -7586.163 0.023 0.011
Chain 1: 1300 -7544.946 0.010 0.010
Chain 1: 1400 -7560.374 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001449 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.49 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61068.593 1.000 1.000
Chain 1: 200 -17221.874 1.773 2.546
Chain 1: 300 -8518.944 1.523 1.022
Chain 1: 400 -8040.380 1.157 1.022
Chain 1: 500 -8136.224 0.928 1.000
Chain 1: 600 -7854.203 0.779 1.000
Chain 1: 700 -7674.277 0.671 0.060
Chain 1: 800 -8559.528 0.600 0.103
Chain 1: 900 -7640.443 0.547 0.103
Chain 1: 1000 -7578.909 0.493 0.103
Chain 1: 1100 -7636.783 0.394 0.060
Chain 1: 1200 -7479.671 0.141 0.036
Chain 1: 1300 -7512.387 0.040 0.023
Chain 1: 1400 -7749.115 0.037 0.023
Chain 1: 1500 -7495.986 0.039 0.031
Chain 1: 1600 -7393.050 0.037 0.023
Chain 1: 1700 -7372.654 0.035 0.021
Chain 1: 1800 -7415.587 0.025 0.014
Chain 1: 1900 -7451.750 0.013 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003303 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.03 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85938.743 1.000 1.000
Chain 1: 200 -12897.146 3.332 5.663
Chain 1: 300 -9394.222 2.345 1.000
Chain 1: 400 -10148.377 1.778 1.000
Chain 1: 500 -8238.675 1.468 0.373
Chain 1: 600 -7999.734 1.229 0.373
Chain 1: 700 -8251.895 1.058 0.232
Chain 1: 800 -8252.174 0.925 0.232
Chain 1: 900 -8286.794 0.823 0.074
Chain 1: 1000 -8146.032 0.742 0.074
Chain 1: 1100 -8332.126 0.645 0.031 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8073.798 0.082 0.031
Chain 1: 1300 -8153.396 0.045 0.030
Chain 1: 1400 -8166.087 0.038 0.022
Chain 1: 1500 -8063.761 0.016 0.017
Chain 1: 1600 -8149.994 0.014 0.013
Chain 1: 1700 -8250.842 0.012 0.012
Chain 1: 1800 -7872.153 0.017 0.013
Chain 1: 1900 -7969.942 0.018 0.013
Chain 1: 2000 -7940.545 0.017 0.012
Chain 1: 2100 -8086.302 0.016 0.012
Chain 1: 2200 -7863.626 0.016 0.012
Chain 1: 2300 -7993.259 0.016 0.013
Chain 1: 2400 -7892.627 0.017 0.013
Chain 1: 2500 -7947.280 0.017 0.013
Chain 1: 2600 -7959.927 0.016 0.013
Chain 1: 2700 -7880.996 0.016 0.013
Chain 1: 2800 -7866.632 0.011 0.012
Chain 1: 2900 -7855.226 0.010 0.010
Chain 1: 3000 -7871.376 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003145 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.45 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8395854.400 1.000 1.000
Chain 1: 200 -1583596.272 2.651 4.302
Chain 1: 300 -890657.718 2.027 1.000
Chain 1: 400 -457249.125 1.757 1.000
Chain 1: 500 -357597.596 1.461 0.948
Chain 1: 600 -232523.368 1.307 0.948
Chain 1: 700 -118669.841 1.258 0.948
Chain 1: 800 -85840.151 1.148 0.948
Chain 1: 900 -66162.484 1.054 0.778
Chain 1: 1000 -50938.204 0.978 0.778
Chain 1: 1100 -38402.792 0.911 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37569.539 0.483 0.382
Chain 1: 1300 -25531.620 0.452 0.382
Chain 1: 1400 -25246.604 0.359 0.326
Chain 1: 1500 -21836.193 0.346 0.326
Chain 1: 1600 -21051.935 0.296 0.299
Chain 1: 1700 -19927.542 0.206 0.297
Chain 1: 1800 -19871.477 0.168 0.156
Chain 1: 1900 -20196.774 0.140 0.056
Chain 1: 2000 -18710.391 0.118 0.056
Chain 1: 2100 -18948.557 0.087 0.037
Chain 1: 2200 -19174.321 0.086 0.037
Chain 1: 2300 -18792.361 0.040 0.020
Chain 1: 2400 -18564.788 0.041 0.020
Chain 1: 2500 -18366.758 0.026 0.016
Chain 1: 2600 -17997.878 0.024 0.016
Chain 1: 2700 -17955.086 0.019 0.013
Chain 1: 2800 -17672.335 0.020 0.016
Chain 1: 2900 -17953.180 0.020 0.016
Chain 1: 3000 -17939.436 0.012 0.013
Chain 1: 3100 -18024.288 0.012 0.012
Chain 1: 3200 -17715.537 0.012 0.016
Chain 1: 3300 -17919.803 0.011 0.012
Chain 1: 3400 -17395.752 0.013 0.016
Chain 1: 3500 -18006.064 0.015 0.016
Chain 1: 3600 -17314.803 0.017 0.016
Chain 1: 3700 -17700.099 0.019 0.017
Chain 1: 3800 -16662.952 0.024 0.022
Chain 1: 3900 -16659.174 0.022 0.022
Chain 1: 4000 -16776.478 0.023 0.022
Chain 1: 4100 -16690.404 0.023 0.022
Chain 1: 4200 -16507.325 0.022 0.022
Chain 1: 4300 -16645.243 0.022 0.022
Chain 1: 4400 -16602.644 0.019 0.011
Chain 1: 4500 -16505.279 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00129 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.9 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12583.289 1.000 1.000
Chain 1: 200 -9505.027 0.662 1.000
Chain 1: 300 -8256.973 0.492 0.324
Chain 1: 400 -8361.164 0.372 0.324
Chain 1: 500 -8356.773 0.298 0.151
Chain 1: 600 -8151.492 0.252 0.151
Chain 1: 700 -8102.520 0.217 0.025
Chain 1: 800 -8079.625 0.190 0.025
Chain 1: 900 -8126.058 0.170 0.012
Chain 1: 1000 -8136.702 0.153 0.012
Chain 1: 1100 -8193.654 0.054 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001379 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.79 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58233.882 1.000 1.000
Chain 1: 200 -17840.418 1.632 2.264
Chain 1: 300 -8782.242 1.432 1.031
Chain 1: 400 -8197.513 1.092 1.031
Chain 1: 500 -8527.733 0.881 1.000
Chain 1: 600 -8928.472 0.742 1.000
Chain 1: 700 -8458.289 0.644 0.071
Chain 1: 800 -8348.696 0.565 0.071
Chain 1: 900 -8000.568 0.507 0.056
Chain 1: 1000 -7815.966 0.459 0.056
Chain 1: 1100 -7803.830 0.359 0.045
Chain 1: 1200 -7862.873 0.133 0.044
Chain 1: 1300 -7715.650 0.032 0.039
Chain 1: 1400 -7903.057 0.027 0.024
Chain 1: 1500 -7637.627 0.027 0.024
Chain 1: 1600 -7585.831 0.023 0.024
Chain 1: 1700 -7612.699 0.018 0.019
Chain 1: 1800 -7649.241 0.017 0.019
Chain 1: 1900 -7645.084 0.013 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004012 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 40.12 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86928.349 1.000 1.000
Chain 1: 200 -13688.051 3.175 5.351
Chain 1: 300 -10058.370 2.237 1.000
Chain 1: 400 -10835.865 1.696 1.000
Chain 1: 500 -8995.453 1.398 0.361
Chain 1: 600 -8525.228 1.174 0.361
Chain 1: 700 -8887.427 1.012 0.205
Chain 1: 800 -9066.599 0.888 0.205
Chain 1: 900 -8916.056 0.791 0.072
Chain 1: 1000 -8696.646 0.715 0.072
Chain 1: 1100 -8848.916 0.616 0.055 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8717.233 0.083 0.041
Chain 1: 1300 -8749.583 0.047 0.025
Chain 1: 1400 -8757.110 0.040 0.020
Chain 1: 1500 -8619.975 0.021 0.017
Chain 1: 1600 -8728.692 0.017 0.017
Chain 1: 1700 -8812.867 0.014 0.016
Chain 1: 1800 -8399.593 0.017 0.016
Chain 1: 1900 -8495.879 0.016 0.015
Chain 1: 2000 -8469.168 0.014 0.012
Chain 1: 2100 -8591.807 0.014 0.012
Chain 1: 2200 -8411.972 0.014 0.012
Chain 1: 2300 -8490.717 0.015 0.012
Chain 1: 2400 -8560.447 0.015 0.012
Chain 1: 2500 -8505.908 0.015 0.011
Chain 1: 2600 -8505.378 0.013 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003473 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.73 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8419740.658 1.000 1.000
Chain 1: 200 -1587386.742 2.652 4.304
Chain 1: 300 -890786.955 2.029 1.000
Chain 1: 400 -458127.419 1.758 1.000
Chain 1: 500 -358104.575 1.462 0.944
Chain 1: 600 -232903.896 1.308 0.944
Chain 1: 700 -119226.463 1.257 0.944
Chain 1: 800 -86478.300 1.147 0.944
Chain 1: 900 -66849.384 1.053 0.782
Chain 1: 1000 -51676.016 0.977 0.782
Chain 1: 1100 -39186.895 0.909 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38364.715 0.480 0.379
Chain 1: 1300 -26357.033 0.448 0.379
Chain 1: 1400 -26078.758 0.354 0.319
Chain 1: 1500 -22675.772 0.341 0.319
Chain 1: 1600 -21895.065 0.291 0.294
Chain 1: 1700 -20773.133 0.201 0.294
Chain 1: 1800 -20718.178 0.164 0.150
Chain 1: 1900 -21044.346 0.136 0.054
Chain 1: 2000 -19557.783 0.114 0.054
Chain 1: 2100 -19796.026 0.083 0.036
Chain 1: 2200 -20022.164 0.082 0.036
Chain 1: 2300 -19639.601 0.039 0.019
Chain 1: 2400 -19411.742 0.039 0.019
Chain 1: 2500 -19213.626 0.025 0.015
Chain 1: 2600 -18844.026 0.023 0.015
Chain 1: 2700 -18801.013 0.018 0.012
Chain 1: 2800 -18517.883 0.019 0.015
Chain 1: 2900 -18799.008 0.019 0.015
Chain 1: 3000 -18785.241 0.012 0.012
Chain 1: 3100 -18870.251 0.011 0.012
Chain 1: 3200 -18560.993 0.012 0.015
Chain 1: 3300 -18765.633 0.011 0.012
Chain 1: 3400 -18240.652 0.012 0.015
Chain 1: 3500 -18852.379 0.015 0.015
Chain 1: 3600 -18159.170 0.016 0.015
Chain 1: 3700 -18545.892 0.018 0.017
Chain 1: 3800 -17505.784 0.023 0.021
Chain 1: 3900 -17501.896 0.021 0.021
Chain 1: 4000 -17619.224 0.022 0.021
Chain 1: 4100 -17533.041 0.022 0.021
Chain 1: 4200 -17349.269 0.021 0.021
Chain 1: 4300 -17487.684 0.021 0.021
Chain 1: 4400 -17444.549 0.018 0.011
Chain 1: 4500 -17347.051 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001283 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.83 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13004.517 1.000 1.000
Chain 1: 200 -9796.179 0.664 1.000
Chain 1: 300 -8409.578 0.497 0.328
Chain 1: 400 -8626.027 0.379 0.328
Chain 1: 500 -8149.079 0.315 0.165
Chain 1: 600 -8324.951 0.266 0.165
Chain 1: 700 -8419.222 0.230 0.059
Chain 1: 800 -8277.555 0.203 0.059
Chain 1: 900 -8280.351 0.181 0.025
Chain 1: 1000 -8288.752 0.163 0.025
Chain 1: 1100 -8387.481 0.064 0.021
Chain 1: 1200 -8232.975 0.033 0.019
Chain 1: 1300 -8202.932 0.017 0.017
Chain 1: 1400 -8188.873 0.015 0.012
Chain 1: 1500 -8283.937 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001391 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.91 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -59301.004 1.000 1.000
Chain 1: 200 -18579.461 1.596 2.192
Chain 1: 300 -9108.040 1.411 1.040
Chain 1: 400 -8236.200 1.084 1.040
Chain 1: 500 -9421.076 0.893 1.000
Chain 1: 600 -8527.401 0.761 1.000
Chain 1: 700 -8636.897 0.654 0.126
Chain 1: 800 -8312.807 0.577 0.126
Chain 1: 900 -8109.428 0.516 0.106
Chain 1: 1000 -8060.504 0.465 0.106
Chain 1: 1100 -8031.903 0.365 0.105
Chain 1: 1200 -7781.219 0.149 0.039
Chain 1: 1300 -7925.959 0.047 0.032
Chain 1: 1400 -7883.583 0.037 0.025
Chain 1: 1500 -7658.309 0.028 0.025
Chain 1: 1600 -7798.583 0.019 0.018
Chain 1: 1700 -7582.426 0.021 0.025
Chain 1: 1800 -7716.893 0.018 0.018
Chain 1: 1900 -7723.966 0.016 0.018
Chain 1: 2000 -7831.732 0.017 0.018
Chain 1: 2100 -7694.510 0.018 0.018
Chain 1: 2200 -8095.062 0.020 0.018
Chain 1: 2300 -7710.372 0.023 0.018
Chain 1: 2400 -7852.058 0.024 0.018
Chain 1: 2500 -7700.918 0.023 0.018
Chain 1: 2600 -7642.080 0.022 0.018
Chain 1: 2700 -7639.568 0.020 0.018
Chain 1: 2800 -7650.119 0.018 0.018
Chain 1: 2900 -7482.397 0.020 0.018
Chain 1: 3000 -7637.745 0.021 0.020
Chain 1: 3100 -7632.187 0.019 0.020
Chain 1: 3200 -7864.002 0.017 0.020
Chain 1: 3300 -7545.570 0.016 0.020
Chain 1: 3400 -7766.362 0.017 0.020
Chain 1: 3500 -7546.227 0.018 0.022
Chain 1: 3600 -7612.321 0.018 0.022
Chain 1: 3700 -7567.007 0.019 0.022
Chain 1: 3800 -7538.608 0.019 0.022
Chain 1: 3900 -7512.806 0.017 0.020
Chain 1: 4000 -7508.258 0.015 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002973 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.73 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85976.809 1.000 1.000
Chain 1: 200 -14229.295 3.021 5.042
Chain 1: 300 -10440.245 2.135 1.000
Chain 1: 400 -12361.088 1.640 1.000
Chain 1: 500 -8832.386 1.392 0.400
Chain 1: 600 -9412.280 1.170 0.400
Chain 1: 700 -8976.729 1.010 0.363
Chain 1: 800 -9388.452 0.889 0.363
Chain 1: 900 -9248.263 0.792 0.155
Chain 1: 1000 -9462.439 0.715 0.155
Chain 1: 1100 -9204.832 0.618 0.062 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8737.421 0.119 0.053
Chain 1: 1300 -9025.811 0.086 0.049
Chain 1: 1400 -8939.930 0.071 0.044
Chain 1: 1500 -8908.886 0.032 0.032
Chain 1: 1600 -8953.940 0.026 0.028
Chain 1: 1700 -9019.316 0.022 0.023
Chain 1: 1800 -8555.739 0.023 0.023
Chain 1: 1900 -8673.046 0.023 0.023
Chain 1: 2000 -8693.740 0.021 0.014
Chain 1: 2100 -8780.441 0.019 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003479 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.79 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8375947.240 1.000 1.000
Chain 1: 200 -1578787.634 2.653 4.305
Chain 1: 300 -890658.734 2.026 1.000
Chain 1: 400 -458906.310 1.755 1.000
Chain 1: 500 -359853.581 1.459 0.941
Chain 1: 600 -234743.138 1.304 0.941
Chain 1: 700 -120485.890 1.254 0.941
Chain 1: 800 -87623.462 1.144 0.941
Chain 1: 900 -67864.043 1.049 0.773
Chain 1: 1000 -52597.987 0.973 0.773
Chain 1: 1100 -40007.812 0.905 0.533 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39181.879 0.476 0.375
Chain 1: 1300 -27048.656 0.444 0.375
Chain 1: 1400 -26763.545 0.351 0.315
Chain 1: 1500 -23328.357 0.338 0.315
Chain 1: 1600 -22539.793 0.288 0.291
Chain 1: 1700 -21401.298 0.199 0.290
Chain 1: 1800 -21343.430 0.161 0.147
Chain 1: 1900 -21670.452 0.134 0.053
Chain 1: 2000 -20173.867 0.112 0.053
Chain 1: 2100 -20412.481 0.082 0.035
Chain 1: 2200 -20640.804 0.081 0.035
Chain 1: 2300 -20256.168 0.038 0.019
Chain 1: 2400 -20027.798 0.038 0.019
Chain 1: 2500 -19830.370 0.024 0.015
Chain 1: 2600 -19459.079 0.023 0.015
Chain 1: 2700 -19415.552 0.018 0.012
Chain 1: 2800 -19132.269 0.019 0.015
Chain 1: 2900 -19414.074 0.019 0.015
Chain 1: 3000 -19399.999 0.011 0.012
Chain 1: 3100 -19485.216 0.011 0.011
Chain 1: 3200 -19175.136 0.011 0.015
Chain 1: 3300 -19380.440 0.010 0.011
Chain 1: 3400 -18854.262 0.012 0.015
Chain 1: 3500 -19468.019 0.014 0.015
Chain 1: 3600 -18772.223 0.016 0.015
Chain 1: 3700 -19160.967 0.018 0.016
Chain 1: 3800 -18117.027 0.022 0.020
Chain 1: 3900 -18113.154 0.021 0.020
Chain 1: 4000 -18230.376 0.021 0.020
Chain 1: 4100 -18144.044 0.021 0.020
Chain 1: 4200 -17959.434 0.021 0.020
Chain 1: 4300 -18098.358 0.020 0.020
Chain 1: 4400 -18054.495 0.018 0.010
Chain 1: 4500 -17956.994 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001242 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.42 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48901.564 1.000 1.000
Chain 1: 200 -20605.419 1.187 1.373
Chain 1: 300 -13890.564 0.952 1.000
Chain 1: 400 -17078.045 0.761 1.000
Chain 1: 500 -13528.293 0.661 0.483
Chain 1: 600 -14544.121 0.563 0.483
Chain 1: 700 -10902.701 0.530 0.334
Chain 1: 800 -14639.497 0.496 0.334
Chain 1: 900 -14724.719 0.441 0.262
Chain 1: 1000 -11235.584 0.428 0.311
Chain 1: 1100 -13324.050 0.344 0.262
Chain 1: 1200 -10070.375 0.239 0.262
Chain 1: 1300 -10037.048 0.191 0.255
Chain 1: 1400 -11156.214 0.182 0.255
Chain 1: 1500 -14916.201 0.181 0.252
Chain 1: 1600 -12933.358 0.189 0.252
Chain 1: 1700 -13511.881 0.160 0.157
Chain 1: 1800 -10222.012 0.167 0.157
Chain 1: 1900 -10765.123 0.171 0.157
Chain 1: 2000 -11215.687 0.144 0.153
Chain 1: 2100 -12751.321 0.141 0.120
Chain 1: 2200 -10510.332 0.130 0.120
Chain 1: 2300 -9629.906 0.139 0.120
Chain 1: 2400 -9434.432 0.131 0.120
Chain 1: 2500 -9585.733 0.107 0.091
Chain 1: 2600 -9260.769 0.095 0.050
Chain 1: 2700 -9154.362 0.092 0.050
Chain 1: 2800 -20730.742 0.116 0.050
Chain 1: 2900 -10441.671 0.209 0.091
Chain 1: 3000 -9657.179 0.213 0.091
Chain 1: 3100 -10646.689 0.211 0.091
Chain 1: 3200 -8993.608 0.208 0.091
Chain 1: 3300 -10427.670 0.212 0.093
Chain 1: 3400 -15254.926 0.242 0.138
Chain 1: 3500 -9588.712 0.299 0.184
Chain 1: 3600 -8929.863 0.303 0.184
Chain 1: 3700 -9014.664 0.303 0.184
Chain 1: 3800 -9028.949 0.247 0.138
Chain 1: 3900 -9424.947 0.153 0.093
Chain 1: 4000 -8622.262 0.154 0.093
Chain 1: 4100 -9124.395 0.150 0.093
Chain 1: 4200 -9589.988 0.137 0.074
Chain 1: 4300 -14135.109 0.155 0.074
Chain 1: 4400 -8727.059 0.186 0.074
Chain 1: 4500 -9534.443 0.135 0.074
Chain 1: 4600 -8799.094 0.136 0.084
Chain 1: 4700 -8698.262 0.136 0.084
Chain 1: 4800 -11228.907 0.159 0.085
Chain 1: 4900 -12097.314 0.161 0.085
Chain 1: 5000 -13563.618 0.163 0.085
Chain 1: 5100 -8497.191 0.217 0.108
Chain 1: 5200 -9197.467 0.220 0.108
Chain 1: 5300 -14747.511 0.225 0.108
Chain 1: 5400 -8741.716 0.232 0.108
Chain 1: 5500 -11507.584 0.248 0.225
Chain 1: 5600 -8304.980 0.278 0.240
Chain 1: 5700 -12636.655 0.311 0.343
Chain 1: 5800 -8478.078 0.337 0.376
Chain 1: 5900 -9411.913 0.340 0.376
Chain 1: 6000 -11756.864 0.349 0.376
Chain 1: 6100 -8589.831 0.327 0.369
Chain 1: 6200 -8869.800 0.322 0.369
Chain 1: 6300 -8599.299 0.288 0.343
Chain 1: 6400 -10454.496 0.237 0.240
Chain 1: 6500 -12876.793 0.231 0.199
Chain 1: 6600 -9506.710 0.228 0.199
Chain 1: 6700 -8531.705 0.206 0.188
Chain 1: 6800 -8766.728 0.159 0.177
Chain 1: 6900 -8580.311 0.151 0.177
Chain 1: 7000 -15060.676 0.174 0.177
Chain 1: 7100 -8402.664 0.217 0.177
Chain 1: 7200 -8584.678 0.216 0.177
Chain 1: 7300 -8402.240 0.215 0.177
Chain 1: 7400 -8546.403 0.199 0.114
Chain 1: 7500 -8374.275 0.182 0.027
Chain 1: 7600 -8509.110 0.148 0.022
Chain 1: 7700 -12213.853 0.167 0.022
Chain 1: 7800 -8550.420 0.207 0.022
Chain 1: 7900 -9535.098 0.215 0.103
Chain 1: 8000 -12009.931 0.193 0.103
Chain 1: 8100 -8298.114 0.158 0.103
Chain 1: 8200 -9584.531 0.170 0.134
Chain 1: 8300 -8160.998 0.185 0.174
Chain 1: 8400 -8554.232 0.188 0.174
Chain 1: 8500 -8729.698 0.188 0.174
Chain 1: 8600 -8168.557 0.193 0.174
Chain 1: 8700 -8283.496 0.164 0.134
Chain 1: 8800 -8116.881 0.123 0.103
Chain 1: 8900 -8741.624 0.120 0.071
Chain 1: 9000 -10609.621 0.117 0.071
Chain 1: 9100 -8313.241 0.100 0.071
Chain 1: 9200 -12239.178 0.119 0.071
Chain 1: 9300 -8174.374 0.151 0.071
Chain 1: 9400 -8135.319 0.147 0.071
Chain 1: 9500 -9388.041 0.158 0.133
Chain 1: 9600 -8372.999 0.164 0.133
Chain 1: 9700 -8111.984 0.165 0.133
Chain 1: 9800 -8460.333 0.167 0.133
Chain 1: 9900 -11696.893 0.188 0.176
Chain 1: 10000 -9847.959 0.189 0.188
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001435 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.35 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57032.540 1.000 1.000
Chain 1: 200 -17466.118 1.633 2.265
Chain 1: 300 -8762.423 1.420 1.000
Chain 1: 400 -8392.265 1.076 1.000
Chain 1: 500 -8489.578 0.863 0.993
Chain 1: 600 -9004.906 0.729 0.993
Chain 1: 700 -7815.594 0.646 0.152
Chain 1: 800 -8139.910 0.570 0.152
Chain 1: 900 -7994.375 0.509 0.057
Chain 1: 1000 -7912.289 0.459 0.057
Chain 1: 1100 -7746.418 0.361 0.044
Chain 1: 1200 -7670.149 0.136 0.040
Chain 1: 1300 -7620.802 0.037 0.021
Chain 1: 1400 -7853.630 0.036 0.021
Chain 1: 1500 -7658.640 0.037 0.025
Chain 1: 1600 -7678.157 0.032 0.021
Chain 1: 1700 -7561.816 0.018 0.018
Chain 1: 1800 -7585.634 0.014 0.015
Chain 1: 1900 -7620.794 0.013 0.010
Chain 1: 2000 -7662.172 0.012 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003395 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.95 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87032.924 1.000 1.000
Chain 1: 200 -13492.159 3.225 5.451
Chain 1: 300 -9834.782 2.274 1.000
Chain 1: 400 -10735.560 1.727 1.000
Chain 1: 500 -8761.150 1.426 0.372
Chain 1: 600 -8310.533 1.198 0.372
Chain 1: 700 -8639.037 1.032 0.225
Chain 1: 800 -9240.140 0.911 0.225
Chain 1: 900 -8679.645 0.817 0.084
Chain 1: 1000 -8334.768 0.740 0.084
Chain 1: 1100 -8654.717 0.643 0.065 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8269.216 0.103 0.065
Chain 1: 1300 -8485.725 0.068 0.054
Chain 1: 1400 -8525.739 0.060 0.047
Chain 1: 1500 -8376.541 0.039 0.041
Chain 1: 1600 -8488.052 0.035 0.038
Chain 1: 1700 -8568.382 0.033 0.037
Chain 1: 1800 -8147.425 0.031 0.037
Chain 1: 1900 -8247.063 0.026 0.026
Chain 1: 2000 -8221.289 0.022 0.018
Chain 1: 2100 -8345.935 0.020 0.015
Chain 1: 2200 -8153.172 0.018 0.015
Chain 1: 2300 -8241.709 0.016 0.013
Chain 1: 2400 -8310.847 0.016 0.013
Chain 1: 2500 -8256.969 0.015 0.012
Chain 1: 2600 -8257.820 0.014 0.011
Chain 1: 2700 -8174.796 0.014 0.011
Chain 1: 2800 -8135.429 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002567 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.67 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8420309.526 1.000 1.000
Chain 1: 200 -1586523.502 2.654 4.307
Chain 1: 300 -890574.621 2.030 1.000
Chain 1: 400 -457337.806 1.759 1.000
Chain 1: 500 -357504.279 1.463 0.947
Chain 1: 600 -232539.009 1.309 0.947
Chain 1: 700 -118977.781 1.258 0.947
Chain 1: 800 -86246.755 1.148 0.947
Chain 1: 900 -66637.933 1.053 0.781
Chain 1: 1000 -51471.791 0.978 0.781
Chain 1: 1100 -38984.139 0.910 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38164.474 0.481 0.380
Chain 1: 1300 -26156.763 0.449 0.380
Chain 1: 1400 -25879.057 0.355 0.320
Chain 1: 1500 -22475.618 0.342 0.320
Chain 1: 1600 -21694.774 0.292 0.295
Chain 1: 1700 -20572.923 0.202 0.294
Chain 1: 1800 -20518.084 0.165 0.151
Chain 1: 1900 -20844.307 0.137 0.055
Chain 1: 2000 -19357.427 0.115 0.055
Chain 1: 2100 -19595.798 0.084 0.036
Chain 1: 2200 -19821.929 0.083 0.036
Chain 1: 2300 -19439.361 0.039 0.020
Chain 1: 2400 -19211.463 0.039 0.020
Chain 1: 2500 -19013.260 0.025 0.016
Chain 1: 2600 -18643.644 0.024 0.016
Chain 1: 2700 -18600.595 0.018 0.012
Chain 1: 2800 -18317.379 0.020 0.015
Chain 1: 2900 -18598.562 0.020 0.015
Chain 1: 3000 -18584.812 0.012 0.012
Chain 1: 3100 -18669.834 0.011 0.012
Chain 1: 3200 -18360.501 0.012 0.015
Chain 1: 3300 -18565.191 0.011 0.012
Chain 1: 3400 -18040.063 0.013 0.015
Chain 1: 3500 -18651.988 0.015 0.015
Chain 1: 3600 -17958.520 0.017 0.015
Chain 1: 3700 -18345.428 0.019 0.017
Chain 1: 3800 -17304.906 0.023 0.021
Chain 1: 3900 -17300.987 0.022 0.021
Chain 1: 4000 -17418.334 0.022 0.021
Chain 1: 4100 -17332.109 0.022 0.021
Chain 1: 4200 -17148.257 0.022 0.021
Chain 1: 4300 -17286.754 0.021 0.021
Chain 1: 4400 -17243.534 0.019 0.011
Chain 1: 4500 -17146.008 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001375 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.75 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48897.238 1.000 1.000
Chain 1: 200 -20030.611 1.221 1.441
Chain 1: 300 -18870.889 0.834 1.000
Chain 1: 400 -18143.790 0.636 1.000
Chain 1: 500 -13446.225 0.578 0.349
Chain 1: 600 -14102.279 0.490 0.349
Chain 1: 700 -14594.372 0.425 0.061
Chain 1: 800 -19293.643 0.402 0.244
Chain 1: 900 -13130.046 0.409 0.244
Chain 1: 1000 -25068.606 0.416 0.349
Chain 1: 1100 -10110.884 0.464 0.349
Chain 1: 1200 -10430.435 0.323 0.244
Chain 1: 1300 -11795.253 0.328 0.244
Chain 1: 1400 -11117.425 0.331 0.244
Chain 1: 1500 -10236.509 0.304 0.116
Chain 1: 1600 -10879.354 0.305 0.116
Chain 1: 1700 -10180.041 0.309 0.116
Chain 1: 1800 -13804.076 0.311 0.116
Chain 1: 1900 -11239.103 0.287 0.116
Chain 1: 2000 -15742.789 0.268 0.116
Chain 1: 2100 -13164.187 0.139 0.116
Chain 1: 2200 -11902.252 0.147 0.116
Chain 1: 2300 -11083.201 0.143 0.106
Chain 1: 2400 -9448.405 0.154 0.173
Chain 1: 2500 -10859.218 0.158 0.173
Chain 1: 2600 -9085.524 0.172 0.195
Chain 1: 2700 -9860.272 0.173 0.195
Chain 1: 2800 -9177.399 0.154 0.173
Chain 1: 2900 -11535.552 0.152 0.173
Chain 1: 3000 -10920.052 0.129 0.130
Chain 1: 3100 -14079.750 0.132 0.130
Chain 1: 3200 -10001.873 0.162 0.173
Chain 1: 3300 -14763.143 0.187 0.195
Chain 1: 3400 -9160.429 0.231 0.204
Chain 1: 3500 -9051.258 0.219 0.204
Chain 1: 3600 -9409.145 0.203 0.204
Chain 1: 3700 -15833.113 0.236 0.224
Chain 1: 3800 -9018.529 0.304 0.323
Chain 1: 3900 -9771.287 0.291 0.323
Chain 1: 4000 -9704.958 0.286 0.323
Chain 1: 4100 -8762.204 0.274 0.323
Chain 1: 4200 -11542.404 0.258 0.241
Chain 1: 4300 -9947.935 0.242 0.160
Chain 1: 4400 -14149.618 0.210 0.160
Chain 1: 4500 -8880.592 0.268 0.241
Chain 1: 4600 -9415.764 0.270 0.241
Chain 1: 4700 -14780.820 0.266 0.241
Chain 1: 4800 -8788.831 0.258 0.241
Chain 1: 4900 -8715.881 0.252 0.241
Chain 1: 5000 -19377.023 0.306 0.297
Chain 1: 5100 -8517.650 0.423 0.363
Chain 1: 5200 -9088.286 0.405 0.363
Chain 1: 5300 -13891.852 0.423 0.363
Chain 1: 5400 -8376.752 0.460 0.550 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 5500 -12860.302 0.435 0.363
Chain 1: 5600 -10658.946 0.450 0.363
Chain 1: 5700 -10148.166 0.419 0.349
Chain 1: 5800 -8848.596 0.365 0.346
Chain 1: 5900 -10445.316 0.380 0.346
Chain 1: 6000 -10218.126 0.327 0.207
Chain 1: 6100 -8427.204 0.221 0.207
Chain 1: 6200 -8084.198 0.219 0.207
Chain 1: 6300 -13217.706 0.223 0.207
Chain 1: 6400 -13423.810 0.159 0.153
Chain 1: 6500 -9167.407 0.170 0.153
Chain 1: 6600 -8495.712 0.157 0.147
Chain 1: 6700 -12166.535 0.183 0.153
Chain 1: 6800 -8331.524 0.214 0.213
Chain 1: 6900 -10992.189 0.223 0.242
Chain 1: 7000 -10942.422 0.221 0.242
Chain 1: 7100 -14976.370 0.227 0.269
Chain 1: 7200 -8177.021 0.306 0.302
Chain 1: 7300 -11532.616 0.296 0.291
Chain 1: 7400 -8708.659 0.327 0.302
Chain 1: 7500 -10512.448 0.298 0.291
Chain 1: 7600 -11655.296 0.299 0.291
Chain 1: 7700 -12969.950 0.279 0.269
Chain 1: 7800 -11772.790 0.244 0.242
Chain 1: 7900 -8180.407 0.263 0.269
Chain 1: 8000 -8552.459 0.267 0.269
Chain 1: 8100 -8073.678 0.246 0.172
Chain 1: 8200 -9871.998 0.181 0.172
Chain 1: 8300 -8390.489 0.170 0.172
Chain 1: 8400 -11357.995 0.163 0.172
Chain 1: 8500 -12359.786 0.154 0.102
Chain 1: 8600 -10133.508 0.167 0.177
Chain 1: 8700 -9044.978 0.168 0.177
Chain 1: 8800 -8266.042 0.168 0.177
Chain 1: 8900 -8486.010 0.126 0.120
Chain 1: 9000 -8774.397 0.125 0.120
Chain 1: 9100 -8199.868 0.126 0.120
Chain 1: 9200 -8450.737 0.111 0.094
Chain 1: 9300 -9543.551 0.105 0.094
Chain 1: 9400 -11057.445 0.093 0.094
Chain 1: 9500 -8102.423 0.121 0.115
Chain 1: 9600 -9457.088 0.113 0.115
Chain 1: 9700 -8135.695 0.117 0.115
Chain 1: 9800 -8922.948 0.117 0.115
Chain 1: 9900 -12372.930 0.142 0.137
Chain 1: 10000 -9368.053 0.171 0.143
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00137 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.7 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58387.689 1.000 1.000
Chain 1: 200 -17798.164 1.640 2.281
Chain 1: 300 -8693.390 1.443 1.047
Chain 1: 400 -8221.844 1.096 1.047
Chain 1: 500 -8302.570 0.879 1.000
Chain 1: 600 -8800.385 0.742 1.000
Chain 1: 700 -7740.896 0.655 0.137
Chain 1: 800 -8113.970 0.579 0.137
Chain 1: 900 -7992.343 0.517 0.057
Chain 1: 1000 -7578.477 0.470 0.057
Chain 1: 1100 -7625.426 0.371 0.057
Chain 1: 1200 -7642.070 0.143 0.055
Chain 1: 1300 -7666.063 0.039 0.046
Chain 1: 1400 -7821.387 0.035 0.020
Chain 1: 1500 -7560.744 0.038 0.034
Chain 1: 1600 -7568.095 0.032 0.020
Chain 1: 1700 -7506.297 0.019 0.015
Chain 1: 1800 -7610.419 0.016 0.014
Chain 1: 1900 -7570.856 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003434 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.34 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86019.533 1.000 1.000
Chain 1: 200 -13511.465 3.183 5.366
Chain 1: 300 -9808.910 2.248 1.000
Chain 1: 400 -11081.879 1.715 1.000
Chain 1: 500 -8796.153 1.424 0.377
Chain 1: 600 -8417.931 1.194 0.377
Chain 1: 700 -8421.020 1.023 0.260
Chain 1: 800 -8626.486 0.898 0.260
Chain 1: 900 -8651.784 0.799 0.115
Chain 1: 1000 -8220.064 0.724 0.115
Chain 1: 1100 -8571.237 0.628 0.053 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8151.488 0.097 0.051
Chain 1: 1300 -8455.343 0.063 0.045
Chain 1: 1400 -8470.100 0.051 0.041
Chain 1: 1500 -8311.853 0.027 0.036
Chain 1: 1600 -8421.564 0.024 0.024
Chain 1: 1700 -8490.851 0.025 0.024
Chain 1: 1800 -8053.197 0.028 0.036
Chain 1: 1900 -8158.839 0.029 0.036
Chain 1: 2000 -8135.446 0.024 0.019
Chain 1: 2100 -8277.673 0.022 0.017
Chain 1: 2200 -8065.550 0.019 0.017
Chain 1: 2300 -8225.736 0.018 0.017
Chain 1: 2400 -8061.342 0.019 0.019
Chain 1: 2500 -8132.881 0.018 0.017
Chain 1: 2600 -8044.892 0.018 0.017
Chain 1: 2700 -8079.141 0.018 0.017
Chain 1: 2800 -8038.879 0.013 0.013
Chain 1: 2900 -8132.531 0.013 0.012
Chain 1: 3000 -7966.506 0.014 0.017
Chain 1: 3100 -8121.781 0.015 0.019
Chain 1: 3200 -7993.415 0.014 0.016
Chain 1: 3300 -8001.398 0.012 0.012
Chain 1: 3400 -8162.908 0.012 0.012
Chain 1: 3500 -8173.397 0.011 0.012
Chain 1: 3600 -7949.829 0.013 0.016
Chain 1: 3700 -8096.380 0.014 0.018
Chain 1: 3800 -7956.241 0.015 0.018
Chain 1: 3900 -7890.615 0.015 0.018
Chain 1: 4000 -7967.038 0.014 0.018
Chain 1: 4100 -7961.978 0.012 0.016
Chain 1: 4200 -7946.022 0.011 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003039 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.39 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8393404.268 1.000 1.000
Chain 1: 200 -1582220.139 2.652 4.305
Chain 1: 300 -890282.216 2.027 1.000
Chain 1: 400 -457641.500 1.757 1.000
Chain 1: 500 -358231.467 1.461 0.945
Chain 1: 600 -233301.347 1.307 0.945
Chain 1: 700 -119457.551 1.256 0.945
Chain 1: 800 -86613.248 1.147 0.945
Chain 1: 900 -66932.036 1.052 0.777
Chain 1: 1000 -51705.712 0.976 0.777
Chain 1: 1100 -39154.611 0.908 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38332.931 0.480 0.379
Chain 1: 1300 -26256.046 0.448 0.379
Chain 1: 1400 -25973.920 0.355 0.321
Chain 1: 1500 -22551.766 0.342 0.321
Chain 1: 1600 -21766.088 0.292 0.294
Chain 1: 1700 -20635.354 0.202 0.294
Chain 1: 1800 -20578.833 0.165 0.152
Chain 1: 1900 -20905.410 0.137 0.055
Chain 1: 2000 -19413.442 0.115 0.055
Chain 1: 2100 -19652.057 0.084 0.036
Chain 1: 2200 -19879.126 0.083 0.036
Chain 1: 2300 -19495.696 0.039 0.020
Chain 1: 2400 -19267.589 0.039 0.020
Chain 1: 2500 -19069.706 0.025 0.016
Chain 1: 2600 -18699.382 0.024 0.016
Chain 1: 2700 -18656.215 0.018 0.012
Chain 1: 2800 -18372.864 0.020 0.015
Chain 1: 2900 -18654.430 0.019 0.015
Chain 1: 3000 -18640.552 0.012 0.012
Chain 1: 3100 -18725.585 0.011 0.012
Chain 1: 3200 -18415.992 0.012 0.015
Chain 1: 3300 -18620.963 0.011 0.012
Chain 1: 3400 -18095.363 0.013 0.015
Chain 1: 3500 -18708.039 0.015 0.015
Chain 1: 3600 -18013.757 0.017 0.015
Chain 1: 3700 -18401.282 0.019 0.017
Chain 1: 3800 -17359.455 0.023 0.021
Chain 1: 3900 -17355.588 0.021 0.021
Chain 1: 4000 -17472.884 0.022 0.021
Chain 1: 4100 -17386.531 0.022 0.021
Chain 1: 4200 -17202.469 0.022 0.021
Chain 1: 4300 -17341.084 0.021 0.021
Chain 1: 4400 -17297.654 0.019 0.011
Chain 1: 4500 -17200.144 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001283 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.83 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13645.764 1.000 1.000
Chain 1: 200 -10006.194 0.682 1.000
Chain 1: 300 -8787.930 0.501 0.364
Chain 1: 400 -8866.893 0.378 0.364
Chain 1: 500 -9133.957 0.308 0.139
Chain 1: 600 -8673.123 0.266 0.139
Chain 1: 700 -8882.287 0.231 0.053
Chain 1: 800 -8682.365 0.205 0.053
Chain 1: 900 -8581.190 0.184 0.029
Chain 1: 1000 -8740.366 0.167 0.029
Chain 1: 1100 -8980.384 0.070 0.027
Chain 1: 1200 -8628.141 0.037 0.027
Chain 1: 1300 -8583.913 0.024 0.024
Chain 1: 1400 -8594.528 0.023 0.024
Chain 1: 1500 -8676.901 0.021 0.023
Chain 1: 1600 -8578.725 0.017 0.018
Chain 1: 1700 -8557.043 0.015 0.012
Chain 1: 1800 -8530.365 0.013 0.011
Chain 1: 1900 -8549.365 0.012 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001402 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.02 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -62658.894 1.000 1.000
Chain 1: 200 -19083.625 1.642 2.283
Chain 1: 300 -9604.858 1.423 1.000
Chain 1: 400 -10034.444 1.078 1.000
Chain 1: 500 -8717.510 0.893 0.987
Chain 1: 600 -8877.699 0.747 0.987
Chain 1: 700 -8942.742 0.641 0.151
Chain 1: 800 -8969.239 0.562 0.151
Chain 1: 900 -8225.933 0.509 0.090
Chain 1: 1000 -7883.014 0.463 0.090
Chain 1: 1100 -8159.815 0.366 0.044
Chain 1: 1200 -7890.724 0.141 0.043
Chain 1: 1300 -7921.565 0.043 0.034
Chain 1: 1400 -8211.095 0.042 0.034
Chain 1: 1500 -7641.456 0.034 0.034
Chain 1: 1600 -7924.474 0.036 0.035
Chain 1: 1700 -7752.985 0.038 0.035
Chain 1: 1800 -7832.542 0.038 0.035
Chain 1: 1900 -7922.687 0.030 0.034
Chain 1: 2000 -7794.621 0.028 0.034
Chain 1: 2100 -7692.785 0.026 0.022
Chain 1: 2200 -8121.727 0.028 0.022
Chain 1: 2300 -7737.197 0.032 0.035
Chain 1: 2400 -7771.188 0.029 0.022
Chain 1: 2500 -7623.444 0.024 0.019
Chain 1: 2600 -7694.453 0.021 0.016
Chain 1: 2700 -7594.920 0.020 0.013
Chain 1: 2800 -7687.263 0.020 0.013
Chain 1: 2900 -7504.900 0.021 0.016
Chain 1: 3000 -7669.962 0.022 0.019
Chain 1: 3100 -7642.496 0.021 0.019
Chain 1: 3200 -7757.887 0.017 0.015
Chain 1: 3300 -7507.398 0.016 0.015
Chain 1: 3400 -7882.239 0.020 0.019
Chain 1: 3500 -7605.551 0.022 0.022
Chain 1: 3600 -7716.842 0.022 0.022
Chain 1: 3700 -7512.199 0.024 0.024
Chain 1: 3800 -7565.848 0.023 0.024
Chain 1: 3900 -7510.461 0.021 0.022
Chain 1: 4000 -7507.071 0.019 0.015
Chain 1: 4100 -7514.956 0.019 0.015
Chain 1: 4200 -7661.945 0.019 0.019
Chain 1: 4300 -7499.799 0.018 0.019
Chain 1: 4400 -7551.786 0.014 0.014
Chain 1: 4500 -7695.044 0.012 0.014
Chain 1: 4600 -7583.457 0.012 0.015
Chain 1: 4700 -7592.636 0.010 0.007 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003168 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.68 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87860.297 1.000 1.000
Chain 1: 200 -14724.579 2.983 4.967
Chain 1: 300 -10869.033 2.107 1.000
Chain 1: 400 -12943.500 1.620 1.000
Chain 1: 500 -9590.979 1.366 0.355
Chain 1: 600 -10383.416 1.151 0.355
Chain 1: 700 -9634.650 0.998 0.350
Chain 1: 800 -9313.558 0.877 0.350
Chain 1: 900 -9620.124 0.784 0.160
Chain 1: 1000 -9040.108 0.712 0.160
Chain 1: 1100 -9292.853 0.614 0.078 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -9149.786 0.119 0.076
Chain 1: 1300 -9411.042 0.086 0.064
Chain 1: 1400 -9383.943 0.071 0.034
Chain 1: 1500 -9330.310 0.036 0.032
Chain 1: 1600 -9366.164 0.029 0.028
Chain 1: 1700 -9429.510 0.022 0.027
Chain 1: 1800 -8972.778 0.024 0.027
Chain 1: 1900 -9071.630 0.022 0.016
Chain 1: 2000 -9089.850 0.015 0.011
Chain 1: 2100 -9193.878 0.014 0.011
Chain 1: 2200 -8948.373 0.015 0.011
Chain 1: 2300 -9060.015 0.013 0.011
Chain 1: 2400 -9124.257 0.014 0.011
Chain 1: 2500 -9064.056 0.014 0.011
Chain 1: 2600 -9103.510 0.014 0.011
Chain 1: 2700 -8991.962 0.015 0.011
Chain 1: 2800 -8938.444 0.010 0.011
Chain 1: 2900 -9045.553 0.010 0.011
Chain 1: 3000 -8959.811 0.011 0.011
Chain 1: 3100 -8924.936 0.010 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003282 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.82 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8424409.551 1.000 1.000
Chain 1: 200 -1583758.158 2.660 4.319
Chain 1: 300 -891365.308 2.032 1.000
Chain 1: 400 -458588.598 1.760 1.000
Chain 1: 500 -358785.032 1.464 0.944
Chain 1: 600 -233842.692 1.309 0.944
Chain 1: 700 -120302.369 1.257 0.944
Chain 1: 800 -87564.412 1.146 0.944
Chain 1: 900 -67952.977 1.051 0.777
Chain 1: 1000 -52796.813 0.975 0.777
Chain 1: 1100 -40309.099 0.906 0.534 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39501.999 0.476 0.374
Chain 1: 1300 -27471.038 0.442 0.374
Chain 1: 1400 -27197.790 0.348 0.310
Chain 1: 1500 -23786.574 0.335 0.310
Chain 1: 1600 -23005.819 0.285 0.289
Chain 1: 1700 -21879.374 0.196 0.287
Chain 1: 1800 -21824.618 0.159 0.143
Chain 1: 1900 -22151.809 0.131 0.051
Chain 1: 2000 -20661.089 0.110 0.051
Chain 1: 2100 -20899.648 0.080 0.034
Chain 1: 2200 -21126.750 0.079 0.034
Chain 1: 2300 -20743.130 0.037 0.018
Chain 1: 2400 -20514.824 0.037 0.018
Chain 1: 2500 -20316.720 0.024 0.015
Chain 1: 2600 -19945.790 0.022 0.015
Chain 1: 2700 -19902.586 0.017 0.011
Chain 1: 2800 -19618.824 0.018 0.014
Chain 1: 2900 -19900.597 0.018 0.014
Chain 1: 3000 -19886.750 0.011 0.011
Chain 1: 3100 -19971.864 0.010 0.011
Chain 1: 3200 -19661.835 0.011 0.014
Chain 1: 3300 -19867.176 0.010 0.011
Chain 1: 3400 -19340.737 0.012 0.014
Chain 1: 3500 -19954.583 0.014 0.014
Chain 1: 3600 -19258.736 0.016 0.014
Chain 1: 3700 -19647.325 0.017 0.016
Chain 1: 3800 -18603.070 0.022 0.020
Chain 1: 3900 -18599.107 0.020 0.020
Chain 1: 4000 -18716.429 0.021 0.020
Chain 1: 4100 -18629.915 0.021 0.020
Chain 1: 4200 -18445.383 0.020 0.020
Chain 1: 4300 -18584.365 0.020 0.020
Chain 1: 4400 -18540.473 0.017 0.010
Chain 1: 4500 -18442.868 0.015 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001354 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.54 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48418.731 1.000 1.000
Chain 1: 200 -17972.330 1.347 1.694
Chain 1: 300 -19686.837 0.927 1.000
Chain 1: 400 -13232.022 0.817 1.000
Chain 1: 500 -17098.856 0.699 0.488
Chain 1: 600 -15769.484 0.597 0.488
Chain 1: 700 -12030.040 0.556 0.311
Chain 1: 800 -12800.379 0.494 0.311
Chain 1: 900 -13813.736 0.447 0.226
Chain 1: 1000 -12947.806 0.409 0.226
Chain 1: 1100 -24911.254 0.357 0.226
Chain 1: 1200 -11650.569 0.302 0.226
Chain 1: 1300 -14467.761 0.312 0.226
Chain 1: 1400 -9528.429 0.315 0.226
Chain 1: 1500 -17715.570 0.339 0.311
Chain 1: 1600 -12084.235 0.377 0.462
Chain 1: 1700 -15671.814 0.369 0.462
Chain 1: 1800 -10567.278 0.411 0.466
Chain 1: 1900 -10575.432 0.404 0.466
Chain 1: 2000 -11070.726 0.402 0.466
Chain 1: 2100 -9531.812 0.370 0.462
Chain 1: 2200 -9215.342 0.259 0.229
Chain 1: 2300 -8722.182 0.246 0.229
Chain 1: 2400 -9732.550 0.204 0.161
Chain 1: 2500 -10002.021 0.161 0.104
Chain 1: 2600 -8791.523 0.128 0.104
Chain 1: 2700 -8901.157 0.106 0.057
Chain 1: 2800 -10008.712 0.069 0.057
Chain 1: 2900 -11737.651 0.084 0.104
Chain 1: 3000 -14522.532 0.098 0.111
Chain 1: 3100 -9370.119 0.137 0.111
Chain 1: 3200 -8590.641 0.143 0.111
Chain 1: 3300 -13174.222 0.172 0.138
Chain 1: 3400 -12372.183 0.168 0.138
Chain 1: 3500 -12767.620 0.168 0.138
Chain 1: 3600 -14531.791 0.167 0.121
Chain 1: 3700 -9221.524 0.223 0.147
Chain 1: 3800 -8613.072 0.219 0.147
Chain 1: 3900 -9872.010 0.217 0.128
Chain 1: 4000 -9006.090 0.208 0.121
Chain 1: 4100 -10709.703 0.169 0.121
Chain 1: 4200 -8513.243 0.185 0.128
Chain 1: 4300 -8827.200 0.154 0.121
Chain 1: 4400 -8620.456 0.150 0.121
Chain 1: 4500 -9470.414 0.156 0.121
Chain 1: 4600 -9008.299 0.149 0.096
Chain 1: 4700 -8231.491 0.101 0.094
Chain 1: 4800 -8469.440 0.096 0.094
Chain 1: 4900 -11556.432 0.110 0.094
Chain 1: 5000 -8951.143 0.130 0.094
Chain 1: 5100 -10441.836 0.128 0.094
Chain 1: 5200 -14611.676 0.131 0.094
Chain 1: 5300 -9884.298 0.175 0.143
Chain 1: 5400 -8455.142 0.190 0.169
Chain 1: 5500 -8713.266 0.184 0.169
Chain 1: 5600 -14897.706 0.220 0.267
Chain 1: 5700 -8245.941 0.291 0.285
Chain 1: 5800 -10656.957 0.311 0.285
Chain 1: 5900 -8866.868 0.305 0.285
Chain 1: 6000 -9966.835 0.287 0.226
Chain 1: 6100 -9419.229 0.278 0.226
Chain 1: 6200 -8107.417 0.266 0.202
Chain 1: 6300 -8982.490 0.228 0.169
Chain 1: 6400 -13252.646 0.243 0.202
Chain 1: 6500 -8453.453 0.297 0.226
Chain 1: 6600 -8548.754 0.256 0.202
Chain 1: 6700 -8985.498 0.181 0.162
Chain 1: 6800 -8488.223 0.164 0.110
Chain 1: 6900 -11581.465 0.170 0.110
Chain 1: 7000 -7962.564 0.205 0.162
Chain 1: 7100 -8248.445 0.202 0.162
Chain 1: 7200 -8833.291 0.193 0.097
Chain 1: 7300 -9141.367 0.186 0.066
Chain 1: 7400 -8854.703 0.157 0.059
Chain 1: 7500 -10166.707 0.114 0.059
Chain 1: 7600 -11658.189 0.125 0.066
Chain 1: 7700 -10380.199 0.133 0.123
Chain 1: 7800 -11078.346 0.133 0.123
Chain 1: 7900 -10851.333 0.109 0.066
Chain 1: 8000 -8109.943 0.097 0.066
Chain 1: 8100 -10628.171 0.117 0.123
Chain 1: 8200 -10629.194 0.111 0.123
Chain 1: 8300 -9707.910 0.117 0.123
Chain 1: 8400 -8044.386 0.134 0.128
Chain 1: 8500 -10268.286 0.143 0.128
Chain 1: 8600 -7996.527 0.158 0.207
Chain 1: 8700 -9024.809 0.158 0.207
Chain 1: 8800 -11160.675 0.170 0.207
Chain 1: 8900 -8394.582 0.201 0.217
Chain 1: 9000 -11171.543 0.192 0.217
Chain 1: 9100 -7931.098 0.209 0.217
Chain 1: 9200 -10306.634 0.232 0.230
Chain 1: 9300 -10366.484 0.224 0.230
Chain 1: 9400 -9924.042 0.207 0.230
Chain 1: 9500 -8200.803 0.207 0.230
Chain 1: 9600 -8128.378 0.179 0.210
Chain 1: 9700 -9973.059 0.186 0.210
Chain 1: 9800 -8146.812 0.190 0.224
Chain 1: 9900 -10915.348 0.182 0.224
Chain 1: 10000 -8074.221 0.192 0.224
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001597 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.97 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56201.070 1.000 1.000
Chain 1: 200 -16885.658 1.664 2.328
Chain 1: 300 -8503.722 1.438 1.000
Chain 1: 400 -8708.361 1.084 1.000
Chain 1: 500 -8370.317 0.876 0.986
Chain 1: 600 -9144.072 0.744 0.986
Chain 1: 700 -8079.041 0.656 0.132
Chain 1: 800 -8019.993 0.575 0.132
Chain 1: 900 -7775.524 0.515 0.085
Chain 1: 1000 -7775.873 0.463 0.085
Chain 1: 1100 -7623.602 0.365 0.040
Chain 1: 1200 -7583.003 0.133 0.031
Chain 1: 1300 -7562.753 0.035 0.023
Chain 1: 1400 -7845.784 0.036 0.031
Chain 1: 1500 -7579.195 0.035 0.031
Chain 1: 1600 -7474.338 0.028 0.020
Chain 1: 1700 -7471.454 0.015 0.014
Chain 1: 1800 -7499.025 0.015 0.014
Chain 1: 1900 -7545.297 0.012 0.006 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004596 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 45.96 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86514.769 1.000 1.000
Chain 1: 200 -13053.119 3.314 5.628
Chain 1: 300 -9545.582 2.332 1.000
Chain 1: 400 -10381.276 1.769 1.000
Chain 1: 500 -8415.394 1.462 0.367
Chain 1: 600 -8155.122 1.224 0.367
Chain 1: 700 -8451.356 1.054 0.234
Chain 1: 800 -8545.286 0.923 0.234
Chain 1: 900 -8439.256 0.822 0.081
Chain 1: 1000 -8170.670 0.743 0.081
Chain 1: 1100 -8447.705 0.647 0.035 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8155.694 0.087 0.035
Chain 1: 1300 -8331.799 0.053 0.033
Chain 1: 1400 -8255.612 0.046 0.033
Chain 1: 1500 -8206.728 0.023 0.032
Chain 1: 1600 -8205.265 0.020 0.021
Chain 1: 1700 -8147.356 0.017 0.013
Chain 1: 1800 -8025.826 0.017 0.015
Chain 1: 1900 -8138.310 0.017 0.015
Chain 1: 2000 -8100.903 0.015 0.014
Chain 1: 2100 -8244.565 0.013 0.014
Chain 1: 2200 -8026.790 0.012 0.014
Chain 1: 2300 -8163.953 0.012 0.014
Chain 1: 2400 -8052.376 0.012 0.014
Chain 1: 2500 -8109.967 0.012 0.014
Chain 1: 2600 -8122.885 0.012 0.014
Chain 1: 2700 -8044.043 0.013 0.014
Chain 1: 2800 -8029.352 0.011 0.014
Chain 1: 2900 -8017.676 0.010 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00366 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 36.6 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8411619.761 1.000 1.000
Chain 1: 200 -1588822.960 2.647 4.294
Chain 1: 300 -891159.533 2.026 1.000
Chain 1: 400 -457309.499 1.756 1.000
Chain 1: 500 -357091.324 1.461 0.949
Chain 1: 600 -231934.962 1.308 0.949
Chain 1: 700 -118423.602 1.258 0.949
Chain 1: 800 -85686.462 1.148 0.949
Chain 1: 900 -66089.989 1.054 0.783
Chain 1: 1000 -50929.304 0.978 0.783
Chain 1: 1100 -38455.600 0.911 0.540 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37629.917 0.483 0.382
Chain 1: 1300 -25655.966 0.452 0.382
Chain 1: 1400 -25377.152 0.358 0.324
Chain 1: 1500 -21982.750 0.345 0.324
Chain 1: 1600 -21203.210 0.295 0.298
Chain 1: 1700 -20086.592 0.205 0.297
Chain 1: 1800 -20032.329 0.167 0.154
Chain 1: 1900 -20357.609 0.139 0.056
Chain 1: 2000 -18875.110 0.117 0.056
Chain 1: 2100 -19113.247 0.086 0.037
Chain 1: 2200 -19338.241 0.085 0.037
Chain 1: 2300 -18956.891 0.040 0.020
Chain 1: 2400 -18729.363 0.040 0.020
Chain 1: 2500 -18531.012 0.026 0.016
Chain 1: 2600 -18162.512 0.024 0.016
Chain 1: 2700 -18119.844 0.019 0.012
Chain 1: 2800 -17836.945 0.020 0.016
Chain 1: 2900 -18117.660 0.020 0.015
Chain 1: 3000 -18104.025 0.012 0.012
Chain 1: 3100 -18188.866 0.011 0.012
Chain 1: 3200 -17880.242 0.012 0.015
Chain 1: 3300 -18084.394 0.011 0.012
Chain 1: 3400 -17560.413 0.013 0.015
Chain 1: 3500 -18170.573 0.015 0.016
Chain 1: 3600 -17479.429 0.017 0.016
Chain 1: 3700 -17864.549 0.019 0.017
Chain 1: 3800 -16827.609 0.024 0.022
Chain 1: 3900 -16823.769 0.022 0.022
Chain 1: 4000 -16941.120 0.023 0.022
Chain 1: 4100 -16855.038 0.023 0.022
Chain 1: 4200 -16671.995 0.022 0.022
Chain 1: 4300 -16809.935 0.022 0.022
Chain 1: 4400 -16767.362 0.019 0.011
Chain 1: 4500 -16669.933 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001287 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.87 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48742.270 1.000 1.000
Chain 1: 200 -16745.467 1.455 1.911
Chain 1: 300 -17367.340 0.982 1.000
Chain 1: 400 -28101.654 0.832 1.000
Chain 1: 500 -16192.706 0.813 0.735
Chain 1: 600 -12552.610 0.726 0.735
Chain 1: 700 -10820.647 0.645 0.382
Chain 1: 800 -15708.819 0.603 0.382
Chain 1: 900 -11095.175 0.582 0.382
Chain 1: 1000 -10682.660 0.528 0.382
Chain 1: 1100 -12881.275 0.445 0.311
Chain 1: 1200 -11608.975 0.265 0.290
Chain 1: 1300 -12345.182 0.267 0.290
Chain 1: 1400 -10485.000 0.247 0.177
Chain 1: 1500 -10839.124 0.177 0.171
Chain 1: 1600 -11670.032 0.155 0.160
Chain 1: 1700 -18921.097 0.177 0.171
Chain 1: 1800 -9481.942 0.245 0.171
Chain 1: 1900 -16090.770 0.245 0.171
Chain 1: 2000 -11748.278 0.278 0.177
Chain 1: 2100 -9356.448 0.287 0.256
Chain 1: 2200 -10409.126 0.286 0.256
Chain 1: 2300 -14853.653 0.310 0.299
Chain 1: 2400 -10028.688 0.340 0.370
Chain 1: 2500 -11122.398 0.347 0.370
Chain 1: 2600 -9290.119 0.359 0.370
Chain 1: 2700 -12518.368 0.347 0.299
Chain 1: 2800 -10173.040 0.270 0.258
Chain 1: 2900 -9437.534 0.237 0.256
Chain 1: 3000 -22551.859 0.258 0.256
Chain 1: 3100 -9913.526 0.360 0.258
Chain 1: 3200 -9503.581 0.354 0.258
Chain 1: 3300 -9560.766 0.325 0.231
Chain 1: 3400 -9471.400 0.278 0.197
Chain 1: 3500 -13531.067 0.298 0.231
Chain 1: 3600 -10489.808 0.307 0.258
Chain 1: 3700 -10233.397 0.284 0.231
Chain 1: 3800 -8755.427 0.278 0.169
Chain 1: 3900 -13896.791 0.307 0.290
Chain 1: 4000 -8830.720 0.306 0.290
Chain 1: 4100 -8768.440 0.179 0.169
Chain 1: 4200 -14925.141 0.216 0.290
Chain 1: 4300 -9644.981 0.270 0.300
Chain 1: 4400 -9253.249 0.274 0.300
Chain 1: 4500 -11075.738 0.260 0.290
Chain 1: 4600 -11089.115 0.231 0.169
Chain 1: 4700 -8580.958 0.258 0.292
Chain 1: 4800 -8751.842 0.243 0.292
Chain 1: 4900 -16528.084 0.253 0.292
Chain 1: 5000 -9499.652 0.270 0.292
Chain 1: 5100 -9515.189 0.269 0.292
Chain 1: 5200 -10375.331 0.236 0.165
Chain 1: 5300 -15995.686 0.217 0.165
Chain 1: 5400 -10051.336 0.272 0.292
Chain 1: 5500 -9462.023 0.261 0.292
Chain 1: 5600 -9319.552 0.263 0.292
Chain 1: 5700 -12775.001 0.261 0.270
Chain 1: 5800 -8639.634 0.306 0.351
Chain 1: 5900 -14869.407 0.301 0.351
Chain 1: 6000 -10809.160 0.265 0.351
Chain 1: 6100 -8634.028 0.290 0.351
Chain 1: 6200 -8693.164 0.282 0.351
Chain 1: 6300 -8873.020 0.249 0.270
Chain 1: 6400 -13659.714 0.225 0.270
Chain 1: 6500 -9797.861 0.258 0.350
Chain 1: 6600 -8345.468 0.274 0.350
Chain 1: 6700 -8285.256 0.248 0.350
Chain 1: 6800 -12884.418 0.236 0.350
Chain 1: 6900 -8615.062 0.243 0.350
Chain 1: 7000 -8337.703 0.209 0.252
Chain 1: 7100 -8650.877 0.187 0.174
Chain 1: 7200 -8338.441 0.191 0.174
Chain 1: 7300 -8255.303 0.190 0.174
Chain 1: 7400 -8268.877 0.155 0.037
Chain 1: 7500 -10763.715 0.138 0.037
Chain 1: 7600 -8436.607 0.149 0.037
Chain 1: 7700 -10351.808 0.166 0.185
Chain 1: 7800 -9728.672 0.137 0.064
Chain 1: 7900 -9683.450 0.088 0.037
Chain 1: 8000 -8480.259 0.099 0.064
Chain 1: 8100 -8379.715 0.096 0.064
Chain 1: 8200 -8495.118 0.094 0.064
Chain 1: 8300 -8177.357 0.097 0.064
Chain 1: 8400 -8474.714 0.100 0.064
Chain 1: 8500 -8254.143 0.080 0.039
Chain 1: 8600 -8667.127 0.057 0.039
Chain 1: 8700 -9059.951 0.043 0.039
Chain 1: 8800 -8917.339 0.038 0.035
Chain 1: 8900 -10817.411 0.055 0.039
Chain 1: 9000 -9691.123 0.053 0.039
Chain 1: 9100 -8962.009 0.059 0.043
Chain 1: 9200 -9154.848 0.060 0.043
Chain 1: 9300 -8802.887 0.060 0.043
Chain 1: 9400 -8785.695 0.057 0.043
Chain 1: 9500 -12962.504 0.087 0.048
Chain 1: 9600 -8342.549 0.137 0.081
Chain 1: 9700 -9270.323 0.143 0.100
Chain 1: 9800 -8404.942 0.152 0.103
Chain 1: 9900 -9532.413 0.146 0.103
Chain 1: 10000 -8756.688 0.143 0.100
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.0014 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -45891.032 1.000 1.000
Chain 1: 200 -15390.549 1.491 1.982
Chain 1: 300 -8634.010 1.255 1.000
Chain 1: 400 -8617.416 0.942 1.000
Chain 1: 500 -8169.787 0.764 0.783
Chain 1: 600 -8061.230 0.639 0.783
Chain 1: 700 -7869.196 0.551 0.055
Chain 1: 800 -8141.606 0.487 0.055
Chain 1: 900 -7873.326 0.436 0.034
Chain 1: 1000 -7691.625 0.395 0.034
Chain 1: 1100 -7820.310 0.297 0.033
Chain 1: 1200 -7694.528 0.100 0.024
Chain 1: 1300 -7538.210 0.024 0.024
Chain 1: 1400 -7626.418 0.025 0.024
Chain 1: 1500 -7568.916 0.020 0.021
Chain 1: 1600 -7721.620 0.021 0.021
Chain 1: 1700 -7467.897 0.022 0.021
Chain 1: 1800 -7557.455 0.020 0.020
Chain 1: 1900 -7520.850 0.017 0.016
Chain 1: 2000 -7561.839 0.015 0.016
Chain 1: 2100 -7548.129 0.013 0.012
Chain 1: 2200 -7652.674 0.013 0.012
Chain 1: 2300 -7560.611 0.012 0.012
Chain 1: 2400 -7584.877 0.011 0.012
Chain 1: 2500 -7591.039 0.011 0.012
Chain 1: 2600 -7486.673 0.010 0.012
Chain 1: 2700 -7506.451 0.007 0.005 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003038 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.38 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86631.519 1.000 1.000
Chain 1: 200 -13405.682 3.231 5.462
Chain 1: 300 -9819.092 2.276 1.000
Chain 1: 400 -10647.944 1.726 1.000
Chain 1: 500 -8765.902 1.424 0.365
Chain 1: 600 -8692.463 1.188 0.365
Chain 1: 700 -8440.730 1.023 0.215
Chain 1: 800 -8933.278 0.902 0.215
Chain 1: 900 -8643.812 0.805 0.078
Chain 1: 1000 -8433.290 0.727 0.078
Chain 1: 1100 -8648.956 0.630 0.055 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8313.562 0.087 0.040
Chain 1: 1300 -8607.102 0.054 0.034
Chain 1: 1400 -8532.707 0.047 0.033
Chain 1: 1500 -8435.633 0.027 0.030
Chain 1: 1600 -8535.856 0.027 0.030
Chain 1: 1700 -8624.496 0.026 0.025
Chain 1: 1800 -8229.398 0.025 0.025
Chain 1: 1900 -8331.216 0.023 0.025
Chain 1: 2000 -8301.728 0.021 0.012
Chain 1: 2100 -8425.309 0.020 0.012
Chain 1: 2200 -8208.434 0.018 0.012
Chain 1: 2300 -8359.940 0.017 0.012
Chain 1: 2400 -8374.372 0.016 0.012
Chain 1: 2500 -8342.924 0.015 0.012
Chain 1: 2600 -8345.293 0.014 0.012
Chain 1: 2700 -8251.692 0.014 0.012
Chain 1: 2800 -8223.472 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003215 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.15 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8407419.195 1.000 1.000
Chain 1: 200 -1590160.357 2.644 4.287
Chain 1: 300 -892641.564 2.023 1.000
Chain 1: 400 -458065.329 1.754 1.000
Chain 1: 500 -357831.166 1.459 0.949
Chain 1: 600 -232640.669 1.306 0.949
Chain 1: 700 -118981.447 1.256 0.949
Chain 1: 800 -86187.267 1.146 0.949
Chain 1: 900 -66563.350 1.052 0.781
Chain 1: 1000 -51382.958 0.976 0.781
Chain 1: 1100 -38881.486 0.908 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38058.304 0.482 0.380
Chain 1: 1300 -26049.343 0.450 0.380
Chain 1: 1400 -25769.543 0.356 0.322
Chain 1: 1500 -22365.084 0.343 0.322
Chain 1: 1600 -21583.037 0.293 0.295
Chain 1: 1700 -20461.879 0.203 0.295
Chain 1: 1800 -20406.830 0.165 0.152
Chain 1: 1900 -20732.538 0.137 0.055
Chain 1: 2000 -19246.723 0.115 0.055
Chain 1: 2100 -19485.140 0.084 0.036
Chain 1: 2200 -19710.741 0.083 0.036
Chain 1: 2300 -19328.769 0.039 0.020
Chain 1: 2400 -19101.030 0.039 0.020
Chain 1: 2500 -18902.731 0.025 0.016
Chain 1: 2600 -18533.680 0.024 0.016
Chain 1: 2700 -18490.868 0.018 0.012
Chain 1: 2800 -18207.730 0.020 0.016
Chain 1: 2900 -18488.759 0.020 0.015
Chain 1: 3000 -18475.121 0.012 0.012
Chain 1: 3100 -18559.990 0.011 0.012
Chain 1: 3200 -18251.029 0.012 0.015
Chain 1: 3300 -18455.466 0.011 0.012
Chain 1: 3400 -17930.862 0.013 0.015
Chain 1: 3500 -18541.928 0.015 0.016
Chain 1: 3600 -17849.709 0.017 0.016
Chain 1: 3700 -18235.616 0.019 0.017
Chain 1: 3800 -17196.941 0.023 0.021
Chain 1: 3900 -17193.076 0.022 0.021
Chain 1: 4000 -17310.439 0.022 0.021
Chain 1: 4100 -17224.218 0.022 0.021
Chain 1: 4200 -17040.839 0.022 0.021
Chain 1: 4300 -17179.022 0.021 0.021
Chain 1: 4400 -17136.147 0.019 0.011
Chain 1: 4500 -17038.679 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001299 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.99 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13301.279 1.000 1.000
Chain 1: 200 -10003.706 0.665 1.000
Chain 1: 300 -8471.952 0.503 0.330
Chain 1: 400 -8189.501 0.386 0.330
Chain 1: 500 -8117.989 0.311 0.181
Chain 1: 600 -7960.891 0.262 0.181
Chain 1: 700 -8143.499 0.228 0.034
Chain 1: 800 -7932.091 0.203 0.034
Chain 1: 900 -8043.577 0.182 0.027
Chain 1: 1000 -8027.159 0.164 0.027
Chain 1: 1100 -8062.606 0.064 0.022
Chain 1: 1200 -7971.298 0.032 0.020
Chain 1: 1300 -7917.082 0.015 0.014
Chain 1: 1400 -7950.689 0.012 0.011
Chain 1: 1500 -8053.911 0.012 0.013
Chain 1: 1600 -7973.043 0.011 0.011
Chain 1: 1700 -7928.715 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001424 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.24 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57216.752 1.000 1.000
Chain 1: 200 -17775.353 1.609 2.219
Chain 1: 300 -8751.451 1.417 1.031
Chain 1: 400 -7919.324 1.089 1.031
Chain 1: 500 -8678.653 0.889 1.000
Chain 1: 600 -9234.456 0.750 1.000
Chain 1: 700 -8353.480 0.658 0.105
Chain 1: 800 -8116.208 0.580 0.105
Chain 1: 900 -7956.402 0.518 0.105
Chain 1: 1000 -7761.436 0.468 0.105
Chain 1: 1100 -7713.799 0.369 0.087
Chain 1: 1200 -7617.058 0.148 0.060
Chain 1: 1300 -7596.155 0.045 0.029
Chain 1: 1400 -7700.713 0.036 0.025
Chain 1: 1500 -7579.040 0.029 0.020
Chain 1: 1600 -7725.110 0.025 0.019
Chain 1: 1700 -7604.168 0.016 0.016
Chain 1: 1800 -7681.873 0.014 0.016
Chain 1: 1900 -7567.506 0.014 0.015
Chain 1: 2000 -7614.914 0.012 0.014
Chain 1: 2100 -7565.075 0.012 0.014
Chain 1: 2200 -7736.229 0.013 0.015
Chain 1: 2300 -7548.065 0.015 0.016
Chain 1: 2400 -7610.853 0.014 0.016
Chain 1: 2500 -7627.098 0.013 0.015
Chain 1: 2600 -7519.545 0.013 0.014
Chain 1: 2700 -7539.856 0.011 0.010
Chain 1: 2800 -7505.354 0.011 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003314 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.14 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86030.187 1.000 1.000
Chain 1: 200 -13761.441 3.126 5.252
Chain 1: 300 -10055.570 2.207 1.000
Chain 1: 400 -11294.174 1.682 1.000
Chain 1: 500 -9019.530 1.396 0.369
Chain 1: 600 -8397.861 1.176 0.369
Chain 1: 700 -8464.428 1.009 0.252
Chain 1: 800 -8902.786 0.889 0.252
Chain 1: 900 -8752.344 0.792 0.110
Chain 1: 1000 -8897.838 0.715 0.110
Chain 1: 1100 -8613.299 0.618 0.074 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8416.208 0.095 0.049
Chain 1: 1300 -8780.555 0.062 0.041
Chain 1: 1400 -8671.227 0.053 0.033
Chain 1: 1500 -8554.868 0.029 0.023
Chain 1: 1600 -8668.190 0.023 0.017
Chain 1: 1700 -8732.939 0.023 0.017
Chain 1: 1800 -8292.511 0.023 0.017
Chain 1: 1900 -8399.204 0.023 0.016
Chain 1: 2000 -8377.457 0.021 0.014
Chain 1: 2100 -8517.361 0.020 0.014
Chain 1: 2200 -8305.657 0.020 0.014
Chain 1: 2300 -8464.930 0.018 0.014
Chain 1: 2400 -8302.793 0.018 0.016
Chain 1: 2500 -8374.064 0.018 0.016
Chain 1: 2600 -8286.036 0.018 0.016
Chain 1: 2700 -8319.575 0.017 0.016
Chain 1: 2800 -8279.107 0.012 0.013
Chain 1: 2900 -8373.264 0.012 0.011
Chain 1: 3000 -8208.843 0.014 0.016
Chain 1: 3100 -8362.131 0.014 0.018
Chain 1: 3200 -8233.606 0.013 0.016
Chain 1: 3300 -8243.524 0.011 0.011
Chain 1: 3400 -8407.542 0.011 0.011
Chain 1: 3500 -8419.124 0.011 0.011
Chain 1: 3600 -8189.947 0.012 0.016
Chain 1: 3700 -8336.996 0.014 0.018
Chain 1: 3800 -8196.095 0.015 0.018
Chain 1: 3900 -8130.263 0.015 0.018
Chain 1: 4000 -8208.401 0.014 0.017
Chain 1: 4100 -8201.681 0.012 0.016
Chain 1: 4200 -8186.334 0.011 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003122 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.22 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8395046.857 1.000 1.000
Chain 1: 200 -1580839.141 2.655 4.311
Chain 1: 300 -890913.924 2.028 1.000
Chain 1: 400 -458420.723 1.757 1.000
Chain 1: 500 -358988.493 1.461 0.943
Chain 1: 600 -233821.132 1.307 0.943
Chain 1: 700 -119823.165 1.256 0.943
Chain 1: 800 -86946.229 1.146 0.943
Chain 1: 900 -67233.515 1.051 0.774
Chain 1: 1000 -51990.964 0.976 0.774
Chain 1: 1100 -39426.113 0.908 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38602.844 0.479 0.378
Chain 1: 1300 -26513.590 0.447 0.378
Chain 1: 1400 -26229.595 0.353 0.319
Chain 1: 1500 -22804.829 0.341 0.319
Chain 1: 1600 -22018.346 0.291 0.293
Chain 1: 1700 -20886.505 0.201 0.293
Chain 1: 1800 -20829.649 0.164 0.150
Chain 1: 1900 -21156.232 0.136 0.054
Chain 1: 2000 -19663.777 0.114 0.054
Chain 1: 2100 -19902.269 0.083 0.036
Chain 1: 2200 -20129.464 0.082 0.036
Chain 1: 2300 -19745.942 0.039 0.019
Chain 1: 2400 -19517.879 0.039 0.019
Chain 1: 2500 -19319.977 0.025 0.015
Chain 1: 2600 -18949.521 0.023 0.015
Chain 1: 2700 -18906.380 0.018 0.012
Chain 1: 2800 -18623.036 0.019 0.015
Chain 1: 2900 -18904.644 0.019 0.015
Chain 1: 3000 -18890.708 0.012 0.012
Chain 1: 3100 -18975.746 0.011 0.012
Chain 1: 3200 -18666.098 0.012 0.015
Chain 1: 3300 -18871.129 0.011 0.012
Chain 1: 3400 -18345.454 0.012 0.015
Chain 1: 3500 -18958.184 0.015 0.015
Chain 1: 3600 -18263.880 0.016 0.015
Chain 1: 3700 -18651.452 0.018 0.017
Chain 1: 3800 -17609.493 0.023 0.021
Chain 1: 3900 -17605.650 0.021 0.021
Chain 1: 4000 -17722.940 0.022 0.021
Chain 1: 4100 -17636.568 0.022 0.021
Chain 1: 4200 -17452.519 0.021 0.021
Chain 1: 4300 -17591.111 0.021 0.021
Chain 1: 4400 -17547.667 0.018 0.011
Chain 1: 4500 -17450.189 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001254 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.54 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49085.493 1.000 1.000
Chain 1: 200 -22705.702 1.081 1.162
Chain 1: 300 -13133.774 0.964 1.000
Chain 1: 400 -25404.495 0.843 1.000
Chain 1: 500 -20362.301 0.724 0.729
Chain 1: 600 -17160.694 0.635 0.729
Chain 1: 700 -13076.635 0.589 0.483
Chain 1: 800 -11870.645 0.528 0.483
Chain 1: 900 -18347.119 0.508 0.353
Chain 1: 1000 -12360.800 0.506 0.483
Chain 1: 1100 -11822.869 0.410 0.353
Chain 1: 1200 -13844.568 0.309 0.312
Chain 1: 1300 -11752.475 0.254 0.248
Chain 1: 1400 -9805.838 0.225 0.199
Chain 1: 1500 -17396.442 0.244 0.199
Chain 1: 1600 -11976.400 0.271 0.312
Chain 1: 1700 -9970.027 0.260 0.201
Chain 1: 1800 -10942.637 0.258 0.201
Chain 1: 1900 -10310.326 0.229 0.199
Chain 1: 2000 -9554.778 0.189 0.178
Chain 1: 2100 -9683.487 0.186 0.178
Chain 1: 2200 -9406.003 0.174 0.178
Chain 1: 2300 -10198.106 0.164 0.089
Chain 1: 2400 -9504.654 0.151 0.079
Chain 1: 2500 -9826.096 0.111 0.078
Chain 1: 2600 -9248.598 0.072 0.073
Chain 1: 2700 -10263.339 0.062 0.073
Chain 1: 2800 -10588.462 0.056 0.062
Chain 1: 2900 -9413.915 0.062 0.073
Chain 1: 3000 -8645.489 0.063 0.073
Chain 1: 3100 -13738.201 0.099 0.078
Chain 1: 3200 -8904.592 0.150 0.089
Chain 1: 3300 -9082.606 0.144 0.089
Chain 1: 3400 -8861.309 0.140 0.089
Chain 1: 3500 -9249.083 0.141 0.089
Chain 1: 3600 -8646.892 0.141 0.089
Chain 1: 3700 -9619.729 0.142 0.089
Chain 1: 3800 -10115.342 0.143 0.089
Chain 1: 3900 -8704.564 0.147 0.089
Chain 1: 4000 -9673.314 0.148 0.100
Chain 1: 4100 -8791.024 0.121 0.100
Chain 1: 4200 -14955.001 0.108 0.100
Chain 1: 4300 -9132.134 0.170 0.100
Chain 1: 4400 -8952.579 0.169 0.100
Chain 1: 4500 -8983.552 0.166 0.100
Chain 1: 4600 -13495.809 0.192 0.101
Chain 1: 4700 -12076.863 0.194 0.117
Chain 1: 4800 -10885.220 0.200 0.117
Chain 1: 4900 -9072.145 0.203 0.117
Chain 1: 5000 -10881.729 0.210 0.166
Chain 1: 5100 -10039.895 0.208 0.166
Chain 1: 5200 -8513.418 0.185 0.166
Chain 1: 5300 -9323.336 0.130 0.117
Chain 1: 5400 -10183.696 0.137 0.117
Chain 1: 5500 -8618.811 0.154 0.166
Chain 1: 5600 -8245.644 0.125 0.117
Chain 1: 5700 -9780.586 0.129 0.157
Chain 1: 5800 -9194.742 0.125 0.157
Chain 1: 5900 -8171.714 0.117 0.125
Chain 1: 6000 -8192.296 0.101 0.087
Chain 1: 6100 -9294.375 0.104 0.119
Chain 1: 6200 -11871.421 0.108 0.119
Chain 1: 6300 -9179.989 0.129 0.125
Chain 1: 6400 -13165.781 0.151 0.157
Chain 1: 6500 -8847.466 0.181 0.157
Chain 1: 6600 -8474.925 0.181 0.157
Chain 1: 6700 -14023.237 0.205 0.217
Chain 1: 6800 -9142.221 0.252 0.293
Chain 1: 6900 -8286.712 0.250 0.293
Chain 1: 7000 -8947.605 0.257 0.293
Chain 1: 7100 -9443.073 0.250 0.293
Chain 1: 7200 -9356.761 0.230 0.293
Chain 1: 7300 -8776.527 0.207 0.103
Chain 1: 7400 -8537.984 0.179 0.074
Chain 1: 7500 -8093.552 0.136 0.066
Chain 1: 7600 -8318.948 0.134 0.066
Chain 1: 7700 -8553.526 0.098 0.055
Chain 1: 7800 -11131.889 0.067 0.055
Chain 1: 7900 -8047.308 0.095 0.055
Chain 1: 8000 -8219.406 0.090 0.052
Chain 1: 8100 -8685.298 0.090 0.054
Chain 1: 8200 -9276.517 0.096 0.055
Chain 1: 8300 -10693.919 0.102 0.055
Chain 1: 8400 -10479.918 0.102 0.055
Chain 1: 8500 -8030.834 0.127 0.064
Chain 1: 8600 -8254.761 0.127 0.064
Chain 1: 8700 -9323.674 0.135 0.115
Chain 1: 8800 -10031.900 0.119 0.071
Chain 1: 8900 -11027.467 0.090 0.071
Chain 1: 9000 -8552.285 0.117 0.090
Chain 1: 9100 -9877.766 0.125 0.115
Chain 1: 9200 -10398.053 0.123 0.115
Chain 1: 9300 -8207.771 0.137 0.115
Chain 1: 9400 -11126.395 0.161 0.134
Chain 1: 9500 -8976.336 0.154 0.134
Chain 1: 9600 -9807.313 0.160 0.134
Chain 1: 9700 -9535.736 0.152 0.134
Chain 1: 9800 -10228.507 0.151 0.134
Chain 1: 9900 -10200.385 0.143 0.134
Chain 1: 10000 -9320.665 0.123 0.094
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001574 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.74 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57813.935 1.000 1.000
Chain 1: 200 -17328.039 1.668 2.336
Chain 1: 300 -8552.347 1.454 1.026
Chain 1: 400 -8144.677 1.103 1.026
Chain 1: 500 -8372.166 0.888 1.000
Chain 1: 600 -8149.659 0.745 1.000
Chain 1: 700 -7942.240 0.642 0.050
Chain 1: 800 -8053.124 0.563 0.050
Chain 1: 900 -7911.190 0.503 0.027
Chain 1: 1000 -7787.415 0.454 0.027
Chain 1: 1100 -7687.772 0.355 0.027
Chain 1: 1200 -7673.445 0.122 0.026
Chain 1: 1300 -7662.640 0.019 0.018
Chain 1: 1400 -7879.522 0.017 0.018
Chain 1: 1500 -7601.540 0.018 0.018
Chain 1: 1600 -7532.712 0.016 0.016
Chain 1: 1700 -7503.698 0.014 0.014
Chain 1: 1800 -7556.910 0.013 0.013
Chain 1: 1900 -7596.860 0.012 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003431 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.31 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86089.957 1.000 1.000
Chain 1: 200 -13209.002 3.259 5.518
Chain 1: 300 -9647.021 2.296 1.000
Chain 1: 400 -10459.284 1.741 1.000
Chain 1: 500 -8575.648 1.437 0.369
Chain 1: 600 -8205.888 1.205 0.369
Chain 1: 700 -8548.415 1.038 0.220
Chain 1: 800 -9076.112 0.916 0.220
Chain 1: 900 -8467.472 0.822 0.078
Chain 1: 1000 -8219.509 0.743 0.078
Chain 1: 1100 -8484.195 0.646 0.072 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8153.991 0.098 0.058
Chain 1: 1300 -8213.495 0.062 0.045
Chain 1: 1400 -8211.217 0.054 0.040
Chain 1: 1500 -8242.570 0.033 0.040
Chain 1: 1600 -8249.144 0.028 0.031
Chain 1: 1700 -8173.314 0.025 0.030
Chain 1: 1800 -8060.751 0.021 0.014
Chain 1: 1900 -8179.609 0.015 0.014
Chain 1: 2000 -8139.726 0.013 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003743 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 37.43 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8431527.670 1.000 1.000
Chain 1: 200 -1589867.553 2.652 4.303
Chain 1: 300 -890583.834 2.029 1.000
Chain 1: 400 -456825.890 1.759 1.000
Chain 1: 500 -356650.015 1.464 0.950
Chain 1: 600 -231650.362 1.310 0.950
Chain 1: 700 -118404.817 1.259 0.950
Chain 1: 800 -85715.998 1.150 0.950
Chain 1: 900 -66166.161 1.055 0.785
Chain 1: 1000 -51042.374 0.979 0.785
Chain 1: 1100 -38597.759 0.911 0.540 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37779.772 0.483 0.381
Chain 1: 1300 -25826.077 0.451 0.381
Chain 1: 1400 -25551.167 0.357 0.322
Chain 1: 1500 -22161.475 0.344 0.322
Chain 1: 1600 -21384.136 0.294 0.296
Chain 1: 1700 -20269.331 0.204 0.295
Chain 1: 1800 -20215.837 0.166 0.153
Chain 1: 1900 -20541.541 0.138 0.055
Chain 1: 2000 -19059.384 0.116 0.055
Chain 1: 2100 -19297.448 0.085 0.036
Chain 1: 2200 -19522.556 0.084 0.036
Chain 1: 2300 -19141.053 0.040 0.020
Chain 1: 2400 -18913.419 0.040 0.020
Chain 1: 2500 -18714.977 0.025 0.016
Chain 1: 2600 -18346.139 0.024 0.016
Chain 1: 2700 -18303.423 0.019 0.012
Chain 1: 2800 -18020.280 0.020 0.016
Chain 1: 2900 -18301.171 0.020 0.015
Chain 1: 3000 -18287.541 0.012 0.012
Chain 1: 3100 -18372.412 0.011 0.012
Chain 1: 3200 -18063.554 0.012 0.015
Chain 1: 3300 -18267.920 0.011 0.012
Chain 1: 3400 -17743.440 0.013 0.015
Chain 1: 3500 -18354.266 0.015 0.016
Chain 1: 3600 -17662.291 0.017 0.016
Chain 1: 3700 -18047.995 0.019 0.017
Chain 1: 3800 -17009.705 0.023 0.021
Chain 1: 3900 -17005.835 0.022 0.021
Chain 1: 4000 -17123.206 0.022 0.021
Chain 1: 4100 -17037.015 0.022 0.021
Chain 1: 4200 -16853.725 0.022 0.021
Chain 1: 4300 -16991.864 0.022 0.021
Chain 1: 4400 -16949.052 0.019 0.011
Chain 1: 4500 -16851.588 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001241 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.41 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48724.464 1.000 1.000
Chain 1: 200 -14012.020 1.739 2.477
Chain 1: 300 -20634.344 1.266 1.000
Chain 1: 400 -13536.375 1.081 1.000
Chain 1: 500 -23152.863 0.948 0.524
Chain 1: 600 -15896.730 0.866 0.524
Chain 1: 700 -15586.985 0.745 0.456
Chain 1: 800 -10643.039 0.710 0.465
Chain 1: 900 -10524.277 0.632 0.456
Chain 1: 1000 -12129.317 0.582 0.456
Chain 1: 1100 -17638.024 0.513 0.415 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -12375.043 0.308 0.415
Chain 1: 1300 -11377.597 0.285 0.415
Chain 1: 1400 -10117.010 0.245 0.312
Chain 1: 1500 -11932.368 0.219 0.152
Chain 1: 1600 -12287.279 0.176 0.132
Chain 1: 1700 -10236.898 0.194 0.152
Chain 1: 1800 -17450.763 0.189 0.152
Chain 1: 1900 -9769.623 0.266 0.200
Chain 1: 2000 -10222.070 0.258 0.200
Chain 1: 2100 -13836.652 0.252 0.200
Chain 1: 2200 -11110.917 0.234 0.200
Chain 1: 2300 -9000.290 0.249 0.235
Chain 1: 2400 -9111.798 0.238 0.235
Chain 1: 2500 -11060.942 0.240 0.235
Chain 1: 2600 -9668.778 0.252 0.235
Chain 1: 2700 -9267.216 0.236 0.235
Chain 1: 2800 -9444.722 0.197 0.176
Chain 1: 2900 -9216.679 0.120 0.144
Chain 1: 3000 -9878.488 0.123 0.144
Chain 1: 3100 -8866.210 0.108 0.114
Chain 1: 3200 -8837.775 0.084 0.067
Chain 1: 3300 -9532.696 0.068 0.067
Chain 1: 3400 -9765.850 0.069 0.067
Chain 1: 3500 -8847.501 0.062 0.067
Chain 1: 3600 -9267.685 0.052 0.045
Chain 1: 3700 -9156.756 0.049 0.045
Chain 1: 3800 -8636.656 0.053 0.060
Chain 1: 3900 -11035.817 0.072 0.067
Chain 1: 4000 -8625.954 0.093 0.073
Chain 1: 4100 -9542.182 0.091 0.073
Chain 1: 4200 -13498.589 0.120 0.096
Chain 1: 4300 -10111.849 0.147 0.104
Chain 1: 4400 -8863.469 0.158 0.141
Chain 1: 4500 -12809.716 0.179 0.217
Chain 1: 4600 -11490.903 0.186 0.217
Chain 1: 4700 -9238.730 0.209 0.244
Chain 1: 4800 -8482.247 0.212 0.244
Chain 1: 4900 -9158.531 0.197 0.244
Chain 1: 5000 -9796.873 0.176 0.141
Chain 1: 5100 -8754.985 0.178 0.141
Chain 1: 5200 -11513.798 0.173 0.141
Chain 1: 5300 -10345.829 0.151 0.119
Chain 1: 5400 -10014.879 0.140 0.115
Chain 1: 5500 -10944.940 0.118 0.113
Chain 1: 5600 -8912.263 0.129 0.113
Chain 1: 5700 -11441.845 0.127 0.113
Chain 1: 5800 -8920.582 0.146 0.119
Chain 1: 5900 -8229.349 0.147 0.119
Chain 1: 6000 -11392.188 0.168 0.221
Chain 1: 6100 -8849.608 0.185 0.228
Chain 1: 6200 -10359.922 0.176 0.221
Chain 1: 6300 -8392.641 0.188 0.228
Chain 1: 6400 -12583.275 0.218 0.234
Chain 1: 6500 -10113.898 0.234 0.244
Chain 1: 6600 -8300.094 0.233 0.244
Chain 1: 6700 -8632.383 0.215 0.244
Chain 1: 6800 -10769.428 0.206 0.234
Chain 1: 6900 -9717.061 0.209 0.234
Chain 1: 7000 -10385.848 0.187 0.219
Chain 1: 7100 -9250.108 0.171 0.198
Chain 1: 7200 -9695.435 0.161 0.198
Chain 1: 7300 -9243.188 0.142 0.123
Chain 1: 7400 -8135.149 0.123 0.123
Chain 1: 7500 -10578.540 0.121 0.123
Chain 1: 7600 -9636.617 0.109 0.108
Chain 1: 7700 -8290.392 0.122 0.123
Chain 1: 7800 -8373.623 0.103 0.108
Chain 1: 7900 -9700.051 0.106 0.123
Chain 1: 8000 -10887.122 0.110 0.123
Chain 1: 8100 -8181.769 0.131 0.136
Chain 1: 8200 -11148.912 0.153 0.137
Chain 1: 8300 -11285.882 0.149 0.137
Chain 1: 8400 -8202.410 0.173 0.162
Chain 1: 8500 -8173.258 0.150 0.137
Chain 1: 8600 -8525.857 0.145 0.137
Chain 1: 8700 -7925.046 0.136 0.109
Chain 1: 8800 -8110.388 0.137 0.109
Chain 1: 8900 -8418.564 0.127 0.076
Chain 1: 9000 -11297.842 0.142 0.076
Chain 1: 9100 -8987.821 0.135 0.076
Chain 1: 9200 -7974.552 0.121 0.076
Chain 1: 9300 -9399.264 0.135 0.127
Chain 1: 9400 -8529.718 0.107 0.102
Chain 1: 9500 -8187.997 0.111 0.102
Chain 1: 9600 -8418.594 0.110 0.102
Chain 1: 9700 -8135.891 0.106 0.102
Chain 1: 9800 -8520.826 0.108 0.102
Chain 1: 9900 -8422.287 0.105 0.102
Chain 1: 10000 -9187.916 0.088 0.083
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.006654 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 66.54 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61614.571 1.000 1.000
Chain 1: 200 -17451.381 1.765 2.531
Chain 1: 300 -8673.858 1.514 1.012
Chain 1: 400 -8140.441 1.152 1.012
Chain 1: 500 -8116.295 0.922 1.000
Chain 1: 600 -8698.664 0.780 1.000
Chain 1: 700 -7682.207 0.687 0.132
Chain 1: 800 -7997.035 0.606 0.132
Chain 1: 900 -7638.857 0.544 0.067
Chain 1: 1000 -7686.800 0.490 0.067
Chain 1: 1100 -7810.262 0.392 0.066
Chain 1: 1200 -7494.414 0.143 0.047
Chain 1: 1300 -7650.608 0.044 0.042
Chain 1: 1400 -7784.231 0.039 0.039
Chain 1: 1500 -7534.383 0.042 0.039
Chain 1: 1600 -7679.106 0.037 0.033
Chain 1: 1700 -7455.052 0.027 0.030
Chain 1: 1800 -7544.064 0.024 0.020
Chain 1: 1900 -7550.571 0.020 0.019
Chain 1: 2000 -7576.172 0.019 0.019
Chain 1: 2100 -7551.247 0.018 0.019
Chain 1: 2200 -7634.552 0.015 0.017
Chain 1: 2300 -7498.811 0.015 0.017
Chain 1: 2400 -7560.835 0.014 0.012
Chain 1: 2500 -7398.727 0.013 0.012
Chain 1: 2600 -7452.654 0.012 0.011
Chain 1: 2700 -7544.145 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003056 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.56 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85422.829 1.000 1.000
Chain 1: 200 -13170.864 3.243 5.486
Chain 1: 300 -9664.845 2.283 1.000
Chain 1: 400 -10487.505 1.732 1.000
Chain 1: 500 -8530.989 1.431 0.363
Chain 1: 600 -8287.039 1.198 0.363
Chain 1: 700 -8448.622 1.029 0.229
Chain 1: 800 -8512.945 0.902 0.229
Chain 1: 900 -8514.815 0.801 0.078
Chain 1: 1000 -8305.451 0.724 0.078
Chain 1: 1100 -8535.013 0.626 0.029 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8216.477 0.082 0.029
Chain 1: 1300 -8269.075 0.046 0.027
Chain 1: 1400 -8343.402 0.039 0.025
Chain 1: 1500 -8288.916 0.017 0.019
Chain 1: 1600 -8296.048 0.014 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002969 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.69 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8392090.914 1.000 1.000
Chain 1: 200 -1580458.786 2.655 4.310
Chain 1: 300 -889385.575 2.029 1.000
Chain 1: 400 -456722.959 1.759 1.000
Chain 1: 500 -357244.199 1.463 0.947
Chain 1: 600 -232504.854 1.308 0.947
Chain 1: 700 -118836.395 1.258 0.947
Chain 1: 800 -86083.297 1.148 0.947
Chain 1: 900 -66432.394 1.054 0.777
Chain 1: 1000 -51220.026 0.978 0.777
Chain 1: 1100 -38698.186 0.910 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37868.733 0.481 0.380
Chain 1: 1300 -25832.010 0.450 0.380
Chain 1: 1400 -25549.485 0.357 0.324
Chain 1: 1500 -22139.205 0.344 0.324
Chain 1: 1600 -21356.024 0.294 0.297
Chain 1: 1700 -20230.846 0.204 0.296
Chain 1: 1800 -20175.024 0.166 0.154
Chain 1: 1900 -20500.505 0.138 0.056
Chain 1: 2000 -19013.738 0.117 0.056
Chain 1: 2100 -19251.840 0.085 0.037
Chain 1: 2200 -19477.835 0.084 0.037
Chain 1: 2300 -19095.661 0.040 0.020
Chain 1: 2400 -18868.003 0.040 0.020
Chain 1: 2500 -18670.164 0.026 0.016
Chain 1: 2600 -18300.955 0.024 0.016
Chain 1: 2700 -18258.134 0.019 0.012
Chain 1: 2800 -17975.349 0.020 0.016
Chain 1: 2900 -18256.280 0.020 0.015
Chain 1: 3000 -18242.480 0.012 0.012
Chain 1: 3100 -18327.357 0.011 0.012
Chain 1: 3200 -18018.513 0.012 0.015
Chain 1: 3300 -18222.892 0.011 0.012
Chain 1: 3400 -17698.678 0.013 0.015
Chain 1: 3500 -18309.282 0.015 0.016
Chain 1: 3600 -17617.634 0.017 0.016
Chain 1: 3700 -18003.192 0.019 0.017
Chain 1: 3800 -16965.540 0.023 0.021
Chain 1: 3900 -16961.781 0.022 0.021
Chain 1: 4000 -17079.049 0.022 0.021
Chain 1: 4100 -16992.942 0.023 0.021
Chain 1: 4200 -16809.786 0.022 0.021
Chain 1: 4300 -16947.744 0.022 0.021
Chain 1: 4400 -16905.032 0.019 0.011
Chain 1: 4500 -16807.676 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001207 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.07 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49382.723 1.000 1.000
Chain 1: 200 -23061.420 1.071 1.141
Chain 1: 300 -14932.986 0.895 1.000
Chain 1: 400 -13535.912 0.697 1.000
Chain 1: 500 -13679.795 0.560 0.544
Chain 1: 600 -14482.923 0.476 0.544
Chain 1: 700 -15540.643 0.418 0.103
Chain 1: 800 -14744.450 0.372 0.103
Chain 1: 900 -19972.429 0.360 0.103
Chain 1: 1000 -13312.698 0.374 0.262
Chain 1: 1100 -10768.860 0.298 0.236
Chain 1: 1200 -11541.824 0.190 0.103
Chain 1: 1300 -12414.880 0.143 0.070
Chain 1: 1400 -12984.990 0.137 0.068
Chain 1: 1500 -10859.806 0.155 0.070
Chain 1: 1600 -10992.891 0.151 0.070
Chain 1: 1700 -13026.477 0.160 0.156
Chain 1: 1800 -10138.793 0.183 0.196
Chain 1: 1900 -10845.291 0.163 0.156
Chain 1: 2000 -10293.724 0.118 0.070
Chain 1: 2100 -18043.543 0.138 0.070
Chain 1: 2200 -10305.179 0.206 0.156
Chain 1: 2300 -16857.320 0.238 0.196
Chain 1: 2400 -9931.359 0.303 0.285
Chain 1: 2500 -10302.206 0.287 0.285
Chain 1: 2600 -13098.829 0.308 0.285
Chain 1: 2700 -15282.203 0.306 0.285
Chain 1: 2800 -12188.962 0.303 0.254
Chain 1: 2900 -14392.701 0.312 0.254
Chain 1: 3000 -9624.000 0.356 0.389
Chain 1: 3100 -10152.519 0.318 0.254
Chain 1: 3200 -10588.058 0.247 0.214
Chain 1: 3300 -12140.576 0.221 0.153
Chain 1: 3400 -13904.701 0.164 0.143
Chain 1: 3500 -10075.362 0.199 0.153
Chain 1: 3600 -9220.912 0.187 0.143
Chain 1: 3700 -20394.862 0.227 0.153
Chain 1: 3800 -9917.494 0.307 0.153
Chain 1: 3900 -14000.482 0.321 0.292
Chain 1: 4000 -9804.677 0.314 0.292
Chain 1: 4100 -9041.154 0.318 0.292
Chain 1: 4200 -10102.149 0.324 0.292
Chain 1: 4300 -16487.518 0.350 0.380
Chain 1: 4400 -8976.391 0.421 0.387
Chain 1: 4500 -10065.232 0.394 0.387
Chain 1: 4600 -8836.423 0.398 0.387
Chain 1: 4700 -12380.651 0.372 0.292
Chain 1: 4800 -17221.753 0.295 0.286
Chain 1: 4900 -12294.279 0.306 0.286
Chain 1: 5000 -13635.660 0.273 0.281
Chain 1: 5100 -8628.088 0.322 0.286
Chain 1: 5200 -11009.482 0.333 0.286
Chain 1: 5300 -13425.331 0.313 0.281
Chain 1: 5400 -13389.952 0.229 0.216
Chain 1: 5500 -8737.951 0.272 0.281
Chain 1: 5600 -8887.847 0.260 0.281
Chain 1: 5700 -8848.652 0.231 0.216
Chain 1: 5800 -8975.801 0.205 0.180
Chain 1: 5900 -16490.060 0.210 0.180
Chain 1: 6000 -8925.877 0.285 0.216
Chain 1: 6100 -13416.710 0.260 0.216
Chain 1: 6200 -9181.207 0.285 0.335
Chain 1: 6300 -8664.363 0.273 0.335
Chain 1: 6400 -12518.314 0.303 0.335
Chain 1: 6500 -9176.187 0.287 0.335
Chain 1: 6600 -11663.044 0.306 0.335
Chain 1: 6700 -13730.329 0.321 0.335
Chain 1: 6800 -8528.447 0.380 0.364
Chain 1: 6900 -8736.595 0.337 0.335
Chain 1: 7000 -8720.128 0.253 0.308
Chain 1: 7100 -8457.883 0.222 0.213
Chain 1: 7200 -8957.160 0.182 0.151
Chain 1: 7300 -8515.788 0.181 0.151
Chain 1: 7400 -8421.115 0.151 0.056
Chain 1: 7500 -9641.490 0.128 0.056
Chain 1: 7600 -11445.479 0.122 0.056
Chain 1: 7700 -12177.259 0.113 0.056
Chain 1: 7800 -9154.230 0.085 0.056
Chain 1: 7900 -8569.772 0.089 0.060
Chain 1: 8000 -8272.036 0.093 0.060
Chain 1: 8100 -11434.129 0.117 0.068
Chain 1: 8200 -10105.265 0.125 0.127
Chain 1: 8300 -13325.961 0.144 0.132
Chain 1: 8400 -8978.247 0.191 0.158
Chain 1: 8500 -11399.357 0.200 0.212
Chain 1: 8600 -9920.623 0.199 0.212
Chain 1: 8700 -10978.896 0.203 0.212
Chain 1: 8800 -8194.481 0.204 0.212
Chain 1: 8900 -8911.493 0.205 0.212
Chain 1: 9000 -12291.405 0.229 0.242
Chain 1: 9100 -8561.238 0.245 0.242
Chain 1: 9200 -9559.824 0.242 0.242
Chain 1: 9300 -8261.458 0.233 0.212
Chain 1: 9400 -8516.642 0.188 0.157
Chain 1: 9500 -8224.027 0.170 0.149
Chain 1: 9600 -9391.828 0.168 0.124
Chain 1: 9700 -8405.798 0.170 0.124
Chain 1: 9800 -8369.992 0.136 0.117
Chain 1: 9900 -10374.573 0.148 0.124
Chain 1: 10000 -12786.403 0.139 0.124
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002403 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 24.03 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58011.869 1.000 1.000
Chain 1: 200 -18199.900 1.594 2.187
Chain 1: 300 -9070.225 1.398 1.007
Chain 1: 400 -8178.182 1.076 1.007
Chain 1: 500 -8489.024 0.868 1.000
Chain 1: 600 -8588.651 0.725 1.000
Chain 1: 700 -8860.573 0.626 0.109
Chain 1: 800 -8307.387 0.556 0.109
Chain 1: 900 -8221.978 0.495 0.067
Chain 1: 1000 -8129.397 0.447 0.067
Chain 1: 1100 -7669.848 0.353 0.060
Chain 1: 1200 -7590.521 0.135 0.037
Chain 1: 1300 -7795.181 0.037 0.031
Chain 1: 1400 -7676.558 0.028 0.026
Chain 1: 1500 -7702.242 0.025 0.015
Chain 1: 1600 -7717.459 0.024 0.015
Chain 1: 1700 -7577.498 0.022 0.015
Chain 1: 1800 -7570.491 0.016 0.011
Chain 1: 1900 -7626.983 0.016 0.011
Chain 1: 2000 -7768.470 0.016 0.015
Chain 1: 2100 -7558.155 0.013 0.015
Chain 1: 2200 -7793.070 0.015 0.018
Chain 1: 2300 -7633.418 0.014 0.018
Chain 1: 2400 -7707.342 0.014 0.018
Chain 1: 2500 -7620.603 0.015 0.018
Chain 1: 2600 -7534.179 0.016 0.018
Chain 1: 2700 -7527.087 0.014 0.011
Chain 1: 2800 -7665.428 0.016 0.018
Chain 1: 2900 -7420.188 0.018 0.018
Chain 1: 3000 -7547.661 0.018 0.018
Chain 1: 3100 -7544.223 0.015 0.017
Chain 1: 3200 -7731.147 0.015 0.017
Chain 1: 3300 -7466.650 0.016 0.017
Chain 1: 3400 -7668.259 0.018 0.018
Chain 1: 3500 -7451.827 0.020 0.024
Chain 1: 3600 -7504.162 0.019 0.024
Chain 1: 3700 -7467.297 0.020 0.024
Chain 1: 3800 -7438.339 0.018 0.024
Chain 1: 3900 -7409.742 0.015 0.017
Chain 1: 4000 -7405.251 0.014 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002884 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 28.84 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85579.034 1.000 1.000
Chain 1: 200 -14118.015 3.031 5.062
Chain 1: 300 -10232.892 2.147 1.000
Chain 1: 400 -12727.028 1.659 1.000
Chain 1: 500 -9934.995 1.384 0.380
Chain 1: 600 -8536.387 1.180 0.380
Chain 1: 700 -8287.031 1.016 0.281
Chain 1: 800 -8726.196 0.895 0.281
Chain 1: 900 -8832.531 0.797 0.196
Chain 1: 1000 -8785.632 0.718 0.196
Chain 1: 1100 -8981.527 0.620 0.164 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8328.245 0.122 0.078
Chain 1: 1300 -8846.250 0.090 0.059
Chain 1: 1400 -8541.397 0.074 0.050
Chain 1: 1500 -8612.384 0.046 0.036
Chain 1: 1600 -8690.941 0.031 0.030
Chain 1: 1700 -8738.763 0.028 0.022
Chain 1: 1800 -8275.575 0.029 0.022
Chain 1: 1900 -8365.096 0.029 0.022
Chain 1: 2000 -8378.629 0.029 0.022
Chain 1: 2100 -8520.011 0.028 0.017
Chain 1: 2200 -8234.966 0.024 0.017
Chain 1: 2300 -8323.412 0.019 0.011
Chain 1: 2400 -8417.110 0.016 0.011
Chain 1: 2500 -8313.586 0.017 0.011
Chain 1: 2600 -8364.275 0.017 0.011
Chain 1: 2700 -8271.068 0.017 0.011
Chain 1: 2800 -8239.627 0.012 0.011
Chain 1: 2900 -8325.325 0.012 0.011
Chain 1: 3000 -8256.593 0.013 0.011
Chain 1: 3100 -8211.361 0.011 0.011
Chain 1: 3200 -8170.160 0.008 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003371 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.71 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8363565.627 1.000 1.000
Chain 1: 200 -1578638.019 2.649 4.298
Chain 1: 300 -890983.448 2.023 1.000
Chain 1: 400 -458169.604 1.754 1.000
Chain 1: 500 -359337.241 1.458 0.945
Chain 1: 600 -234363.919 1.304 0.945
Chain 1: 700 -120310.754 1.253 0.945
Chain 1: 800 -87473.915 1.143 0.945
Chain 1: 900 -67753.322 1.049 0.772
Chain 1: 1000 -52507.628 0.973 0.772
Chain 1: 1100 -39925.273 0.904 0.533 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39107.595 0.477 0.375
Chain 1: 1300 -26970.179 0.444 0.375
Chain 1: 1400 -26687.261 0.351 0.315
Chain 1: 1500 -23249.427 0.338 0.315
Chain 1: 1600 -22461.009 0.288 0.291
Chain 1: 1700 -21321.537 0.199 0.290
Chain 1: 1800 -21263.651 0.162 0.148
Chain 1: 1900 -21591.227 0.134 0.053
Chain 1: 2000 -20093.136 0.113 0.053
Chain 1: 2100 -20331.938 0.082 0.035
Chain 1: 2200 -20560.591 0.081 0.035
Chain 1: 2300 -20175.530 0.038 0.019
Chain 1: 2400 -19947.004 0.038 0.019
Chain 1: 2500 -19749.518 0.024 0.015
Chain 1: 2600 -19377.801 0.023 0.015
Chain 1: 2700 -19334.227 0.018 0.012
Chain 1: 2800 -19050.675 0.019 0.015
Chain 1: 2900 -19332.696 0.019 0.015
Chain 1: 3000 -19318.670 0.012 0.012
Chain 1: 3100 -19403.902 0.011 0.011
Chain 1: 3200 -19093.540 0.011 0.015
Chain 1: 3300 -19299.114 0.010 0.011
Chain 1: 3400 -18772.338 0.012 0.015
Chain 1: 3500 -19386.927 0.014 0.015
Chain 1: 3600 -18690.134 0.016 0.015
Chain 1: 3700 -19079.576 0.018 0.016
Chain 1: 3800 -18034.002 0.022 0.020
Chain 1: 3900 -18030.103 0.021 0.020
Chain 1: 4000 -18147.327 0.021 0.020
Chain 1: 4100 -18060.860 0.021 0.020
Chain 1: 4200 -17875.971 0.021 0.020
Chain 1: 4300 -18015.119 0.020 0.020
Chain 1: 4400 -17970.981 0.018 0.010
Chain 1: 4500 -17873.423 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001302 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.02 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48441.371 1.000 1.000
Chain 1: 200 -16668.682 1.453 1.906
Chain 1: 300 -17301.461 0.981 1.000
Chain 1: 400 -15911.646 0.758 1.000
Chain 1: 500 -11309.083 0.687 0.407
Chain 1: 600 -11756.919 0.579 0.407
Chain 1: 700 -11365.311 0.501 0.087
Chain 1: 800 -10930.840 0.444 0.087
Chain 1: 900 -10474.852 0.399 0.044
Chain 1: 1000 -12776.234 0.377 0.087
Chain 1: 1100 -21221.489 0.317 0.087
Chain 1: 1200 -13431.879 0.184 0.087
Chain 1: 1300 -12702.967 0.187 0.087
Chain 1: 1400 -18918.281 0.211 0.180
Chain 1: 1500 -10619.265 0.248 0.180
Chain 1: 1600 -10784.591 0.246 0.180
Chain 1: 1700 -10181.661 0.248 0.180
Chain 1: 1800 -12814.772 0.265 0.205
Chain 1: 1900 -10547.024 0.282 0.215
Chain 1: 2000 -11012.828 0.268 0.215
Chain 1: 2100 -13402.681 0.246 0.205
Chain 1: 2200 -10146.780 0.220 0.205
Chain 1: 2300 -10965.107 0.222 0.205
Chain 1: 2400 -8547.669 0.218 0.205
Chain 1: 2500 -8933.601 0.144 0.178
Chain 1: 2600 -9373.685 0.147 0.178
Chain 1: 2700 -8782.252 0.148 0.178
Chain 1: 2800 -10363.751 0.142 0.153
Chain 1: 2900 -10342.590 0.121 0.075
Chain 1: 3000 -10117.827 0.119 0.075
Chain 1: 3100 -15069.216 0.134 0.075
Chain 1: 3200 -17206.932 0.114 0.075
Chain 1: 3300 -9578.216 0.187 0.124
Chain 1: 3400 -8851.956 0.167 0.082
Chain 1: 3500 -11028.533 0.182 0.124
Chain 1: 3600 -8740.631 0.203 0.153
Chain 1: 3700 -9102.837 0.201 0.153
Chain 1: 3800 -9597.471 0.191 0.124
Chain 1: 3900 -11207.488 0.205 0.144
Chain 1: 4000 -13870.032 0.222 0.192
Chain 1: 4100 -9337.021 0.237 0.192
Chain 1: 4200 -8869.832 0.230 0.192
Chain 1: 4300 -8494.237 0.155 0.144
Chain 1: 4400 -12663.693 0.180 0.192
Chain 1: 4500 -14605.687 0.173 0.144
Chain 1: 4600 -8394.495 0.221 0.144
Chain 1: 4700 -10623.064 0.238 0.192
Chain 1: 4800 -8432.961 0.259 0.210
Chain 1: 4900 -8844.714 0.249 0.210
Chain 1: 5000 -10690.408 0.247 0.210
Chain 1: 5100 -8837.858 0.220 0.210
Chain 1: 5200 -12924.533 0.246 0.210
Chain 1: 5300 -12633.034 0.244 0.210
Chain 1: 5400 -8668.621 0.257 0.210
Chain 1: 5500 -12896.802 0.276 0.260
Chain 1: 5600 -10023.398 0.231 0.260
Chain 1: 5700 -8474.266 0.228 0.260
Chain 1: 5800 -8563.690 0.203 0.210
Chain 1: 5900 -8848.854 0.202 0.210
Chain 1: 6000 -8400.121 0.190 0.210
Chain 1: 6100 -8378.566 0.169 0.183
Chain 1: 6200 -8558.955 0.140 0.053
Chain 1: 6300 -8277.858 0.141 0.053
Chain 1: 6400 -9090.159 0.104 0.053
Chain 1: 6500 -9085.180 0.071 0.034
Chain 1: 6600 -8261.818 0.053 0.034
Chain 1: 6700 -10376.341 0.055 0.034
Chain 1: 6800 -8758.740 0.072 0.053
Chain 1: 6900 -8733.387 0.069 0.053
Chain 1: 7000 -8809.874 0.065 0.034
Chain 1: 7100 -8001.470 0.075 0.089
Chain 1: 7200 -8086.571 0.074 0.089
Chain 1: 7300 -10379.995 0.092 0.100
Chain 1: 7400 -9161.501 0.097 0.101
Chain 1: 7500 -7956.804 0.112 0.133
Chain 1: 7600 -8268.972 0.105 0.133
Chain 1: 7700 -8550.494 0.088 0.101
Chain 1: 7800 -8320.146 0.073 0.038
Chain 1: 7900 -8184.657 0.074 0.038
Chain 1: 8000 -7968.407 0.076 0.038
Chain 1: 8100 -8267.619 0.069 0.036
Chain 1: 8200 -10045.889 0.086 0.038
Chain 1: 8300 -8079.995 0.088 0.038
Chain 1: 8400 -8353.288 0.078 0.036
Chain 1: 8500 -8102.879 0.066 0.033
Chain 1: 8600 -8759.152 0.070 0.033
Chain 1: 8700 -8312.585 0.072 0.036
Chain 1: 8800 -10000.748 0.086 0.054
Chain 1: 8900 -10901.373 0.093 0.075
Chain 1: 9000 -9106.112 0.110 0.083
Chain 1: 9100 -8074.552 0.119 0.128
Chain 1: 9200 -7749.374 0.105 0.083
Chain 1: 9300 -10208.536 0.105 0.083
Chain 1: 9400 -7931.955 0.131 0.128
Chain 1: 9500 -8091.765 0.129 0.128
Chain 1: 9600 -8010.593 0.123 0.128
Chain 1: 9700 -9713.517 0.135 0.169
Chain 1: 9800 -11581.287 0.134 0.161
Chain 1: 9900 -8578.745 0.161 0.175
Chain 1: 10000 -8997.507 0.146 0.161
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00139 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.9 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56450.752 1.000 1.000
Chain 1: 200 -17027.907 1.658 2.315
Chain 1: 300 -8561.071 1.435 1.000
Chain 1: 400 -8841.531 1.084 1.000
Chain 1: 500 -8284.690 0.881 0.989
Chain 1: 600 -8699.590 0.742 0.989
Chain 1: 700 -7797.164 0.652 0.116
Chain 1: 800 -8526.126 0.582 0.116
Chain 1: 900 -7899.868 0.526 0.085
Chain 1: 1000 -7893.125 0.473 0.085
Chain 1: 1100 -7835.403 0.374 0.079
Chain 1: 1200 -7592.271 0.146 0.067
Chain 1: 1300 -7635.794 0.047 0.048
Chain 1: 1400 -7925.844 0.048 0.048
Chain 1: 1500 -7607.189 0.045 0.042
Chain 1: 1600 -7496.312 0.042 0.037
Chain 1: 1700 -7482.819 0.031 0.032
Chain 1: 1800 -7545.292 0.023 0.015
Chain 1: 1900 -7589.030 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.0036 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 36 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86432.738 1.000 1.000
Chain 1: 200 -13137.471 3.290 5.579
Chain 1: 300 -9588.342 2.316 1.000
Chain 1: 400 -10482.004 1.759 1.000
Chain 1: 500 -8491.741 1.454 0.370
Chain 1: 600 -8142.833 1.219 0.370
Chain 1: 700 -8243.785 1.046 0.234
Chain 1: 800 -8848.327 0.924 0.234
Chain 1: 900 -8411.385 0.827 0.085
Chain 1: 1000 -8292.492 0.746 0.085
Chain 1: 1100 -8486.527 0.648 0.068 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8201.082 0.094 0.052
Chain 1: 1300 -8351.900 0.059 0.043
Chain 1: 1400 -8319.651 0.050 0.035
Chain 1: 1500 -8207.723 0.028 0.023
Chain 1: 1600 -8309.361 0.025 0.018
Chain 1: 1700 -8396.341 0.025 0.018
Chain 1: 1800 -8009.105 0.023 0.018
Chain 1: 1900 -8111.686 0.019 0.014
Chain 1: 2000 -8081.558 0.018 0.014
Chain 1: 2100 -8213.338 0.017 0.014
Chain 1: 2200 -7999.263 0.017 0.014
Chain 1: 2300 -8140.988 0.017 0.014
Chain 1: 2400 -8153.336 0.016 0.014
Chain 1: 2500 -8121.414 0.015 0.013
Chain 1: 2600 -8121.159 0.014 0.013
Chain 1: 2700 -8029.398 0.014 0.013
Chain 1: 2800 -8005.426 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003016 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.16 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8408926.555 1.000 1.000
Chain 1: 200 -1584872.277 2.653 4.306
Chain 1: 300 -890182.543 2.029 1.000
Chain 1: 400 -457166.132 1.758 1.000
Chain 1: 500 -357406.127 1.462 0.947
Chain 1: 600 -232359.228 1.308 0.947
Chain 1: 700 -118691.200 1.258 0.947
Chain 1: 800 -85941.877 1.149 0.947
Chain 1: 900 -66307.831 1.054 0.780
Chain 1: 1000 -51121.184 0.978 0.780
Chain 1: 1100 -38620.866 0.911 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37795.035 0.482 0.381
Chain 1: 1300 -25781.444 0.451 0.381
Chain 1: 1400 -25500.860 0.357 0.324
Chain 1: 1500 -22096.643 0.345 0.324
Chain 1: 1600 -21315.087 0.295 0.297
Chain 1: 1700 -20192.806 0.204 0.296
Chain 1: 1800 -20137.569 0.166 0.154
Chain 1: 1900 -20463.212 0.138 0.056
Chain 1: 2000 -18977.560 0.117 0.056
Chain 1: 2100 -19215.681 0.085 0.037
Chain 1: 2200 -19441.490 0.084 0.037
Chain 1: 2300 -19059.384 0.040 0.020
Chain 1: 2400 -18831.701 0.040 0.020
Chain 1: 2500 -18633.661 0.026 0.016
Chain 1: 2600 -18264.452 0.024 0.016
Chain 1: 2700 -18221.622 0.019 0.012
Chain 1: 2800 -17938.716 0.020 0.016
Chain 1: 2900 -18219.648 0.020 0.015
Chain 1: 3000 -18205.896 0.012 0.012
Chain 1: 3100 -18290.803 0.011 0.012
Chain 1: 3200 -17981.858 0.012 0.015
Chain 1: 3300 -18186.298 0.011 0.012
Chain 1: 3400 -17661.886 0.013 0.015
Chain 1: 3500 -18272.737 0.015 0.016
Chain 1: 3600 -17580.733 0.017 0.016
Chain 1: 3700 -17966.530 0.019 0.017
Chain 1: 3800 -16928.287 0.023 0.021
Chain 1: 3900 -16924.479 0.022 0.021
Chain 1: 4000 -17041.780 0.023 0.021
Chain 1: 4100 -16955.636 0.023 0.021
Chain 1: 4200 -16772.346 0.022 0.021
Chain 1: 4300 -16910.426 0.022 0.021
Chain 1: 4400 -16867.611 0.019 0.011
Chain 1: 4500 -16770.200 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001292 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.92 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49438.146 1.000 1.000
Chain 1: 200 -23925.020 1.033 1.066
Chain 1: 300 -19014.166 0.775 1.000
Chain 1: 400 -14689.822 0.655 1.000
Chain 1: 500 -16196.567 0.542 0.294
Chain 1: 600 -11914.768 0.512 0.359
Chain 1: 700 -15724.745 0.473 0.294
Chain 1: 800 -14730.842 0.423 0.294
Chain 1: 900 -14966.643 0.377 0.258
Chain 1: 1000 -13249.129 0.353 0.258
Chain 1: 1100 -14025.189 0.258 0.242
Chain 1: 1200 -27335.242 0.200 0.242
Chain 1: 1300 -21511.803 0.201 0.242
Chain 1: 1400 -11841.417 0.254 0.242
Chain 1: 1500 -12228.840 0.248 0.242
Chain 1: 1600 -10580.968 0.227 0.156
Chain 1: 1700 -15950.768 0.237 0.156
Chain 1: 1800 -10547.342 0.281 0.271
Chain 1: 1900 -10064.308 0.284 0.271
Chain 1: 2000 -11127.141 0.281 0.271
Chain 1: 2100 -10214.701 0.284 0.271
Chain 1: 2200 -11379.308 0.246 0.156
Chain 1: 2300 -11010.074 0.222 0.102
Chain 1: 2400 -16265.468 0.173 0.102
Chain 1: 2500 -10896.746 0.219 0.156
Chain 1: 2600 -16183.417 0.236 0.323
Chain 1: 2700 -15335.410 0.208 0.102
Chain 1: 2800 -9425.680 0.219 0.102
Chain 1: 2900 -9227.487 0.217 0.102
Chain 1: 3000 -9621.037 0.211 0.102
Chain 1: 3100 -10141.267 0.207 0.102
Chain 1: 3200 -9412.809 0.205 0.077
Chain 1: 3300 -10439.522 0.211 0.098
Chain 1: 3400 -10290.068 0.181 0.077
Chain 1: 3500 -10001.697 0.134 0.055
Chain 1: 3600 -10316.594 0.105 0.051
Chain 1: 3700 -9703.709 0.105 0.051
Chain 1: 3800 -9927.647 0.045 0.041
Chain 1: 3900 -13499.745 0.069 0.051
Chain 1: 4000 -16738.259 0.084 0.063
Chain 1: 4100 -14537.475 0.094 0.077
Chain 1: 4200 -9196.811 0.145 0.098
Chain 1: 4300 -10451.889 0.147 0.120
Chain 1: 4400 -12735.627 0.163 0.151
Chain 1: 4500 -9371.189 0.196 0.179
Chain 1: 4600 -9334.669 0.194 0.179
Chain 1: 4700 -9515.995 0.189 0.179
Chain 1: 4800 -12347.941 0.210 0.193
Chain 1: 4900 -13683.711 0.193 0.179
Chain 1: 5000 -10325.723 0.207 0.179
Chain 1: 5100 -8816.706 0.209 0.179
Chain 1: 5200 -9038.327 0.153 0.171
Chain 1: 5300 -10054.486 0.151 0.171
Chain 1: 5400 -8668.472 0.149 0.160
Chain 1: 5500 -12688.793 0.145 0.160
Chain 1: 5600 -12510.688 0.146 0.160
Chain 1: 5700 -9817.387 0.171 0.171
Chain 1: 5800 -9833.258 0.149 0.160
Chain 1: 5900 -11612.199 0.154 0.160
Chain 1: 6000 -8741.642 0.155 0.160
Chain 1: 6100 -12494.399 0.167 0.160
Chain 1: 6200 -9099.313 0.202 0.274
Chain 1: 6300 -8855.560 0.195 0.274
Chain 1: 6400 -8857.908 0.179 0.274
Chain 1: 6500 -8717.096 0.149 0.153
Chain 1: 6600 -9632.127 0.157 0.153
Chain 1: 6700 -8867.083 0.138 0.095
Chain 1: 6800 -9089.801 0.140 0.095
Chain 1: 6900 -8964.987 0.127 0.086
Chain 1: 7000 -15774.308 0.137 0.086
Chain 1: 7100 -9546.583 0.172 0.086
Chain 1: 7200 -8593.684 0.146 0.086
Chain 1: 7300 -8787.664 0.145 0.086
Chain 1: 7400 -8936.738 0.147 0.086
Chain 1: 7500 -9505.040 0.151 0.086
Chain 1: 7600 -11760.043 0.161 0.086
Chain 1: 7700 -8830.618 0.186 0.111
Chain 1: 7800 -11703.041 0.208 0.192
Chain 1: 7900 -8471.318 0.244 0.245
Chain 1: 8000 -8594.170 0.203 0.192
Chain 1: 8100 -10022.099 0.152 0.142
Chain 1: 8200 -12081.181 0.158 0.170
Chain 1: 8300 -8912.263 0.191 0.192
Chain 1: 8400 -13229.450 0.222 0.245
Chain 1: 8500 -8679.354 0.268 0.326
Chain 1: 8600 -11361.234 0.273 0.326
Chain 1: 8700 -9808.584 0.255 0.245
Chain 1: 8800 -12228.427 0.251 0.236
Chain 1: 8900 -10829.467 0.225 0.198
Chain 1: 9000 -9387.034 0.239 0.198
Chain 1: 9100 -8585.100 0.235 0.198
Chain 1: 9200 -11453.670 0.243 0.236
Chain 1: 9300 -11312.607 0.208 0.198
Chain 1: 9400 -8518.150 0.208 0.198
Chain 1: 9500 -8821.439 0.159 0.158
Chain 1: 9600 -9178.088 0.140 0.154
Chain 1: 9700 -11216.867 0.142 0.154
Chain 1: 9800 -8485.659 0.154 0.154
Chain 1: 9900 -9663.004 0.154 0.154
Chain 1: 10000 -8396.149 0.153 0.151
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001384 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.84 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57305.506 1.000 1.000
Chain 1: 200 -17756.809 1.614 2.227
Chain 1: 300 -8927.202 1.405 1.000
Chain 1: 400 -8284.818 1.073 1.000
Chain 1: 500 -8794.728 0.870 0.989
Chain 1: 600 -8896.861 0.727 0.989
Chain 1: 700 -8326.607 0.633 0.078
Chain 1: 800 -8294.510 0.554 0.078
Chain 1: 900 -8111.916 0.495 0.068
Chain 1: 1000 -7670.113 0.452 0.068
Chain 1: 1100 -7662.201 0.352 0.058
Chain 1: 1200 -7685.143 0.129 0.058
Chain 1: 1300 -7804.349 0.032 0.023
Chain 1: 1400 -7935.643 0.026 0.017
Chain 1: 1500 -7656.586 0.024 0.017
Chain 1: 1600 -7827.433 0.025 0.022
Chain 1: 1700 -7578.357 0.021 0.022
Chain 1: 1800 -7791.116 0.023 0.023
Chain 1: 1900 -7692.205 0.022 0.022
Chain 1: 2000 -7760.890 0.018 0.017
Chain 1: 2100 -7586.697 0.020 0.022
Chain 1: 2200 -7997.353 0.025 0.023
Chain 1: 2300 -7647.248 0.028 0.027
Chain 1: 2400 -7741.947 0.027 0.027
Chain 1: 2500 -7652.795 0.025 0.023
Chain 1: 2600 -7621.515 0.023 0.023
Chain 1: 2700 -7618.693 0.020 0.013
Chain 1: 2800 -7610.474 0.017 0.012
Chain 1: 2900 -7492.798 0.017 0.012
Chain 1: 3000 -7636.149 0.018 0.016
Chain 1: 3100 -7625.217 0.016 0.012
Chain 1: 3200 -7823.835 0.014 0.012
Chain 1: 3300 -7549.338 0.013 0.012
Chain 1: 3400 -7768.391 0.014 0.016
Chain 1: 3500 -7532.534 0.016 0.019
Chain 1: 3600 -7598.764 0.017 0.019
Chain 1: 3700 -7547.526 0.017 0.019
Chain 1: 3800 -7546.243 0.017 0.019
Chain 1: 3900 -7513.562 0.016 0.019
Chain 1: 4000 -7507.821 0.014 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003182 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.82 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86919.484 1.000 1.000
Chain 1: 200 -13854.217 3.137 5.274
Chain 1: 300 -10155.561 2.213 1.000
Chain 1: 400 -11165.339 1.682 1.000
Chain 1: 500 -9073.143 1.392 0.364
Chain 1: 600 -9307.770 1.164 0.364
Chain 1: 700 -9509.913 1.001 0.231
Chain 1: 800 -8528.390 0.890 0.231
Chain 1: 900 -8478.437 0.792 0.115
Chain 1: 1000 -9243.970 0.721 0.115
Chain 1: 1100 -8693.817 0.627 0.090 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -9067.520 0.104 0.083
Chain 1: 1300 -8578.624 0.073 0.063
Chain 1: 1400 -8648.513 0.065 0.057
Chain 1: 1500 -8604.389 0.042 0.041
Chain 1: 1600 -8619.631 0.040 0.041
Chain 1: 1700 -8505.567 0.039 0.041
Chain 1: 1800 -8559.605 0.028 0.013
Chain 1: 1900 -8434.070 0.029 0.015
Chain 1: 2000 -8497.773 0.022 0.013
Chain 1: 2100 -8637.075 0.017 0.013
Chain 1: 2200 -8438.718 0.015 0.013
Chain 1: 2300 -8590.555 0.011 0.013
Chain 1: 2400 -8430.895 0.013 0.015
Chain 1: 2500 -8500.110 0.013 0.015
Chain 1: 2600 -8414.155 0.014 0.015
Chain 1: 2700 -8446.580 0.013 0.015
Chain 1: 2800 -8407.454 0.013 0.015
Chain 1: 2900 -8499.618 0.012 0.011
Chain 1: 3000 -8325.726 0.013 0.016
Chain 1: 3100 -8489.580 0.014 0.018
Chain 1: 3200 -8362.447 0.013 0.015
Chain 1: 3300 -8371.550 0.011 0.011
Chain 1: 3400 -8523.250 0.011 0.011
Chain 1: 3500 -8514.747 0.010 0.011
Chain 1: 3600 -8319.708 0.012 0.015
Chain 1: 3700 -8462.863 0.013 0.017
Chain 1: 3800 -8326.552 0.014 0.017
Chain 1: 3900 -8261.765 0.014 0.017
Chain 1: 4000 -8336.019 0.013 0.016
Chain 1: 4100 -8327.094 0.011 0.015
Chain 1: 4200 -8315.639 0.010 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003125 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.25 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8402658.006 1.000 1.000
Chain 1: 200 -1582330.340 2.655 4.310
Chain 1: 300 -889862.549 2.029 1.000
Chain 1: 400 -457363.365 1.759 1.000
Chain 1: 500 -357921.058 1.462 0.946
Chain 1: 600 -232960.427 1.308 0.946
Chain 1: 700 -119397.964 1.257 0.946
Chain 1: 800 -86690.562 1.147 0.946
Chain 1: 900 -67068.289 1.052 0.778
Chain 1: 1000 -51897.830 0.976 0.778
Chain 1: 1100 -39404.490 0.908 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38586.743 0.479 0.377
Chain 1: 1300 -26557.615 0.446 0.377
Chain 1: 1400 -26280.221 0.353 0.317
Chain 1: 1500 -22871.095 0.340 0.317
Chain 1: 1600 -22089.845 0.290 0.293
Chain 1: 1700 -20964.312 0.200 0.292
Chain 1: 1800 -20909.119 0.163 0.149
Chain 1: 1900 -21235.625 0.135 0.054
Chain 1: 2000 -19746.824 0.113 0.054
Chain 1: 2100 -19985.125 0.083 0.035
Chain 1: 2200 -20211.825 0.082 0.035
Chain 1: 2300 -19828.714 0.038 0.019
Chain 1: 2400 -19600.677 0.039 0.019
Chain 1: 2500 -19402.752 0.025 0.015
Chain 1: 2600 -19032.534 0.023 0.015
Chain 1: 2700 -18989.456 0.018 0.012
Chain 1: 2800 -18706.210 0.019 0.015
Chain 1: 2900 -18987.537 0.019 0.015
Chain 1: 3000 -18973.685 0.012 0.012
Chain 1: 3100 -19058.742 0.011 0.012
Chain 1: 3200 -18749.207 0.011 0.015
Chain 1: 3300 -18954.123 0.011 0.012
Chain 1: 3400 -18428.681 0.012 0.015
Chain 1: 3500 -19041.150 0.014 0.015
Chain 1: 3600 -18346.999 0.016 0.015
Chain 1: 3700 -18734.387 0.018 0.017
Chain 1: 3800 -17692.913 0.023 0.021
Chain 1: 3900 -17689.037 0.021 0.021
Chain 1: 4000 -17806.309 0.022 0.021
Chain 1: 4100 -17720.030 0.022 0.021
Chain 1: 4200 -17536.038 0.021 0.021
Chain 1: 4300 -17674.610 0.021 0.021
Chain 1: 4400 -17631.196 0.018 0.010
Chain 1: 4500 -17533.695 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001334 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.34 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12503.258 1.000 1.000
Chain 1: 200 -9165.087 0.682 1.000
Chain 1: 300 -8037.208 0.502 0.364
Chain 1: 400 -8191.162 0.381 0.364
Chain 1: 500 -7775.799 0.315 0.140
Chain 1: 600 -7913.410 0.266 0.140
Chain 1: 700 -7849.238 0.229 0.053
Chain 1: 800 -7882.590 0.201 0.053
Chain 1: 900 -7956.655 0.180 0.019
Chain 1: 1000 -7923.690 0.162 0.019
Chain 1: 1100 -7865.457 0.063 0.017
Chain 1: 1200 -7848.960 0.027 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001578 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.78 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61904.587 1.000 1.000
Chain 1: 200 -18130.276 1.707 2.414
Chain 1: 300 -8952.805 1.480 1.025
Chain 1: 400 -9398.276 1.122 1.025
Chain 1: 500 -7859.425 0.937 1.000
Chain 1: 600 -8756.813 0.798 1.000
Chain 1: 700 -7867.561 0.700 0.196
Chain 1: 800 -7810.125 0.613 0.196
Chain 1: 900 -7654.680 0.547 0.113
Chain 1: 1000 -7857.304 0.495 0.113
Chain 1: 1100 -7874.013 0.395 0.102
Chain 1: 1200 -7764.527 0.155 0.047
Chain 1: 1300 -7840.689 0.054 0.026
Chain 1: 1400 -7654.545 0.052 0.024
Chain 1: 1500 -7542.832 0.033 0.020
Chain 1: 1600 -7770.390 0.026 0.020
Chain 1: 1700 -7575.146 0.017 0.020
Chain 1: 1800 -7640.315 0.017 0.020
Chain 1: 1900 -7575.915 0.016 0.015
Chain 1: 2000 -7657.161 0.015 0.014
Chain 1: 2100 -7571.831 0.016 0.014
Chain 1: 2200 -7705.241 0.016 0.015
Chain 1: 2300 -7551.613 0.017 0.017
Chain 1: 2400 -7574.085 0.015 0.015
Chain 1: 2500 -7599.582 0.014 0.011
Chain 1: 2600 -7506.396 0.012 0.011
Chain 1: 2700 -7530.686 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003147 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.47 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86841.842 1.000 1.000
Chain 1: 200 -13695.343 3.170 5.341
Chain 1: 300 -9952.147 2.239 1.000
Chain 1: 400 -11540.348 1.714 1.000
Chain 1: 500 -8784.775 1.434 0.376
Chain 1: 600 -8488.284 1.201 0.376
Chain 1: 700 -8471.070 1.029 0.314
Chain 1: 800 -8554.860 0.902 0.314
Chain 1: 900 -8652.246 0.803 0.138
Chain 1: 1000 -8520.411 0.724 0.138
Chain 1: 1100 -8727.800 0.627 0.035 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8255.502 0.098 0.035
Chain 1: 1300 -8584.203 0.064 0.035
Chain 1: 1400 -8591.925 0.051 0.024
Chain 1: 1500 -8436.621 0.021 0.018
Chain 1: 1600 -8549.983 0.019 0.015
Chain 1: 1700 -8604.608 0.019 0.015
Chain 1: 1800 -8155.874 0.024 0.018
Chain 1: 1900 -8265.157 0.024 0.018
Chain 1: 2000 -8249.356 0.023 0.018
Chain 1: 2100 -8387.793 0.022 0.017
Chain 1: 2200 -8160.354 0.019 0.017
Chain 1: 2300 -8260.475 0.017 0.013
Chain 1: 2400 -8332.709 0.017 0.013
Chain 1: 2500 -8272.159 0.016 0.013
Chain 1: 2600 -8288.903 0.015 0.012
Chain 1: 2700 -8195.382 0.016 0.012
Chain 1: 2800 -8141.282 0.011 0.011
Chain 1: 2900 -8247.148 0.011 0.011
Chain 1: 3000 -8084.895 0.013 0.012
Chain 1: 3100 -8225.829 0.013 0.012
Chain 1: 3200 -8095.029 0.011 0.012
Chain 1: 3300 -8320.894 0.013 0.013
Chain 1: 3400 -8354.985 0.012 0.013
Chain 1: 3500 -8194.971 0.014 0.016
Chain 1: 3600 -8051.851 0.015 0.017
Chain 1: 3700 -8198.353 0.016 0.018
Chain 1: 3800 -8052.634 0.017 0.018
Chain 1: 3900 -7986.902 0.017 0.018
Chain 1: 4000 -8096.980 0.016 0.018
Chain 1: 4100 -8062.015 0.015 0.018
Chain 1: 4200 -8047.947 0.013 0.018
Chain 1: 4300 -8081.344 0.011 0.014
Chain 1: 4400 -8038.165 0.011 0.014
Chain 1: 4500 -8136.159 0.010 0.012
Chain 1: 4600 -8027.902 0.010 0.012 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003454 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.54 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8446687.498 1.000 1.000
Chain 1: 200 -1595137.162 2.648 4.295
Chain 1: 300 -893397.093 2.027 1.000
Chain 1: 400 -458485.401 1.757 1.000
Chain 1: 500 -357892.930 1.462 0.949
Chain 1: 600 -232413.406 1.308 0.949
Chain 1: 700 -118957.079 1.258 0.949
Chain 1: 800 -86263.115 1.148 0.949
Chain 1: 900 -66696.998 1.053 0.785
Chain 1: 1000 -51593.899 0.977 0.785
Chain 1: 1100 -39153.798 0.909 0.540 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38346.577 0.481 0.379
Chain 1: 1300 -26378.619 0.448 0.379
Chain 1: 1400 -26107.478 0.354 0.318
Chain 1: 1500 -22714.097 0.341 0.318
Chain 1: 1600 -21937.151 0.291 0.293
Chain 1: 1700 -20819.470 0.201 0.293
Chain 1: 1800 -20766.054 0.163 0.149
Chain 1: 1900 -21092.750 0.135 0.054
Chain 1: 2000 -19607.336 0.114 0.054
Chain 1: 2100 -19845.558 0.083 0.035
Chain 1: 2200 -20071.678 0.082 0.035
Chain 1: 2300 -19689.066 0.039 0.019
Chain 1: 2400 -19461.054 0.039 0.019
Chain 1: 2500 -19262.661 0.025 0.015
Chain 1: 2600 -18892.607 0.023 0.015
Chain 1: 2700 -18849.572 0.018 0.012
Chain 1: 2800 -18565.954 0.019 0.015
Chain 1: 2900 -18847.380 0.019 0.015
Chain 1: 3000 -18833.637 0.012 0.012
Chain 1: 3100 -18918.678 0.011 0.012
Chain 1: 3200 -18609.049 0.012 0.015
Chain 1: 3300 -18814.041 0.011 0.012
Chain 1: 3400 -18288.225 0.012 0.015
Chain 1: 3500 -18901.028 0.015 0.015
Chain 1: 3600 -18206.528 0.016 0.015
Chain 1: 3700 -18594.082 0.018 0.017
Chain 1: 3800 -17551.839 0.023 0.021
Chain 1: 3900 -17547.883 0.021 0.021
Chain 1: 4000 -17665.258 0.022 0.021
Chain 1: 4100 -17578.863 0.022 0.021
Chain 1: 4200 -17394.722 0.021 0.021
Chain 1: 4300 -17533.431 0.021 0.021
Chain 1: 4400 -17489.893 0.018 0.011
Chain 1: 4500 -17392.340 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001294 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.94 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48816.265 1.000 1.000
Chain 1: 200 -12451.151 1.960 2.921
Chain 1: 300 -15531.262 1.373 1.000
Chain 1: 400 -12756.651 1.084 1.000
Chain 1: 500 -16199.588 0.910 0.218
Chain 1: 600 -11799.595 0.820 0.373
Chain 1: 700 -14037.206 0.726 0.218
Chain 1: 800 -23643.783 0.686 0.373
Chain 1: 900 -19934.581 0.630 0.218
Chain 1: 1000 -12080.750 0.632 0.373
Chain 1: 1100 -13972.837 0.546 0.218 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -12354.984 0.267 0.213
Chain 1: 1300 -12555.335 0.249 0.213
Chain 1: 1400 -11376.823 0.237 0.186
Chain 1: 1500 -23314.941 0.267 0.186
Chain 1: 1600 -25163.195 0.237 0.159
Chain 1: 1700 -9848.878 0.377 0.186
Chain 1: 1800 -12327.805 0.356 0.186
Chain 1: 1900 -17246.427 0.366 0.201
Chain 1: 2000 -10753.185 0.362 0.201
Chain 1: 2100 -9927.481 0.356 0.201
Chain 1: 2200 -16780.196 0.384 0.285
Chain 1: 2300 -10597.358 0.441 0.408
Chain 1: 2400 -9306.241 0.444 0.408
Chain 1: 2500 -16837.985 0.438 0.408
Chain 1: 2600 -9318.077 0.511 0.447 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2700 -9780.762 0.361 0.408
Chain 1: 2800 -11442.995 0.355 0.408
Chain 1: 2900 -9550.834 0.346 0.408
Chain 1: 3000 -9671.460 0.287 0.198
Chain 1: 3100 -13777.380 0.309 0.298
Chain 1: 3200 -11022.525 0.293 0.250
Chain 1: 3300 -10184.288 0.243 0.198
Chain 1: 3400 -9227.887 0.239 0.198
Chain 1: 3500 -9554.756 0.198 0.145
Chain 1: 3600 -10566.839 0.127 0.104
Chain 1: 3700 -17040.915 0.160 0.145
Chain 1: 3800 -11280.661 0.197 0.198
Chain 1: 3900 -10782.522 0.181 0.104
Chain 1: 4000 -10091.488 0.187 0.104
Chain 1: 4100 -10071.690 0.157 0.096
Chain 1: 4200 -10622.033 0.137 0.082
Chain 1: 4300 -9641.281 0.139 0.096
Chain 1: 4400 -8776.184 0.139 0.096
Chain 1: 4500 -8571.640 0.138 0.096
Chain 1: 4600 -8945.512 0.132 0.068
Chain 1: 4700 -8487.016 0.100 0.054
Chain 1: 4800 -8906.736 0.054 0.052
Chain 1: 4900 -9721.799 0.057 0.054
Chain 1: 5000 -9753.909 0.051 0.052
Chain 1: 5100 -8826.179 0.061 0.054
Chain 1: 5200 -9053.663 0.058 0.054
Chain 1: 5300 -9072.834 0.048 0.047
Chain 1: 5400 -8691.802 0.043 0.044
Chain 1: 5500 -12262.659 0.070 0.047
Chain 1: 5600 -10448.029 0.083 0.054
Chain 1: 5700 -13439.915 0.100 0.084
Chain 1: 5800 -8718.742 0.149 0.105
Chain 1: 5900 -8813.585 0.142 0.105
Chain 1: 6000 -9171.501 0.145 0.105
Chain 1: 6100 -8829.740 0.139 0.044
Chain 1: 6200 -9426.044 0.143 0.063
Chain 1: 6300 -9814.339 0.146 0.063
Chain 1: 6400 -9599.394 0.144 0.063
Chain 1: 6500 -8564.460 0.127 0.063
Chain 1: 6600 -9275.588 0.118 0.063
Chain 1: 6700 -8480.452 0.105 0.063
Chain 1: 6800 -12325.034 0.082 0.063
Chain 1: 6900 -10457.137 0.098 0.077
Chain 1: 7000 -8748.523 0.114 0.094
Chain 1: 7100 -9312.030 0.116 0.094
Chain 1: 7200 -8732.504 0.117 0.094
Chain 1: 7300 -10452.470 0.129 0.121
Chain 1: 7400 -8653.400 0.148 0.165
Chain 1: 7500 -8881.510 0.138 0.165
Chain 1: 7600 -8674.342 0.133 0.165
Chain 1: 7700 -10401.222 0.140 0.166
Chain 1: 7800 -8702.188 0.128 0.166
Chain 1: 7900 -8590.708 0.112 0.165
Chain 1: 8000 -8901.597 0.096 0.066
Chain 1: 8100 -8290.691 0.097 0.074
Chain 1: 8200 -10856.519 0.114 0.165
Chain 1: 8300 -8318.820 0.128 0.166
Chain 1: 8400 -11999.072 0.138 0.166
Chain 1: 8500 -9175.208 0.166 0.195
Chain 1: 8600 -12590.779 0.191 0.236
Chain 1: 8700 -10801.213 0.191 0.236
Chain 1: 8800 -8504.580 0.198 0.270
Chain 1: 8900 -9608.012 0.209 0.270
Chain 1: 9000 -8516.031 0.218 0.270
Chain 1: 9100 -9453.155 0.221 0.270
Chain 1: 9200 -9206.698 0.200 0.270
Chain 1: 9300 -10746.693 0.183 0.166
Chain 1: 9400 -10267.244 0.157 0.143
Chain 1: 9500 -8492.458 0.147 0.143
Chain 1: 9600 -8327.928 0.122 0.128
Chain 1: 9700 -9243.014 0.116 0.115
Chain 1: 9800 -8697.737 0.095 0.099
Chain 1: 9900 -9190.017 0.089 0.099
Chain 1: 10000 -9103.655 0.077 0.063
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001426 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.26 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57240.094 1.000 1.000
Chain 1: 200 -17489.689 1.636 2.273
Chain 1: 300 -8783.631 1.421 1.000
Chain 1: 400 -8447.645 1.076 1.000
Chain 1: 500 -8469.382 0.861 0.991
Chain 1: 600 -8555.395 0.719 0.991
Chain 1: 700 -8692.961 0.619 0.040
Chain 1: 800 -8103.730 0.551 0.073
Chain 1: 900 -7952.789 0.492 0.040
Chain 1: 1000 -7871.018 0.443 0.040
Chain 1: 1100 -7796.606 0.344 0.019
Chain 1: 1200 -7704.750 0.118 0.016
Chain 1: 1300 -7628.024 0.020 0.012
Chain 1: 1400 -7696.219 0.017 0.010
Chain 1: 1500 -7675.372 0.017 0.010
Chain 1: 1600 -7729.835 0.017 0.010
Chain 1: 1700 -7610.463 0.017 0.010
Chain 1: 1800 -7644.665 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003167 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.67 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86057.766 1.000 1.000
Chain 1: 200 -13549.242 3.176 5.351
Chain 1: 300 -9963.598 2.237 1.000
Chain 1: 400 -10863.043 1.699 1.000
Chain 1: 500 -8888.042 1.403 0.360
Chain 1: 600 -8448.805 1.178 0.360
Chain 1: 700 -8676.520 1.014 0.222
Chain 1: 800 -8958.693 0.891 0.222
Chain 1: 900 -8759.664 0.794 0.083
Chain 1: 1000 -8552.153 0.717 0.083
Chain 1: 1100 -8808.213 0.620 0.052 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8444.556 0.089 0.043
Chain 1: 1300 -8648.129 0.056 0.031
Chain 1: 1400 -8655.930 0.048 0.029
Chain 1: 1500 -8551.000 0.027 0.026
Chain 1: 1600 -8653.247 0.023 0.024
Chain 1: 1700 -8741.824 0.021 0.024
Chain 1: 1800 -8337.051 0.023 0.024
Chain 1: 1900 -8435.542 0.022 0.024
Chain 1: 2000 -8407.158 0.019 0.012
Chain 1: 2100 -8526.986 0.018 0.012
Chain 1: 2200 -8335.572 0.016 0.012
Chain 1: 2300 -8470.998 0.015 0.012
Chain 1: 2400 -8345.914 0.017 0.014
Chain 1: 2500 -8410.743 0.016 0.014
Chain 1: 2600 -8434.436 0.015 0.014
Chain 1: 2700 -8352.725 0.015 0.014
Chain 1: 2800 -8325.294 0.011 0.012
Chain 1: 2900 -8380.697 0.010 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00348 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.8 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8388983.106 1.000 1.000
Chain 1: 200 -1580933.491 2.653 4.306
Chain 1: 300 -890589.087 2.027 1.000
Chain 1: 400 -457534.681 1.757 1.000
Chain 1: 500 -358340.048 1.461 0.946
Chain 1: 600 -233244.121 1.307 0.946
Chain 1: 700 -119392.818 1.256 0.946
Chain 1: 800 -86592.670 1.147 0.946
Chain 1: 900 -66909.990 1.052 0.775
Chain 1: 1000 -51684.648 0.976 0.775
Chain 1: 1100 -39143.203 0.908 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38315.545 0.480 0.379
Chain 1: 1300 -26251.930 0.448 0.379
Chain 1: 1400 -25968.058 0.355 0.320
Chain 1: 1500 -22550.717 0.342 0.320
Chain 1: 1600 -21766.012 0.292 0.295
Chain 1: 1700 -20637.222 0.202 0.294
Chain 1: 1800 -20580.787 0.165 0.152
Chain 1: 1900 -20906.758 0.137 0.055
Chain 1: 2000 -19417.239 0.115 0.055
Chain 1: 2100 -19655.475 0.084 0.036
Chain 1: 2200 -19882.121 0.083 0.036
Chain 1: 2300 -19499.224 0.039 0.020
Chain 1: 2400 -19271.385 0.039 0.020
Chain 1: 2500 -19073.594 0.025 0.016
Chain 1: 2600 -18703.806 0.023 0.016
Chain 1: 2700 -18660.807 0.018 0.012
Chain 1: 2800 -18377.853 0.020 0.015
Chain 1: 2900 -18659.038 0.019 0.015
Chain 1: 3000 -18645.139 0.012 0.012
Chain 1: 3100 -18730.124 0.011 0.012
Chain 1: 3200 -18420.918 0.012 0.015
Chain 1: 3300 -18625.581 0.011 0.012
Chain 1: 3400 -18100.773 0.012 0.015
Chain 1: 3500 -18712.325 0.015 0.015
Chain 1: 3600 -18019.443 0.017 0.015
Chain 1: 3700 -18405.940 0.018 0.017
Chain 1: 3800 -17366.407 0.023 0.021
Chain 1: 3900 -17362.618 0.021 0.021
Chain 1: 4000 -17479.861 0.022 0.021
Chain 1: 4100 -17393.695 0.022 0.021
Chain 1: 4200 -17210.117 0.021 0.021
Chain 1: 4300 -17348.361 0.021 0.021
Chain 1: 4400 -17305.318 0.018 0.011
Chain 1: 4500 -17207.914 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001281 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.81 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12576.392 1.000 1.000
Chain 1: 200 -9529.530 0.660 1.000
Chain 1: 300 -8204.424 0.494 0.320
Chain 1: 400 -8351.999 0.375 0.320
Chain 1: 500 -8308.928 0.301 0.162
Chain 1: 600 -8174.525 0.253 0.162
Chain 1: 700 -8122.361 0.218 0.018
Chain 1: 800 -8142.739 0.191 0.018
Chain 1: 900 -8084.679 0.171 0.016
Chain 1: 1000 -8155.276 0.155 0.016
Chain 1: 1100 -8334.157 0.057 0.016
Chain 1: 1200 -8129.942 0.027 0.016
Chain 1: 1300 -8064.864 0.012 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001493 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.93 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -62497.285 1.000 1.000
Chain 1: 200 -17993.145 1.737 2.473
Chain 1: 300 -8942.501 1.495 1.012
Chain 1: 400 -8486.971 1.135 1.012
Chain 1: 500 -8702.907 0.913 1.000
Chain 1: 600 -9203.822 0.770 1.000
Chain 1: 700 -7797.943 0.686 0.180
Chain 1: 800 -8160.899 0.605 0.180
Chain 1: 900 -7950.010 0.541 0.054
Chain 1: 1000 -7775.460 0.489 0.054
Chain 1: 1100 -7813.703 0.390 0.054
Chain 1: 1200 -7706.001 0.144 0.044
Chain 1: 1300 -7763.000 0.043 0.027
Chain 1: 1400 -7880.514 0.039 0.025
Chain 1: 1500 -7566.628 0.041 0.027
Chain 1: 1600 -7752.919 0.038 0.024
Chain 1: 1700 -7545.652 0.023 0.024
Chain 1: 1800 -7634.849 0.019 0.022
Chain 1: 1900 -7650.915 0.017 0.015
Chain 1: 2000 -7643.904 0.015 0.014
Chain 1: 2100 -7616.732 0.015 0.014
Chain 1: 2200 -7724.575 0.015 0.014
Chain 1: 2300 -7568.587 0.016 0.015
Chain 1: 2400 -7678.754 0.016 0.014
Chain 1: 2500 -7495.800 0.014 0.014
Chain 1: 2600 -7540.772 0.013 0.014
Chain 1: 2700 -7566.433 0.010 0.012
Chain 1: 2800 -7606.417 0.009 0.006 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003063 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.63 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86032.085 1.000 1.000
Chain 1: 200 -13687.114 3.143 5.286
Chain 1: 300 -10070.181 2.215 1.000
Chain 1: 400 -10814.230 1.678 1.000
Chain 1: 500 -9036.318 1.382 0.359
Chain 1: 600 -8748.877 1.157 0.359
Chain 1: 700 -8697.474 0.993 0.197
Chain 1: 800 -9350.017 0.877 0.197
Chain 1: 900 -8805.601 0.787 0.070
Chain 1: 1000 -8614.565 0.710 0.070
Chain 1: 1100 -8940.315 0.614 0.069 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8560.976 0.090 0.062
Chain 1: 1300 -8776.907 0.056 0.044
Chain 1: 1400 -8784.365 0.050 0.036
Chain 1: 1500 -8635.818 0.032 0.033
Chain 1: 1600 -8748.380 0.030 0.025
Chain 1: 1700 -8834.763 0.030 0.025
Chain 1: 1800 -8425.818 0.028 0.025
Chain 1: 1900 -8521.498 0.023 0.022
Chain 1: 2000 -8494.235 0.021 0.017
Chain 1: 2100 -8615.600 0.019 0.014
Chain 1: 2200 -8456.879 0.016 0.014
Chain 1: 2300 -8519.908 0.014 0.013
Chain 1: 2400 -8586.471 0.015 0.013
Chain 1: 2500 -8531.938 0.014 0.011
Chain 1: 2600 -8530.257 0.013 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003362 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.62 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8364101.875 1.000 1.000
Chain 1: 200 -1577777.771 2.651 4.301
Chain 1: 300 -890142.093 2.025 1.000
Chain 1: 400 -457366.784 1.755 1.000
Chain 1: 500 -358570.628 1.459 0.946
Chain 1: 600 -233577.297 1.305 0.946
Chain 1: 700 -119676.199 1.255 0.946
Chain 1: 800 -86831.905 1.145 0.946
Chain 1: 900 -67128.571 1.050 0.773
Chain 1: 1000 -51884.960 0.975 0.773
Chain 1: 1100 -39318.727 0.907 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38490.839 0.479 0.378
Chain 1: 1300 -26402.727 0.447 0.378
Chain 1: 1400 -26117.169 0.354 0.320
Chain 1: 1500 -22692.696 0.341 0.320
Chain 1: 1600 -21905.635 0.291 0.294
Chain 1: 1700 -20774.084 0.202 0.294
Chain 1: 1800 -20717.086 0.164 0.151
Chain 1: 1900 -21043.133 0.136 0.054
Chain 1: 2000 -19551.665 0.115 0.054
Chain 1: 2100 -19790.157 0.084 0.036
Chain 1: 2200 -20017.016 0.083 0.036
Chain 1: 2300 -19633.889 0.039 0.020
Chain 1: 2400 -19405.981 0.039 0.020
Chain 1: 2500 -19208.164 0.025 0.015
Chain 1: 2600 -18838.341 0.023 0.015
Chain 1: 2700 -18795.271 0.018 0.012
Chain 1: 2800 -18512.274 0.019 0.015
Chain 1: 2900 -18793.523 0.019 0.015
Chain 1: 3000 -18779.658 0.012 0.012
Chain 1: 3100 -18864.666 0.011 0.012
Chain 1: 3200 -18555.371 0.012 0.015
Chain 1: 3300 -18760.062 0.011 0.012
Chain 1: 3400 -18235.141 0.012 0.015
Chain 1: 3500 -18846.899 0.015 0.015
Chain 1: 3600 -18153.740 0.016 0.015
Chain 1: 3700 -18540.490 0.018 0.017
Chain 1: 3800 -17500.514 0.023 0.021
Chain 1: 3900 -17496.705 0.021 0.021
Chain 1: 4000 -17613.955 0.022 0.021
Chain 1: 4100 -17527.778 0.022 0.021
Chain 1: 4200 -17344.096 0.021 0.021
Chain 1: 4300 -17482.424 0.021 0.021
Chain 1: 4400 -17439.313 0.018 0.011
Chain 1: 4500 -17341.895 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001306 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.06 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12198.221 1.000 1.000
Chain 1: 200 -9103.807 0.670 1.000
Chain 1: 300 -7999.445 0.493 0.340
Chain 1: 400 -8111.977 0.373 0.340
Chain 1: 500 -8039.396 0.300 0.138
Chain 1: 600 -7907.164 0.253 0.138
Chain 1: 700 -7824.757 0.218 0.017
Chain 1: 800 -7835.541 0.191 0.017
Chain 1: 900 -7739.298 0.171 0.014
Chain 1: 1000 -7885.616 0.156 0.017
Chain 1: 1100 -7944.269 0.057 0.014
Chain 1: 1200 -7842.105 0.024 0.013
Chain 1: 1300 -7793.812 0.011 0.012
Chain 1: 1400 -7821.269 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001393 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.93 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61480.139 1.000 1.000
Chain 1: 200 -17744.099 1.732 2.465
Chain 1: 300 -8751.701 1.497 1.028
Chain 1: 400 -9236.393 1.136 1.028
Chain 1: 500 -8338.696 0.930 1.000
Chain 1: 600 -8520.005 0.779 1.000
Chain 1: 700 -7773.648 0.681 0.108
Chain 1: 800 -8134.593 0.602 0.108
Chain 1: 900 -7983.250 0.537 0.096
Chain 1: 1000 -7656.783 0.488 0.096
Chain 1: 1100 -7624.794 0.388 0.052
Chain 1: 1200 -7565.356 0.142 0.044
Chain 1: 1300 -7564.384 0.040 0.043
Chain 1: 1400 -7812.694 0.037 0.032
Chain 1: 1500 -7544.434 0.030 0.032
Chain 1: 1600 -7561.649 0.028 0.032
Chain 1: 1700 -7452.105 0.020 0.019
Chain 1: 1800 -7528.524 0.017 0.015
Chain 1: 1900 -7493.688 0.015 0.010
Chain 1: 2000 -7541.451 0.012 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003048 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.48 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86046.092 1.000 1.000
Chain 1: 200 -13358.881 3.221 5.441
Chain 1: 300 -9766.278 2.270 1.000
Chain 1: 400 -10745.902 1.725 1.000
Chain 1: 500 -8687.595 1.427 0.368
Chain 1: 600 -8255.130 1.198 0.368
Chain 1: 700 -8418.534 1.030 0.237
Chain 1: 800 -8723.866 0.905 0.237
Chain 1: 900 -8662.060 0.806 0.091
Chain 1: 1000 -8428.124 0.728 0.091
Chain 1: 1100 -8656.746 0.631 0.052 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8300.219 0.091 0.043
Chain 1: 1300 -8550.289 0.057 0.035
Chain 1: 1400 -8495.221 0.048 0.029
Chain 1: 1500 -8350.804 0.026 0.028
Chain 1: 1600 -8460.797 0.022 0.026
Chain 1: 1700 -8548.188 0.022 0.026
Chain 1: 1800 -8146.155 0.023 0.026
Chain 1: 1900 -8243.672 0.023 0.026
Chain 1: 2000 -8215.364 0.021 0.017
Chain 1: 2100 -8335.152 0.020 0.014
Chain 1: 2200 -8141.136 0.018 0.014
Chain 1: 2300 -8278.857 0.017 0.014
Chain 1: 2400 -8154.084 0.018 0.015
Chain 1: 2500 -8218.968 0.017 0.014
Chain 1: 2600 -8242.414 0.016 0.014
Chain 1: 2700 -8160.862 0.016 0.014
Chain 1: 2800 -8133.510 0.011 0.012
Chain 1: 2900 -8188.930 0.010 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003308 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.08 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8404371.337 1.000 1.000
Chain 1: 200 -1584506.346 2.652 4.304
Chain 1: 300 -891135.561 2.027 1.000
Chain 1: 400 -457835.227 1.757 1.000
Chain 1: 500 -358356.987 1.461 0.946
Chain 1: 600 -233128.201 1.307 0.946
Chain 1: 700 -119197.982 1.257 0.946
Chain 1: 800 -86398.919 1.147 0.946
Chain 1: 900 -66705.892 1.053 0.778
Chain 1: 1000 -51483.210 0.977 0.778
Chain 1: 1100 -38943.981 0.909 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38114.310 0.481 0.380
Chain 1: 1300 -26055.609 0.449 0.380
Chain 1: 1400 -25771.216 0.356 0.322
Chain 1: 1500 -22356.347 0.343 0.322
Chain 1: 1600 -21571.826 0.293 0.296
Chain 1: 1700 -20443.992 0.203 0.295
Chain 1: 1800 -20387.616 0.166 0.153
Chain 1: 1900 -20713.551 0.138 0.055
Chain 1: 2000 -19224.631 0.116 0.055
Chain 1: 2100 -19462.694 0.085 0.036
Chain 1: 2200 -19689.351 0.084 0.036
Chain 1: 2300 -19306.497 0.039 0.020
Chain 1: 2400 -19078.700 0.040 0.020
Chain 1: 2500 -18880.884 0.025 0.016
Chain 1: 2600 -18511.083 0.024 0.016
Chain 1: 2700 -18468.071 0.018 0.012
Chain 1: 2800 -18185.157 0.020 0.016
Chain 1: 2900 -18466.318 0.020 0.015
Chain 1: 3000 -18452.395 0.012 0.012
Chain 1: 3100 -18537.383 0.011 0.012
Chain 1: 3200 -18228.167 0.012 0.015
Chain 1: 3300 -18432.819 0.011 0.012
Chain 1: 3400 -17908.016 0.013 0.015
Chain 1: 3500 -18519.536 0.015 0.016
Chain 1: 3600 -17826.715 0.017 0.016
Chain 1: 3700 -18213.179 0.019 0.017
Chain 1: 3800 -17173.679 0.023 0.021
Chain 1: 3900 -17169.894 0.022 0.021
Chain 1: 4000 -17287.148 0.022 0.021
Chain 1: 4100 -17200.998 0.022 0.021
Chain 1: 4200 -17017.407 0.022 0.021
Chain 1: 4300 -17155.639 0.021 0.021
Chain 1: 4400 -17112.595 0.019 0.011
Chain 1: 4500 -17015.218 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001402 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.02 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12186.113 1.000 1.000
Chain 1: 200 -9048.808 0.673 1.000
Chain 1: 300 -8060.294 0.490 0.347
Chain 1: 400 -8088.847 0.368 0.347
Chain 1: 500 -7940.541 0.298 0.123
Chain 1: 600 -7866.506 0.250 0.123
Chain 1: 700 -7784.465 0.216 0.019
Chain 1: 800 -7817.171 0.189 0.019
Chain 1: 900 -7982.887 0.171 0.019
Chain 1: 1000 -7824.468 0.156 0.020
Chain 1: 1100 -7865.454 0.056 0.019
Chain 1: 1200 -7819.068 0.022 0.011
Chain 1: 1300 -7758.166 0.011 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001416 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.16 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -45988.717 1.000 1.000
Chain 1: 200 -15243.040 1.509 2.017
Chain 1: 300 -8567.788 1.265 1.000
Chain 1: 400 -8519.585 0.950 1.000
Chain 1: 500 -8353.289 0.764 0.779
Chain 1: 600 -8176.948 0.641 0.779
Chain 1: 700 -7795.179 0.556 0.049
Chain 1: 800 -8028.163 0.490 0.049
Chain 1: 900 -7747.833 0.440 0.036
Chain 1: 1000 -7793.879 0.396 0.036
Chain 1: 1100 -7658.304 0.298 0.029
Chain 1: 1200 -7623.695 0.097 0.022
Chain 1: 1300 -7608.454 0.019 0.020
Chain 1: 1400 -7719.527 0.020 0.020
Chain 1: 1500 -7604.472 0.020 0.018
Chain 1: 1600 -7504.287 0.019 0.015
Chain 1: 1700 -7516.572 0.014 0.014
Chain 1: 1800 -7561.404 0.012 0.013
Chain 1: 1900 -7599.247 0.009 0.006 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002587 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.87 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86041.841 1.000 1.000
Chain 1: 200 -13256.019 3.245 5.491
Chain 1: 300 -9687.171 2.286 1.000
Chain 1: 400 -10565.338 1.736 1.000
Chain 1: 500 -8623.327 1.434 0.368
Chain 1: 600 -8219.452 1.203 0.368
Chain 1: 700 -8357.351 1.033 0.225
Chain 1: 800 -8939.114 0.912 0.225
Chain 1: 900 -8504.630 0.817 0.083
Chain 1: 1000 -8356.438 0.737 0.083
Chain 1: 1100 -8581.674 0.639 0.065 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8263.320 0.094 0.051
Chain 1: 1300 -8413.801 0.059 0.049
Chain 1: 1400 -8420.468 0.051 0.039
Chain 1: 1500 -8287.952 0.030 0.026
Chain 1: 1600 -8396.069 0.026 0.018
Chain 1: 1700 -8480.862 0.026 0.018
Chain 1: 1800 -8087.625 0.024 0.018
Chain 1: 1900 -8188.477 0.020 0.018
Chain 1: 2000 -8159.103 0.019 0.016
Chain 1: 2100 -8281.903 0.018 0.015
Chain 1: 2200 -8063.738 0.016 0.015
Chain 1: 2300 -8217.267 0.016 0.015
Chain 1: 2400 -8231.288 0.017 0.015
Chain 1: 2500 -8200.471 0.015 0.013
Chain 1: 2600 -8203.029 0.014 0.012
Chain 1: 2700 -8109.308 0.014 0.012
Chain 1: 2800 -8080.648 0.010 0.012 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003189 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.89 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8391273.839 1.000 1.000
Chain 1: 200 -1584712.971 2.648 4.295
Chain 1: 300 -890591.336 2.025 1.000
Chain 1: 400 -457177.332 1.756 1.000
Chain 1: 500 -357531.970 1.460 0.948
Chain 1: 600 -232498.888 1.307 0.948
Chain 1: 700 -118847.633 1.256 0.948
Chain 1: 800 -86081.412 1.147 0.948
Chain 1: 900 -66447.355 1.052 0.779
Chain 1: 1000 -51262.602 0.977 0.779
Chain 1: 1100 -38758.009 0.909 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37935.195 0.482 0.381
Chain 1: 1300 -25917.362 0.450 0.381
Chain 1: 1400 -25637.292 0.356 0.323
Chain 1: 1500 -22231.109 0.344 0.323
Chain 1: 1600 -21448.820 0.294 0.296
Chain 1: 1700 -20326.140 0.204 0.295
Chain 1: 1800 -20270.982 0.166 0.153
Chain 1: 1900 -20596.743 0.138 0.055
Chain 1: 2000 -19110.373 0.116 0.055
Chain 1: 2100 -19348.674 0.085 0.036
Chain 1: 2200 -19574.499 0.084 0.036
Chain 1: 2300 -19192.351 0.040 0.020
Chain 1: 2400 -18964.593 0.040 0.020
Chain 1: 2500 -18766.469 0.025 0.016
Chain 1: 2600 -18397.221 0.024 0.016
Chain 1: 2700 -18354.393 0.019 0.012
Chain 1: 2800 -18071.329 0.020 0.016
Chain 1: 2900 -18352.389 0.020 0.015
Chain 1: 3000 -18338.653 0.012 0.012
Chain 1: 3100 -18423.541 0.011 0.012
Chain 1: 3200 -18114.556 0.012 0.015
Chain 1: 3300 -18319.038 0.011 0.012
Chain 1: 3400 -17794.447 0.013 0.015
Chain 1: 3500 -18405.549 0.015 0.016
Chain 1: 3600 -17713.275 0.017 0.016
Chain 1: 3700 -18099.243 0.019 0.017
Chain 1: 3800 -17060.525 0.023 0.021
Chain 1: 3900 -17056.698 0.022 0.021
Chain 1: 4000 -17174.020 0.022 0.021
Chain 1: 4100 -17087.810 0.022 0.021
Chain 1: 4200 -16904.443 0.022 0.021
Chain 1: 4300 -17042.590 0.022 0.021
Chain 1: 4400 -16999.691 0.019 0.011
Chain 1: 4500 -16902.267 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001346 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.46 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12337.952 1.000 1.000
Chain 1: 200 -9206.745 0.670 1.000
Chain 1: 300 -8051.006 0.495 0.340
Chain 1: 400 -8201.963 0.376 0.340
Chain 1: 500 -8084.805 0.303 0.144
Chain 1: 600 -8009.428 0.254 0.144
Chain 1: 700 -7923.271 0.220 0.018
Chain 1: 800 -7962.844 0.193 0.018
Chain 1: 900 -8084.625 0.173 0.015
Chain 1: 1000 -7977.376 0.157 0.015
Chain 1: 1100 -8025.820 0.058 0.014
Chain 1: 1200 -7961.674 0.024 0.013
Chain 1: 1300 -7893.986 0.011 0.011
Chain 1: 1400 -7907.744 0.009 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001412 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.12 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -46188.706 1.000 1.000
Chain 1: 200 -15443.582 1.495 1.991
Chain 1: 300 -8678.136 1.257 1.000
Chain 1: 400 -8657.491 0.943 1.000
Chain 1: 500 -8170.648 0.766 0.780
Chain 1: 600 -8833.670 0.651 0.780
Chain 1: 700 -8082.209 0.571 0.093
Chain 1: 800 -8234.639 0.502 0.093
Chain 1: 900 -7978.422 0.450 0.075
Chain 1: 1000 -7957.675 0.405 0.075
Chain 1: 1100 -7664.302 0.309 0.060
Chain 1: 1200 -7679.498 0.110 0.038
Chain 1: 1300 -7642.104 0.033 0.032
Chain 1: 1400 -7927.616 0.036 0.036
Chain 1: 1500 -7636.195 0.034 0.036
Chain 1: 1600 -7617.072 0.027 0.032
Chain 1: 1700 -7577.532 0.018 0.019
Chain 1: 1800 -7623.846 0.017 0.006 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003327 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.27 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86578.137 1.000 1.000
Chain 1: 200 -13438.878 3.221 5.442
Chain 1: 300 -9849.843 2.269 1.000
Chain 1: 400 -10821.073 1.724 1.000
Chain 1: 500 -8771.551 1.426 0.364
Chain 1: 600 -8525.567 1.193 0.364
Chain 1: 700 -8587.541 1.024 0.234
Chain 1: 800 -8836.418 0.899 0.234
Chain 1: 900 -8660.903 0.802 0.090
Chain 1: 1000 -8438.246 0.724 0.090
Chain 1: 1100 -8701.149 0.627 0.030 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8245.296 0.088 0.030
Chain 1: 1300 -8568.486 0.056 0.030
Chain 1: 1400 -8564.173 0.047 0.029
Chain 1: 1500 -8443.515 0.025 0.028
Chain 1: 1600 -8549.611 0.023 0.026
Chain 1: 1700 -8635.533 0.024 0.026
Chain 1: 1800 -8235.238 0.026 0.026
Chain 1: 1900 -8334.862 0.025 0.026
Chain 1: 2000 -8306.058 0.022 0.014
Chain 1: 2100 -8425.985 0.021 0.014
Chain 1: 2200 -8214.837 0.018 0.014
Chain 1: 2300 -8366.080 0.016 0.014
Chain 1: 2400 -8247.544 0.017 0.014
Chain 1: 2500 -8310.414 0.017 0.014
Chain 1: 2600 -8331.722 0.016 0.014
Chain 1: 2700 -8250.813 0.016 0.014
Chain 1: 2800 -8224.895 0.011 0.012
Chain 1: 2900 -8280.268 0.011 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003072 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.72 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8414310.013 1.000 1.000
Chain 1: 200 -1586898.513 2.651 4.302
Chain 1: 300 -891645.896 2.027 1.000
Chain 1: 400 -458068.176 1.757 1.000
Chain 1: 500 -358297.326 1.461 0.947
Chain 1: 600 -233011.873 1.307 0.947
Chain 1: 700 -119162.163 1.257 0.947
Chain 1: 800 -86377.292 1.147 0.947
Chain 1: 900 -66705.822 1.053 0.780
Chain 1: 1000 -51497.276 0.977 0.780
Chain 1: 1100 -38976.865 0.909 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38149.388 0.481 0.380
Chain 1: 1300 -26113.181 0.449 0.380
Chain 1: 1400 -25831.517 0.356 0.321
Chain 1: 1500 -22421.691 0.343 0.321
Chain 1: 1600 -21638.559 0.293 0.295
Chain 1: 1700 -20513.476 0.203 0.295
Chain 1: 1800 -20457.718 0.165 0.152
Chain 1: 1900 -20783.568 0.137 0.055
Chain 1: 2000 -19295.951 0.115 0.055
Chain 1: 2100 -19534.209 0.084 0.036
Chain 1: 2200 -19760.436 0.083 0.036
Chain 1: 2300 -19377.911 0.039 0.020
Chain 1: 2400 -19150.096 0.039 0.020
Chain 1: 2500 -18952.139 0.025 0.016
Chain 1: 2600 -18582.626 0.024 0.016
Chain 1: 2700 -18539.665 0.018 0.012
Chain 1: 2800 -18256.686 0.020 0.015
Chain 1: 2900 -18537.771 0.020 0.015
Chain 1: 3000 -18523.953 0.012 0.012
Chain 1: 3100 -18608.917 0.011 0.012
Chain 1: 3200 -18299.796 0.012 0.015
Chain 1: 3300 -18504.355 0.011 0.012
Chain 1: 3400 -17979.660 0.013 0.015
Chain 1: 3500 -18590.994 0.015 0.015
Chain 1: 3600 -17898.341 0.017 0.015
Chain 1: 3700 -18284.646 0.019 0.017
Chain 1: 3800 -17245.438 0.023 0.021
Chain 1: 3900 -17241.601 0.022 0.021
Chain 1: 4000 -17358.896 0.022 0.021
Chain 1: 4100 -17272.737 0.022 0.021
Chain 1: 4200 -17089.185 0.022 0.021
Chain 1: 4300 -17227.423 0.021 0.021
Chain 1: 4400 -17184.431 0.019 0.011
Chain 1: 4500 -17086.990 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001234 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.34 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12323.632 1.000 1.000
Chain 1: 200 -9147.413 0.674 1.000
Chain 1: 300 -8265.272 0.485 0.347
Chain 1: 400 -8280.936 0.364 0.347
Chain 1: 500 -8175.918 0.294 0.107
Chain 1: 600 -8053.899 0.247 0.107
Chain 1: 700 -7981.156 0.213 0.015
Chain 1: 800 -7991.011 0.187 0.015
Chain 1: 900 -7888.298 0.167 0.013
Chain 1: 1000 -8035.189 0.153 0.015
Chain 1: 1100 -8044.652 0.053 0.013
Chain 1: 1200 -8013.001 0.018 0.013
Chain 1: 1300 -7949.563 0.008 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001408 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.08 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61766.091 1.000 1.000
Chain 1: 200 -17787.911 1.736 2.472
Chain 1: 300 -8810.943 1.497 1.019
Chain 1: 400 -9189.460 1.133 1.019
Chain 1: 500 -7780.355 0.943 1.000
Chain 1: 600 -8483.967 0.799 1.000
Chain 1: 700 -8118.159 0.692 0.181
Chain 1: 800 -8175.084 0.606 0.181
Chain 1: 900 -7913.246 0.542 0.083
Chain 1: 1000 -7777.909 0.490 0.083
Chain 1: 1100 -7563.487 0.393 0.045
Chain 1: 1200 -7721.353 0.148 0.041
Chain 1: 1300 -7701.817 0.046 0.033
Chain 1: 1400 -7820.451 0.043 0.028
Chain 1: 1500 -7586.376 0.028 0.028
Chain 1: 1600 -7615.750 0.020 0.020
Chain 1: 1700 -7508.418 0.017 0.017
Chain 1: 1800 -7544.450 0.017 0.017
Chain 1: 1900 -7558.964 0.014 0.015
Chain 1: 2000 -7572.111 0.012 0.014
Chain 1: 2100 -7575.416 0.010 0.005 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002715 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 27.15 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86338.579 1.000 1.000
Chain 1: 200 -13454.211 3.209 5.417
Chain 1: 300 -9886.928 2.259 1.000
Chain 1: 400 -10858.416 1.717 1.000
Chain 1: 500 -8813.909 1.420 0.361
Chain 1: 600 -8658.766 1.186 0.361
Chain 1: 700 -8712.121 1.018 0.232
Chain 1: 800 -8647.772 0.891 0.232
Chain 1: 900 -8678.344 0.793 0.089
Chain 1: 1000 -8511.714 0.715 0.089
Chain 1: 1100 -8762.143 0.618 0.029 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8470.120 0.080 0.029
Chain 1: 1300 -8627.641 0.046 0.020
Chain 1: 1400 -8628.864 0.037 0.018
Chain 1: 1500 -8495.942 0.015 0.018
Chain 1: 1600 -8602.130 0.015 0.016
Chain 1: 1700 -8689.067 0.015 0.016
Chain 1: 1800 -8298.153 0.019 0.018
Chain 1: 1900 -8399.632 0.020 0.018
Chain 1: 2000 -8370.088 0.018 0.016
Chain 1: 2100 -8496.723 0.017 0.015
Chain 1: 2200 -8282.618 0.016 0.015
Chain 1: 2300 -8428.560 0.016 0.015
Chain 1: 2400 -8444.002 0.016 0.015
Chain 1: 2500 -8410.461 0.015 0.012
Chain 1: 2600 -8412.300 0.014 0.012
Chain 1: 2700 -8319.260 0.014 0.012
Chain 1: 2800 -8292.530 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003107 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.07 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8416505.773 1.000 1.000
Chain 1: 200 -1588000.943 2.650 4.300
Chain 1: 300 -891011.528 2.027 1.000
Chain 1: 400 -457655.907 1.757 1.000
Chain 1: 500 -357553.092 1.462 0.947
Chain 1: 600 -232541.271 1.308 0.947
Chain 1: 700 -118926.160 1.257 0.947
Chain 1: 800 -86205.859 1.148 0.947
Chain 1: 900 -66589.267 1.053 0.782
Chain 1: 1000 -51419.381 0.977 0.782
Chain 1: 1100 -38928.152 0.909 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38107.492 0.481 0.380
Chain 1: 1300 -26098.193 0.449 0.380
Chain 1: 1400 -25819.830 0.356 0.321
Chain 1: 1500 -22416.980 0.343 0.321
Chain 1: 1600 -21636.219 0.293 0.295
Chain 1: 1700 -20514.223 0.203 0.295
Chain 1: 1800 -20459.359 0.165 0.152
Chain 1: 1900 -20785.111 0.137 0.055
Chain 1: 2000 -19299.626 0.115 0.055
Chain 1: 2100 -19537.626 0.084 0.036
Chain 1: 2200 -19763.501 0.083 0.036
Chain 1: 2300 -19381.394 0.039 0.020
Chain 1: 2400 -19153.697 0.039 0.020
Chain 1: 2500 -18955.701 0.025 0.016
Chain 1: 2600 -18586.244 0.024 0.016
Chain 1: 2700 -18543.439 0.018 0.012
Chain 1: 2800 -18260.409 0.020 0.015
Chain 1: 2900 -18541.515 0.020 0.015
Chain 1: 3000 -18527.695 0.012 0.012
Chain 1: 3100 -18612.582 0.011 0.012
Chain 1: 3200 -18303.542 0.012 0.015
Chain 1: 3300 -18508.123 0.011 0.012
Chain 1: 3400 -17983.491 0.013 0.015
Chain 1: 3500 -18594.557 0.015 0.015
Chain 1: 3600 -17902.391 0.017 0.015
Chain 1: 3700 -18288.281 0.019 0.017
Chain 1: 3800 -17249.654 0.023 0.021
Chain 1: 3900 -17245.871 0.022 0.021
Chain 1: 4000 -17363.176 0.022 0.021
Chain 1: 4100 -17276.939 0.022 0.021
Chain 1: 4200 -17093.647 0.022 0.021
Chain 1: 4300 -17231.710 0.021 0.021
Chain 1: 4400 -17188.820 0.019 0.011
Chain 1: 4500 -17091.454 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001386 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.86 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48562.613 1.000 1.000
Chain 1: 200 -13238.117 1.834 2.668
Chain 1: 300 -28193.560 1.400 1.000
Chain 1: 400 -14272.521 1.294 1.000
Chain 1: 500 -13123.666 1.052 0.975
Chain 1: 600 -15507.390 0.903 0.975
Chain 1: 700 -19215.853 0.801 0.530
Chain 1: 800 -14381.624 0.743 0.530
Chain 1: 900 -12556.415 0.677 0.336
Chain 1: 1000 -12059.603 0.613 0.336
Chain 1: 1100 -9633.149 0.538 0.252 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -11599.818 0.288 0.193
Chain 1: 1300 -11913.702 0.238 0.170
Chain 1: 1400 -9800.061 0.162 0.170
Chain 1: 1500 -10284.863 0.158 0.170
Chain 1: 1600 -26840.674 0.204 0.193
Chain 1: 1700 -16007.481 0.253 0.216
Chain 1: 1800 -12141.759 0.251 0.216
Chain 1: 1900 -14515.240 0.253 0.216
Chain 1: 2000 -11961.943 0.270 0.216
Chain 1: 2100 -9114.666 0.276 0.216
Chain 1: 2200 -10195.211 0.270 0.216
Chain 1: 2300 -9386.550 0.276 0.216
Chain 1: 2400 -9445.429 0.255 0.213
Chain 1: 2500 -9453.767 0.250 0.213
Chain 1: 2600 -8920.384 0.194 0.164
Chain 1: 2700 -11936.942 0.152 0.164
Chain 1: 2800 -8839.355 0.155 0.164
Chain 1: 2900 -8759.940 0.140 0.106
Chain 1: 3000 -8805.625 0.119 0.086
Chain 1: 3100 -8983.784 0.090 0.060
Chain 1: 3200 -13840.456 0.114 0.060
Chain 1: 3300 -9285.945 0.155 0.060
Chain 1: 3400 -12025.017 0.177 0.228
Chain 1: 3500 -9226.499 0.207 0.253
Chain 1: 3600 -9293.774 0.202 0.253
Chain 1: 3700 -9043.523 0.179 0.228
Chain 1: 3800 -8480.585 0.151 0.066
Chain 1: 3900 -12210.871 0.180 0.228
Chain 1: 4000 -9990.046 0.202 0.228
Chain 1: 4100 -12678.501 0.221 0.228
Chain 1: 4200 -10216.236 0.210 0.228
Chain 1: 4300 -9510.393 0.169 0.222
Chain 1: 4400 -8444.800 0.159 0.212
Chain 1: 4500 -13729.700 0.167 0.212
Chain 1: 4600 -8232.999 0.233 0.222
Chain 1: 4700 -8976.902 0.238 0.222
Chain 1: 4800 -8263.081 0.240 0.222
Chain 1: 4900 -8429.568 0.212 0.212
Chain 1: 5000 -9660.216 0.202 0.127
Chain 1: 5100 -8423.375 0.196 0.127
Chain 1: 5200 -13410.808 0.209 0.127
Chain 1: 5300 -8065.809 0.268 0.147
Chain 1: 5400 -14966.322 0.301 0.372
Chain 1: 5500 -11710.424 0.290 0.278
Chain 1: 5600 -15358.411 0.247 0.238
Chain 1: 5700 -11029.069 0.278 0.278
Chain 1: 5800 -8382.686 0.301 0.316
Chain 1: 5900 -12747.000 0.334 0.342
Chain 1: 6000 -8825.151 0.365 0.372
Chain 1: 6100 -8822.422 0.351 0.372
Chain 1: 6200 -8350.402 0.319 0.342
Chain 1: 6300 -8114.237 0.256 0.316
Chain 1: 6400 -8244.842 0.211 0.278
Chain 1: 6500 -11584.825 0.212 0.288
Chain 1: 6600 -8156.493 0.231 0.316
Chain 1: 6700 -8180.674 0.192 0.288
Chain 1: 6800 -11571.664 0.189 0.288
Chain 1: 6900 -9744.267 0.174 0.188
Chain 1: 7000 -8191.489 0.148 0.188
Chain 1: 7100 -14430.671 0.192 0.190
Chain 1: 7200 -8424.981 0.257 0.288
Chain 1: 7300 -9057.486 0.261 0.288
Chain 1: 7400 -7920.375 0.274 0.288
Chain 1: 7500 -8839.469 0.256 0.190
Chain 1: 7600 -8575.883 0.217 0.188
Chain 1: 7700 -8183.368 0.221 0.188
Chain 1: 7800 -10042.023 0.210 0.185
Chain 1: 7900 -9232.632 0.200 0.144
Chain 1: 8000 -10941.404 0.197 0.144
Chain 1: 8100 -8244.832 0.186 0.144
Chain 1: 8200 -10967.910 0.140 0.144
Chain 1: 8300 -7923.679 0.171 0.156
Chain 1: 8400 -10679.973 0.183 0.185
Chain 1: 8500 -7890.022 0.208 0.248
Chain 1: 8600 -8512.935 0.212 0.248
Chain 1: 8700 -7908.016 0.215 0.248
Chain 1: 8800 -8327.848 0.202 0.248
Chain 1: 8900 -12251.098 0.225 0.258
Chain 1: 9000 -10319.660 0.228 0.258
Chain 1: 9100 -9899.123 0.199 0.248
Chain 1: 9200 -10266.031 0.178 0.187
Chain 1: 9300 -9263.685 0.151 0.108
Chain 1: 9400 -8247.227 0.137 0.108
Chain 1: 9500 -10982.144 0.127 0.108
Chain 1: 9600 -7961.013 0.157 0.123
Chain 1: 9700 -8223.651 0.153 0.123
Chain 1: 9800 -8444.983 0.150 0.123
Chain 1: 9900 -9981.406 0.134 0.123
Chain 1: 10000 -8267.572 0.136 0.123
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001385 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.85 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57627.260 1.000 1.000
Chain 1: 200 -17189.176 1.676 2.353
Chain 1: 300 -8519.582 1.457 1.018
Chain 1: 400 -7863.205 1.113 1.018
Chain 1: 500 -8344.085 0.902 1.000
Chain 1: 600 -8193.824 0.755 1.000
Chain 1: 700 -8316.445 0.649 0.083
Chain 1: 800 -8846.598 0.576 0.083
Chain 1: 900 -7767.570 0.527 0.083
Chain 1: 1000 -7740.801 0.475 0.083
Chain 1: 1100 -7770.177 0.375 0.060
Chain 1: 1200 -7733.959 0.140 0.058
Chain 1: 1300 -7746.314 0.039 0.018
Chain 1: 1400 -7889.812 0.032 0.018
Chain 1: 1500 -7651.186 0.029 0.018
Chain 1: 1600 -7557.530 0.029 0.015
Chain 1: 1700 -7543.246 0.028 0.012
Chain 1: 1800 -7576.395 0.022 0.005 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002967 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.67 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85658.970 1.000 1.000
Chain 1: 200 -13010.821 3.292 5.584
Chain 1: 300 -9487.398 2.318 1.000
Chain 1: 400 -10416.526 1.761 1.000
Chain 1: 500 -8398.016 1.457 0.371
Chain 1: 600 -8046.580 1.221 0.371
Chain 1: 700 -8340.741 1.052 0.240
Chain 1: 800 -8566.446 0.924 0.240
Chain 1: 900 -8313.410 0.824 0.089
Chain 1: 1000 -8103.370 0.745 0.089
Chain 1: 1100 -8363.407 0.648 0.044 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8032.934 0.093 0.041
Chain 1: 1300 -8230.835 0.059 0.035
Chain 1: 1400 -8229.023 0.050 0.031
Chain 1: 1500 -8130.444 0.027 0.030
Chain 1: 1600 -8219.068 0.024 0.026
Chain 1: 1700 -8309.578 0.021 0.026
Chain 1: 1800 -7925.348 0.024 0.026
Chain 1: 1900 -8027.584 0.022 0.024
Chain 1: 2000 -7997.103 0.020 0.013
Chain 1: 2100 -8133.396 0.018 0.013
Chain 1: 2200 -7916.593 0.017 0.013
Chain 1: 2300 -8058.129 0.016 0.013
Chain 1: 2400 -8067.603 0.016 0.013
Chain 1: 2500 -8035.978 0.015 0.013
Chain 1: 2600 -8033.431 0.014 0.013
Chain 1: 2700 -7943.380 0.014 0.013
Chain 1: 2800 -7922.305 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003075 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.75 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8430273.267 1.000 1.000
Chain 1: 200 -1589523.542 2.652 4.304
Chain 1: 300 -890319.270 2.030 1.000
Chain 1: 400 -457010.656 1.759 1.000
Chain 1: 500 -356906.065 1.464 0.948
Chain 1: 600 -231985.581 1.309 0.948
Chain 1: 700 -118438.503 1.259 0.948
Chain 1: 800 -85708.749 1.150 0.948
Chain 1: 900 -66097.944 1.055 0.785
Chain 1: 1000 -50934.958 0.979 0.785
Chain 1: 1100 -38456.536 0.912 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37632.591 0.483 0.382
Chain 1: 1300 -25645.415 0.452 0.382
Chain 1: 1400 -25367.223 0.358 0.324
Chain 1: 1500 -21969.508 0.345 0.324
Chain 1: 1600 -21189.652 0.295 0.298
Chain 1: 1700 -20070.662 0.205 0.297
Chain 1: 1800 -20016.164 0.167 0.155
Chain 1: 1900 -20341.748 0.139 0.056
Chain 1: 2000 -18857.766 0.117 0.056
Chain 1: 2100 -19095.901 0.086 0.037
Chain 1: 2200 -19321.315 0.085 0.037
Chain 1: 2300 -18939.560 0.040 0.020
Chain 1: 2400 -18711.907 0.040 0.020
Chain 1: 2500 -18513.721 0.026 0.016
Chain 1: 2600 -18144.771 0.024 0.016
Chain 1: 2700 -18101.987 0.019 0.012
Chain 1: 2800 -17818.997 0.020 0.016
Chain 1: 2900 -18099.899 0.020 0.016
Chain 1: 3000 -18086.206 0.012 0.012
Chain 1: 3100 -18171.094 0.011 0.012
Chain 1: 3200 -17862.243 0.012 0.016
Chain 1: 3300 -18066.586 0.011 0.012
Chain 1: 3400 -17542.266 0.013 0.016
Chain 1: 3500 -18152.944 0.015 0.016
Chain 1: 3600 -17461.130 0.017 0.016
Chain 1: 3700 -17846.765 0.019 0.017
Chain 1: 3800 -16808.802 0.024 0.022
Chain 1: 3900 -16804.955 0.022 0.022
Chain 1: 4000 -16922.297 0.023 0.022
Chain 1: 4100 -16836.159 0.023 0.022
Chain 1: 4200 -16652.905 0.022 0.022
Chain 1: 4300 -16790.979 0.022 0.022
Chain 1: 4400 -16748.215 0.019 0.011
Chain 1: 4500 -16650.784 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001327 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.27 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48673.352 1.000 1.000
Chain 1: 200 -16441.553 1.480 1.960
Chain 1: 300 -18603.029 1.026 1.000
Chain 1: 400 -19564.135 0.781 1.000
Chain 1: 500 -30591.174 0.697 0.360
Chain 1: 600 -11097.829 0.874 1.000
Chain 1: 700 -16023.352 0.793 0.360
Chain 1: 800 -15153.850 0.701 0.360
Chain 1: 900 -15845.824 0.628 0.307
Chain 1: 1000 -12263.950 0.594 0.307
Chain 1: 1100 -10620.903 0.510 0.292 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -12034.503 0.325 0.155
Chain 1: 1300 -9926.036 0.335 0.212
Chain 1: 1400 -10200.213 0.333 0.212
Chain 1: 1500 -11595.408 0.309 0.155
Chain 1: 1600 -10922.343 0.139 0.120
Chain 1: 1700 -9608.632 0.122 0.120
Chain 1: 1800 -10987.132 0.129 0.125
Chain 1: 1900 -10634.278 0.128 0.125
Chain 1: 2000 -19280.119 0.144 0.125
Chain 1: 2100 -18919.592 0.130 0.120
Chain 1: 2200 -10044.518 0.207 0.125
Chain 1: 2300 -9327.997 0.193 0.120
Chain 1: 2400 -9956.554 0.197 0.120
Chain 1: 2500 -11153.030 0.196 0.107
Chain 1: 2600 -9329.905 0.209 0.125
Chain 1: 2700 -9856.349 0.201 0.107
Chain 1: 2800 -16417.612 0.228 0.107
Chain 1: 2900 -10130.178 0.287 0.195
Chain 1: 3000 -16310.323 0.280 0.195
Chain 1: 3100 -9096.628 0.357 0.379
Chain 1: 3200 -9326.231 0.271 0.195
Chain 1: 3300 -9420.885 0.265 0.195
Chain 1: 3400 -14235.337 0.292 0.338
Chain 1: 3500 -9146.750 0.337 0.379
Chain 1: 3600 -9922.780 0.325 0.379
Chain 1: 3700 -8886.830 0.332 0.379
Chain 1: 3800 -8580.454 0.295 0.338
Chain 1: 3900 -9378.308 0.242 0.117
Chain 1: 4000 -8646.541 0.212 0.085
Chain 1: 4100 -9201.588 0.139 0.085
Chain 1: 4200 -8792.703 0.141 0.085
Chain 1: 4300 -14849.545 0.181 0.085
Chain 1: 4400 -9739.005 0.200 0.085
Chain 1: 4500 -9552.805 0.146 0.085
Chain 1: 4600 -13010.183 0.165 0.085
Chain 1: 4700 -8380.739 0.208 0.085
Chain 1: 4800 -8746.871 0.209 0.085
Chain 1: 4900 -8620.269 0.202 0.085
Chain 1: 5000 -10594.810 0.212 0.186
Chain 1: 5100 -8805.233 0.226 0.203
Chain 1: 5200 -9935.878 0.233 0.203
Chain 1: 5300 -9494.123 0.197 0.186
Chain 1: 5400 -11115.004 0.159 0.146
Chain 1: 5500 -8544.987 0.187 0.186
Chain 1: 5600 -8694.889 0.162 0.146
Chain 1: 5700 -8559.883 0.109 0.114
Chain 1: 5800 -8828.219 0.107 0.114
Chain 1: 5900 -11620.306 0.130 0.146
Chain 1: 6000 -8692.954 0.145 0.146
Chain 1: 6100 -8889.434 0.127 0.114
Chain 1: 6200 -12310.368 0.143 0.146
Chain 1: 6300 -14352.972 0.153 0.146
Chain 1: 6400 -12545.976 0.153 0.144
Chain 1: 6500 -9377.721 0.156 0.144
Chain 1: 6600 -8547.276 0.164 0.144
Chain 1: 6700 -8864.024 0.166 0.144
Chain 1: 6800 -9663.156 0.172 0.144
Chain 1: 6900 -11619.569 0.164 0.144
Chain 1: 7000 -8885.251 0.162 0.144
Chain 1: 7100 -8838.567 0.160 0.144
Chain 1: 7200 -8588.276 0.135 0.142
Chain 1: 7300 -8724.287 0.122 0.097
Chain 1: 7400 -8393.050 0.112 0.083
Chain 1: 7500 -9021.582 0.085 0.070
Chain 1: 7600 -8359.940 0.083 0.070
Chain 1: 7700 -9732.840 0.094 0.079
Chain 1: 7800 -11848.649 0.103 0.079
Chain 1: 7900 -8637.216 0.124 0.079
Chain 1: 8000 -8432.058 0.095 0.070
Chain 1: 8100 -9337.338 0.105 0.079
Chain 1: 8200 -10151.699 0.110 0.080
Chain 1: 8300 -8197.075 0.132 0.097
Chain 1: 8400 -12025.513 0.160 0.141
Chain 1: 8500 -8275.970 0.198 0.179
Chain 1: 8600 -10059.428 0.208 0.179
Chain 1: 8700 -10880.840 0.201 0.179
Chain 1: 8800 -8128.596 0.217 0.238
Chain 1: 8900 -9392.178 0.194 0.177
Chain 1: 9000 -8891.091 0.197 0.177
Chain 1: 9100 -8538.557 0.191 0.177
Chain 1: 9200 -8917.163 0.188 0.177
Chain 1: 9300 -12949.932 0.195 0.177
Chain 1: 9400 -8412.878 0.217 0.177
Chain 1: 9500 -8693.583 0.175 0.135
Chain 1: 9600 -8190.902 0.163 0.075
Chain 1: 9700 -8241.217 0.156 0.061
Chain 1: 9800 -8848.460 0.129 0.061
Chain 1: 9900 -9042.123 0.118 0.056
Chain 1: 10000 -8945.883 0.114 0.042
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001572 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.72 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56881.433 1.000 1.000
Chain 1: 200 -17365.583 1.638 2.276
Chain 1: 300 -8678.740 1.425 1.001
Chain 1: 400 -8333.557 1.079 1.001
Chain 1: 500 -8178.570 0.867 1.000
Chain 1: 600 -8078.881 0.725 1.000
Chain 1: 700 -7994.686 0.623 0.041
Chain 1: 800 -7628.038 0.551 0.048
Chain 1: 900 -7932.468 0.494 0.041
Chain 1: 1000 -7623.800 0.449 0.041
Chain 1: 1100 -7755.261 0.350 0.040
Chain 1: 1200 -7693.874 0.124 0.038
Chain 1: 1300 -7708.840 0.024 0.019
Chain 1: 1400 -7806.531 0.021 0.017
Chain 1: 1500 -7577.219 0.022 0.017
Chain 1: 1600 -7695.943 0.022 0.017
Chain 1: 1700 -7456.625 0.024 0.030
Chain 1: 1800 -7579.695 0.021 0.017
Chain 1: 1900 -7500.413 0.018 0.016
Chain 1: 2000 -7519.243 0.015 0.015
Chain 1: 2100 -7537.899 0.013 0.013
Chain 1: 2200 -7657.851 0.014 0.015
Chain 1: 2300 -7552.409 0.015 0.015
Chain 1: 2400 -7599.153 0.015 0.015
Chain 1: 2500 -7449.721 0.014 0.015
Chain 1: 2600 -7486.612 0.012 0.014
Chain 1: 2700 -7519.134 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003155 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.55 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86346.189 1.000 1.000
Chain 1: 200 -13435.867 3.213 5.427
Chain 1: 300 -9829.821 2.264 1.000
Chain 1: 400 -10553.812 1.715 1.000
Chain 1: 500 -8779.630 1.413 0.367
Chain 1: 600 -8307.345 1.187 0.367
Chain 1: 700 -8353.122 1.018 0.202
Chain 1: 800 -8862.650 0.898 0.202
Chain 1: 900 -8624.190 0.801 0.069
Chain 1: 1000 -8434.415 0.723 0.069
Chain 1: 1100 -8685.375 0.626 0.057 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8208.100 0.089 0.057
Chain 1: 1300 -8414.489 0.055 0.057
Chain 1: 1400 -8547.166 0.050 0.029
Chain 1: 1500 -8413.562 0.031 0.028
Chain 1: 1600 -8525.638 0.027 0.025
Chain 1: 1700 -8608.617 0.027 0.025
Chain 1: 1800 -8202.493 0.027 0.025
Chain 1: 1900 -8300.874 0.025 0.023
Chain 1: 2000 -8272.543 0.023 0.016
Chain 1: 2100 -8392.676 0.022 0.016
Chain 1: 2200 -8185.253 0.018 0.016
Chain 1: 2300 -8336.540 0.018 0.016
Chain 1: 2400 -8343.646 0.016 0.014
Chain 1: 2500 -8314.347 0.015 0.013
Chain 1: 2600 -8313.250 0.014 0.012
Chain 1: 2700 -8225.369 0.014 0.012
Chain 1: 2800 -8191.485 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003268 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.68 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8409996.152 1.000 1.000
Chain 1: 200 -1584936.413 2.653 4.306
Chain 1: 300 -892201.643 2.028 1.000
Chain 1: 400 -458387.009 1.757 1.000
Chain 1: 500 -358674.894 1.461 0.946
Chain 1: 600 -233441.261 1.307 0.946
Chain 1: 700 -119406.133 1.257 0.946
Chain 1: 800 -86556.854 1.147 0.946
Chain 1: 900 -66846.945 1.053 0.776
Chain 1: 1000 -51598.756 0.977 0.776
Chain 1: 1100 -39042.188 0.909 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38209.561 0.481 0.380
Chain 1: 1300 -26135.957 0.449 0.380
Chain 1: 1400 -25850.215 0.356 0.322
Chain 1: 1500 -22430.511 0.343 0.322
Chain 1: 1600 -21644.620 0.293 0.296
Chain 1: 1700 -20515.198 0.203 0.295
Chain 1: 1800 -20458.293 0.165 0.152
Chain 1: 1900 -20784.246 0.137 0.055
Chain 1: 2000 -19294.108 0.116 0.055
Chain 1: 2100 -19532.437 0.085 0.036
Chain 1: 2200 -19759.152 0.084 0.036
Chain 1: 2300 -19376.182 0.039 0.020
Chain 1: 2400 -19148.321 0.039 0.020
Chain 1: 2500 -18950.483 0.025 0.016
Chain 1: 2600 -18580.726 0.024 0.016
Chain 1: 2700 -18537.661 0.018 0.012
Chain 1: 2800 -18254.739 0.020 0.015
Chain 1: 2900 -18535.884 0.020 0.015
Chain 1: 3000 -18522.027 0.012 0.012
Chain 1: 3100 -18607.042 0.011 0.012
Chain 1: 3200 -18297.787 0.012 0.015
Chain 1: 3300 -18502.428 0.011 0.012
Chain 1: 3400 -17977.590 0.013 0.015
Chain 1: 3500 -18589.178 0.015 0.015
Chain 1: 3600 -17896.205 0.017 0.015
Chain 1: 3700 -18282.816 0.019 0.017
Chain 1: 3800 -17243.112 0.023 0.021
Chain 1: 3900 -17239.289 0.022 0.021
Chain 1: 4000 -17356.565 0.022 0.021
Chain 1: 4100 -17270.412 0.022 0.021
Chain 1: 4200 -17086.739 0.022 0.021
Chain 1: 4300 -17225.048 0.021 0.021
Chain 1: 4400 -17181.992 0.019 0.011
Chain 1: 4500 -17084.545 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00121 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.1 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12453.394 1.000 1.000
Chain 1: 200 -9366.825 0.665 1.000
Chain 1: 300 -8200.627 0.491 0.330
Chain 1: 400 -8359.819 0.373 0.330
Chain 1: 500 -8278.915 0.300 0.142
Chain 1: 600 -8142.026 0.253 0.142
Chain 1: 700 -8068.748 0.218 0.019
Chain 1: 800 -8078.501 0.191 0.019
Chain 1: 900 -7982.615 0.171 0.017
Chain 1: 1000 -8193.716 0.157 0.019
Chain 1: 1100 -8103.529 0.058 0.017
Chain 1: 1200 -8087.554 0.025 0.012
Chain 1: 1300 -8039.654 0.011 0.011
Chain 1: 1400 -8061.212 0.010 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001434 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.34 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58180.389 1.000 1.000
Chain 1: 200 -17747.734 1.639 2.278
Chain 1: 300 -8672.743 1.442 1.046
Chain 1: 400 -8195.234 1.096 1.046
Chain 1: 500 -8075.956 0.880 1.000
Chain 1: 600 -8611.426 0.743 1.000
Chain 1: 700 -8218.887 0.644 0.062
Chain 1: 800 -8390.601 0.566 0.062
Chain 1: 900 -7813.918 0.511 0.062
Chain 1: 1000 -7799.361 0.460 0.062
Chain 1: 1100 -7641.037 0.362 0.058
Chain 1: 1200 -7621.331 0.135 0.048
Chain 1: 1300 -7743.169 0.032 0.021
Chain 1: 1400 -7811.209 0.027 0.020
Chain 1: 1500 -7566.175 0.029 0.021
Chain 1: 1600 -7761.033 0.025 0.021
Chain 1: 1700 -7479.547 0.024 0.021
Chain 1: 1800 -7576.624 0.023 0.021
Chain 1: 1900 -7615.359 0.016 0.016
Chain 1: 2000 -7529.776 0.017 0.016
Chain 1: 2100 -7523.340 0.015 0.013
Chain 1: 2200 -7697.423 0.017 0.016
Chain 1: 2300 -7517.625 0.018 0.023
Chain 1: 2400 -7603.533 0.018 0.023
Chain 1: 2500 -7542.120 0.016 0.013
Chain 1: 2600 -7492.613 0.014 0.011
Chain 1: 2700 -7531.930 0.011 0.011
Chain 1: 2800 -7473.350 0.010 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002969 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.69 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86566.336 1.000 1.000
Chain 1: 200 -13534.186 3.198 5.396
Chain 1: 300 -9971.358 2.251 1.000
Chain 1: 400 -10721.610 1.706 1.000
Chain 1: 500 -8861.390 1.407 0.357
Chain 1: 600 -8499.966 1.179 0.357
Chain 1: 700 -8534.448 1.011 0.210
Chain 1: 800 -8955.630 0.891 0.210
Chain 1: 900 -8715.473 0.795 0.070
Chain 1: 1000 -8478.845 0.718 0.070
Chain 1: 1100 -8858.654 0.623 0.047 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8490.326 0.087 0.043
Chain 1: 1300 -8574.218 0.052 0.043
Chain 1: 1400 -8526.276 0.046 0.043
Chain 1: 1500 -8559.436 0.025 0.028
Chain 1: 1600 -8559.150 0.021 0.028
Chain 1: 1700 -8482.099 0.022 0.028
Chain 1: 1800 -8366.452 0.018 0.014
Chain 1: 1900 -8485.677 0.017 0.014
Chain 1: 2000 -8445.817 0.015 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003689 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 36.89 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8410896.174 1.000 1.000
Chain 1: 200 -1586266.110 2.651 4.302
Chain 1: 300 -890793.399 2.028 1.000
Chain 1: 400 -457865.545 1.757 1.000
Chain 1: 500 -357948.685 1.462 0.946
Chain 1: 600 -232893.380 1.307 0.946
Chain 1: 700 -119159.759 1.257 0.946
Chain 1: 800 -86398.622 1.147 0.946
Chain 1: 900 -66746.878 1.053 0.781
Chain 1: 1000 -51550.170 0.977 0.781
Chain 1: 1100 -39040.541 0.909 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38212.715 0.481 0.379
Chain 1: 1300 -26190.788 0.449 0.379
Chain 1: 1400 -25909.491 0.355 0.320
Chain 1: 1500 -22503.643 0.342 0.320
Chain 1: 1600 -21721.520 0.292 0.295
Chain 1: 1700 -20598.309 0.202 0.294
Chain 1: 1800 -20542.928 0.165 0.151
Chain 1: 1900 -20868.627 0.137 0.055
Chain 1: 2000 -19382.284 0.115 0.055
Chain 1: 2100 -19620.393 0.084 0.036
Chain 1: 2200 -19846.423 0.083 0.036
Chain 1: 2300 -19464.139 0.039 0.020
Chain 1: 2400 -19236.399 0.039 0.020
Chain 1: 2500 -19038.411 0.025 0.016
Chain 1: 2600 -18669.063 0.023 0.016
Chain 1: 2700 -18626.132 0.018 0.012
Chain 1: 2800 -18343.212 0.020 0.015
Chain 1: 2900 -18624.231 0.019 0.015
Chain 1: 3000 -18610.416 0.012 0.012
Chain 1: 3100 -18695.365 0.011 0.012
Chain 1: 3200 -18386.335 0.012 0.015
Chain 1: 3300 -18590.816 0.011 0.012
Chain 1: 3400 -18066.281 0.013 0.015
Chain 1: 3500 -18677.337 0.015 0.015
Chain 1: 3600 -17985.051 0.017 0.015
Chain 1: 3700 -18371.091 0.018 0.017
Chain 1: 3800 -17332.419 0.023 0.021
Chain 1: 3900 -17328.597 0.021 0.021
Chain 1: 4000 -17445.909 0.022 0.021
Chain 1: 4100 -17359.771 0.022 0.021
Chain 1: 4200 -17176.336 0.021 0.021
Chain 1: 4300 -17314.499 0.021 0.021
Chain 1: 4400 -17271.603 0.019 0.011
Chain 1: 4500 -17174.185 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001122 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 11.22 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -11944.269 1.000 1.000
Chain 1: 200 -8868.762 0.673 1.000
Chain 1: 300 -7915.829 0.489 0.347
Chain 1: 400 -8014.219 0.370 0.347
Chain 1: 500 -7729.971 0.303 0.120
Chain 1: 600 -7732.574 0.253 0.120
Chain 1: 700 -7713.156 0.217 0.037
Chain 1: 800 -7693.714 0.190 0.037
Chain 1: 900 -7884.390 0.172 0.024
Chain 1: 1000 -7731.534 0.157 0.024
Chain 1: 1100 -7899.920 0.059 0.021
Chain 1: 1200 -7696.169 0.027 0.021
Chain 1: 1300 -7671.056 0.015 0.020
Chain 1: 1400 -7680.948 0.014 0.020
Chain 1: 1500 -7764.618 0.011 0.011
Chain 1: 1600 -7751.784 0.011 0.011
Chain 1: 1700 -7663.010 0.012 0.012
Chain 1: 1800 -7649.537 0.012 0.012
Chain 1: 1900 -7631.721 0.010 0.011
Chain 1: 2000 -7613.000 0.008 0.003 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001457 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.57 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56679.778 1.000 1.000
Chain 1: 200 -16959.910 1.671 2.342
Chain 1: 300 -8495.207 1.446 1.000
Chain 1: 400 -8629.280 1.088 1.000
Chain 1: 500 -8011.803 0.886 0.996
Chain 1: 600 -8658.146 0.751 0.996
Chain 1: 700 -7831.140 0.659 0.106
Chain 1: 800 -8253.351 0.583 0.106
Chain 1: 900 -7616.539 0.527 0.084
Chain 1: 1000 -7684.773 0.475 0.084
Chain 1: 1100 -7648.642 0.376 0.077
Chain 1: 1200 -7531.646 0.143 0.075
Chain 1: 1300 -7635.108 0.045 0.051
Chain 1: 1400 -7749.179 0.045 0.051
Chain 1: 1500 -7559.295 0.040 0.025
Chain 1: 1600 -7464.487 0.034 0.016
Chain 1: 1700 -7446.234 0.023 0.015
Chain 1: 1800 -7481.437 0.019 0.014
Chain 1: 1900 -7529.011 0.011 0.013
Chain 1: 2000 -7524.334 0.010 0.013
Chain 1: 2100 -7558.086 0.010 0.013
Chain 1: 2200 -7621.608 0.009 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003236 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.36 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86001.309 1.000 1.000
Chain 1: 200 -12994.597 3.309 5.618
Chain 1: 300 -9496.030 2.329 1.000
Chain 1: 400 -10253.843 1.765 1.000
Chain 1: 500 -8378.362 1.457 0.368
Chain 1: 600 -8208.636 1.218 0.368
Chain 1: 700 -8356.080 1.046 0.224
Chain 1: 800 -8674.070 0.920 0.224
Chain 1: 900 -8355.240 0.822 0.074
Chain 1: 1000 -8128.906 0.743 0.074
Chain 1: 1100 -8346.509 0.645 0.038 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8115.256 0.086 0.037
Chain 1: 1300 -8278.535 0.051 0.028
Chain 1: 1400 -8203.296 0.045 0.028
Chain 1: 1500 -8153.873 0.023 0.026
Chain 1: 1600 -8152.299 0.021 0.026
Chain 1: 1700 -8095.338 0.020 0.026
Chain 1: 1800 -7973.841 0.018 0.020
Chain 1: 1900 -8086.060 0.015 0.015
Chain 1: 2000 -8048.960 0.013 0.014
Chain 1: 2100 -8193.321 0.012 0.014
Chain 1: 2200 -7974.932 0.012 0.014
Chain 1: 2300 -8107.819 0.012 0.014
Chain 1: 2400 -8002.753 0.012 0.014
Chain 1: 2500 -8057.641 0.012 0.014
Chain 1: 2600 -8070.359 0.012 0.014
Chain 1: 2700 -7991.461 0.013 0.014
Chain 1: 2800 -7977.406 0.011 0.013
Chain 1: 2900 -7967.176 0.010 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003316 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.16 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8413935.315 1.000 1.000
Chain 1: 200 -1585650.750 2.653 4.306
Chain 1: 300 -890743.751 2.029 1.000
Chain 1: 400 -457690.354 1.758 1.000
Chain 1: 500 -357778.777 1.462 0.946
Chain 1: 600 -232544.012 1.308 0.946
Chain 1: 700 -118674.021 1.259 0.946
Chain 1: 800 -85879.467 1.149 0.946
Chain 1: 900 -66208.535 1.054 0.780
Chain 1: 1000 -50995.166 0.979 0.780
Chain 1: 1100 -38478.004 0.911 0.539 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37644.897 0.483 0.382
Chain 1: 1300 -25624.919 0.452 0.382
Chain 1: 1400 -25341.381 0.358 0.325
Chain 1: 1500 -21936.252 0.346 0.325
Chain 1: 1600 -21153.639 0.296 0.298
Chain 1: 1700 -20031.230 0.205 0.297
Chain 1: 1800 -19975.668 0.167 0.155
Chain 1: 1900 -20300.927 0.139 0.056
Chain 1: 2000 -18815.849 0.117 0.056
Chain 1: 2100 -19053.854 0.086 0.037
Chain 1: 2200 -19279.480 0.085 0.037
Chain 1: 2300 -18897.654 0.040 0.020
Chain 1: 2400 -18670.098 0.040 0.020
Chain 1: 2500 -18472.091 0.026 0.016
Chain 1: 2600 -18103.222 0.024 0.016
Chain 1: 2700 -18060.434 0.019 0.012
Chain 1: 2800 -17777.717 0.020 0.016
Chain 1: 2900 -18058.511 0.020 0.016
Chain 1: 3000 -18044.729 0.012 0.012
Chain 1: 3100 -18129.602 0.011 0.012
Chain 1: 3200 -17820.888 0.012 0.016
Chain 1: 3300 -18025.119 0.011 0.012
Chain 1: 3400 -17501.143 0.013 0.016
Chain 1: 3500 -18111.374 0.015 0.016
Chain 1: 3600 -17420.166 0.017 0.016
Chain 1: 3700 -17805.399 0.019 0.017
Chain 1: 3800 -16768.404 0.024 0.022
Chain 1: 3900 -16764.627 0.022 0.022
Chain 1: 4000 -16881.923 0.023 0.022
Chain 1: 4100 -16795.877 0.023 0.022
Chain 1: 4200 -16612.811 0.022 0.022
Chain 1: 4300 -16750.711 0.022 0.022
Chain 1: 4400 -16708.116 0.019 0.011
Chain 1: 4500 -16610.758 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001406 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.06 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48174.848 1.000 1.000
Chain 1: 200 -16233.881 1.484 1.968
Chain 1: 300 -14909.097 1.019 1.000
Chain 1: 400 -24214.038 0.860 1.000
Chain 1: 500 -10987.648 0.929 1.000
Chain 1: 600 -16260.093 0.828 1.000
Chain 1: 700 -14408.737 0.728 0.384
Chain 1: 800 -10325.070 0.687 0.396
Chain 1: 900 -10606.091 0.613 0.384
Chain 1: 1000 -11904.928 0.563 0.384
Chain 1: 1100 -12409.018 0.467 0.324
Chain 1: 1200 -10271.960 0.291 0.208
Chain 1: 1300 -11908.658 0.296 0.208
Chain 1: 1400 -20884.468 0.300 0.208
Chain 1: 1500 -20471.766 0.182 0.137
Chain 1: 1600 -12094.020 0.219 0.137
Chain 1: 1700 -10363.996 0.223 0.167
Chain 1: 1800 -9485.807 0.192 0.137
Chain 1: 1900 -10574.770 0.200 0.137
Chain 1: 2000 -9366.498 0.202 0.137
Chain 1: 2100 -9588.481 0.200 0.137
Chain 1: 2200 -9244.118 0.183 0.129
Chain 1: 2300 -12036.485 0.193 0.129
Chain 1: 2400 -8734.625 0.187 0.129
Chain 1: 2500 -11221.654 0.208 0.167
Chain 1: 2600 -9409.872 0.158 0.167
Chain 1: 2700 -8883.564 0.147 0.129
Chain 1: 2800 -9714.937 0.146 0.129
Chain 1: 2900 -17067.588 0.179 0.193
Chain 1: 3000 -8878.812 0.258 0.222
Chain 1: 3100 -8963.433 0.257 0.222
Chain 1: 3200 -8627.840 0.257 0.222
Chain 1: 3300 -8708.510 0.235 0.193
Chain 1: 3400 -8641.477 0.198 0.086
Chain 1: 3500 -8709.348 0.176 0.059
Chain 1: 3600 -12590.949 0.188 0.059
Chain 1: 3700 -10394.618 0.203 0.086
Chain 1: 3800 -14099.620 0.221 0.211
Chain 1: 3900 -8654.554 0.241 0.211
Chain 1: 4000 -9611.105 0.158 0.100
Chain 1: 4100 -8798.223 0.167 0.100
Chain 1: 4200 -12174.619 0.191 0.211
Chain 1: 4300 -13516.039 0.200 0.211
Chain 1: 4400 -10144.091 0.232 0.263
Chain 1: 4500 -8660.681 0.248 0.263
Chain 1: 4600 -8178.896 0.223 0.211
Chain 1: 4700 -8708.907 0.208 0.171
Chain 1: 4800 -11165.817 0.204 0.171
Chain 1: 4900 -8483.173 0.173 0.171
Chain 1: 5000 -14297.337 0.204 0.220
Chain 1: 5100 -9346.359 0.247 0.277
Chain 1: 5200 -12470.494 0.245 0.251
Chain 1: 5300 -9460.266 0.266 0.316
Chain 1: 5400 -12503.453 0.258 0.251
Chain 1: 5500 -8581.298 0.286 0.316
Chain 1: 5600 -8266.587 0.284 0.316
Chain 1: 5700 -11595.367 0.307 0.316
Chain 1: 5800 -8541.984 0.320 0.318
Chain 1: 5900 -8207.137 0.293 0.318
Chain 1: 6000 -9504.174 0.266 0.287
Chain 1: 6100 -8833.965 0.220 0.251
Chain 1: 6200 -8071.694 0.205 0.243
Chain 1: 6300 -8683.551 0.180 0.136
Chain 1: 6400 -12632.953 0.187 0.136
Chain 1: 6500 -8262.801 0.194 0.136
Chain 1: 6600 -8246.163 0.191 0.136
Chain 1: 6700 -12628.854 0.197 0.136
Chain 1: 6800 -9080.415 0.200 0.136
Chain 1: 6900 -11782.939 0.219 0.229
Chain 1: 7000 -8760.550 0.240 0.313
Chain 1: 7100 -8354.704 0.237 0.313
Chain 1: 7200 -9770.987 0.242 0.313
Chain 1: 7300 -11645.565 0.251 0.313
Chain 1: 7400 -8568.922 0.256 0.345
Chain 1: 7500 -12211.765 0.233 0.298
Chain 1: 7600 -8503.079 0.276 0.345
Chain 1: 7700 -8219.105 0.245 0.298
Chain 1: 7800 -9254.456 0.217 0.229
Chain 1: 7900 -9794.268 0.199 0.161
Chain 1: 8000 -10968.868 0.176 0.145
Chain 1: 8100 -11953.111 0.179 0.145
Chain 1: 8200 -10695.338 0.176 0.118
Chain 1: 8300 -10355.036 0.163 0.112
Chain 1: 8400 -8158.562 0.155 0.112
Chain 1: 8500 -8224.486 0.125 0.107
Chain 1: 8600 -9992.367 0.100 0.107
Chain 1: 8700 -8847.338 0.109 0.112
Chain 1: 8800 -8043.648 0.108 0.107
Chain 1: 8900 -8559.170 0.108 0.107
Chain 1: 9000 -9023.296 0.103 0.100
Chain 1: 9100 -8563.038 0.100 0.100
Chain 1: 9200 -9101.462 0.094 0.060
Chain 1: 9300 -10677.758 0.106 0.100
Chain 1: 9400 -8090.947 0.111 0.100
Chain 1: 9500 -8131.745 0.110 0.100
Chain 1: 9600 -8327.354 0.095 0.060
Chain 1: 9700 -8262.234 0.083 0.059
Chain 1: 9800 -9106.498 0.082 0.059
Chain 1: 9900 -8502.235 0.083 0.059
Chain 1: 10000 -8960.693 0.083 0.059
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001429 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.29 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61108.482 1.000 1.000
Chain 1: 200 -17202.205 1.776 2.552
Chain 1: 300 -8601.256 1.517 1.000
Chain 1: 400 -8103.716 1.153 1.000
Chain 1: 500 -7887.134 0.928 1.000
Chain 1: 600 -8543.565 0.786 1.000
Chain 1: 700 -7855.603 0.687 0.088
Chain 1: 800 -7989.643 0.603 0.088
Chain 1: 900 -7836.999 0.538 0.077
Chain 1: 1000 -7789.775 0.485 0.077
Chain 1: 1100 -7646.447 0.387 0.061
Chain 1: 1200 -7616.379 0.132 0.027
Chain 1: 1300 -7681.632 0.033 0.019
Chain 1: 1400 -7707.954 0.027 0.019
Chain 1: 1500 -7588.779 0.026 0.017
Chain 1: 1600 -7515.475 0.019 0.016
Chain 1: 1700 -7499.345 0.010 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002932 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.32 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85935.374 1.000 1.000
Chain 1: 200 -12951.214 3.318 5.635
Chain 1: 300 -9532.354 2.331 1.000
Chain 1: 400 -9957.011 1.759 1.000
Chain 1: 500 -8431.151 1.444 0.359
Chain 1: 600 -8458.033 1.203 0.359
Chain 1: 700 -8474.502 1.032 0.181
Chain 1: 800 -8628.387 0.905 0.181
Chain 1: 900 -8428.519 0.807 0.043
Chain 1: 1000 -8251.506 0.729 0.043
Chain 1: 1100 -8461.637 0.631 0.025 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8217.521 0.070 0.025
Chain 1: 1300 -8371.893 0.036 0.024
Chain 1: 1400 -8268.037 0.033 0.021
Chain 1: 1500 -8263.198 0.015 0.018
Chain 1: 1600 -8369.022 0.016 0.018
Chain 1: 1700 -8438.507 0.017 0.018
Chain 1: 1800 -8105.573 0.019 0.021
Chain 1: 1900 -8197.965 0.018 0.018
Chain 1: 2000 -8172.224 0.016 0.013
Chain 1: 2100 -8323.999 0.016 0.013
Chain 1: 2200 -8099.310 0.015 0.013
Chain 1: 2300 -8175.112 0.014 0.013
Chain 1: 2400 -8230.856 0.014 0.011
Chain 1: 2500 -8200.485 0.014 0.011
Chain 1: 2600 -8191.707 0.013 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002582 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.82 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8412452.902 1.000 1.000
Chain 1: 200 -1587839.376 2.649 4.298
Chain 1: 300 -891332.084 2.026 1.000
Chain 1: 400 -457297.443 1.757 1.000
Chain 1: 500 -357156.442 1.462 0.949
Chain 1: 600 -232005.754 1.308 0.949
Chain 1: 700 -118410.918 1.258 0.949
Chain 1: 800 -85649.545 1.149 0.949
Chain 1: 900 -66031.526 1.054 0.781
Chain 1: 1000 -50846.072 0.979 0.781
Chain 1: 1100 -38352.634 0.911 0.539 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37521.195 0.484 0.383
Chain 1: 1300 -25534.958 0.452 0.383
Chain 1: 1400 -25252.087 0.359 0.326
Chain 1: 1500 -21855.685 0.346 0.326
Chain 1: 1600 -21074.910 0.296 0.299
Chain 1: 1700 -19957.356 0.206 0.297
Chain 1: 1800 -19902.553 0.168 0.155
Chain 1: 1900 -20227.342 0.139 0.056
Chain 1: 2000 -18745.638 0.117 0.056
Chain 1: 2100 -18983.414 0.086 0.037
Chain 1: 2200 -19208.267 0.085 0.037
Chain 1: 2300 -18827.279 0.040 0.020
Chain 1: 2400 -18600.002 0.040 0.020
Chain 1: 2500 -18401.771 0.026 0.016
Chain 1: 2600 -18033.569 0.024 0.016
Chain 1: 2700 -17991.062 0.019 0.013
Chain 1: 2800 -17708.446 0.020 0.016
Chain 1: 2900 -17989.002 0.020 0.016
Chain 1: 3000 -17975.300 0.012 0.013
Chain 1: 3100 -18060.047 0.011 0.012
Chain 1: 3200 -17751.720 0.012 0.016
Chain 1: 3300 -17955.701 0.011 0.012
Chain 1: 3400 -17432.275 0.013 0.016
Chain 1: 3500 -18041.530 0.015 0.016
Chain 1: 3600 -17351.702 0.017 0.016
Chain 1: 3700 -17735.871 0.019 0.017
Chain 1: 3800 -16700.855 0.024 0.022
Chain 1: 3900 -16697.139 0.022 0.022
Chain 1: 4000 -16814.450 0.023 0.022
Chain 1: 4100 -16728.432 0.023 0.022
Chain 1: 4200 -16545.882 0.022 0.022
Chain 1: 4300 -16683.446 0.022 0.022
Chain 1: 4400 -16641.226 0.019 0.011
Chain 1: 4500 -16543.936 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001298 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.98 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49513.390 1.000 1.000
Chain 1: 200 -15323.327 1.616 2.231
Chain 1: 300 -13305.073 1.128 1.000
Chain 1: 400 -15923.696 0.887 1.000
Chain 1: 500 -13109.407 0.752 0.215
Chain 1: 600 -15221.886 0.650 0.215
Chain 1: 700 -16505.980 0.568 0.164
Chain 1: 800 -12334.658 0.540 0.215
Chain 1: 900 -13301.797 0.488 0.164
Chain 1: 1000 -12219.506 0.448 0.164
Chain 1: 1100 -11565.462 0.353 0.152
Chain 1: 1200 -12535.925 0.138 0.139
Chain 1: 1300 -10806.416 0.139 0.139
Chain 1: 1400 -13305.231 0.141 0.139
Chain 1: 1500 -10679.338 0.144 0.139
Chain 1: 1600 -13036.682 0.149 0.160
Chain 1: 1700 -9912.687 0.172 0.181
Chain 1: 1800 -16814.392 0.180 0.181
Chain 1: 1900 -11319.465 0.221 0.188
Chain 1: 2000 -10173.846 0.223 0.188
Chain 1: 2100 -10080.863 0.218 0.188
Chain 1: 2200 -9863.387 0.213 0.188
Chain 1: 2300 -9823.135 0.197 0.188
Chain 1: 2400 -14838.558 0.212 0.246
Chain 1: 2500 -16388.044 0.197 0.181
Chain 1: 2600 -10065.857 0.242 0.315
Chain 1: 2700 -11013.964 0.219 0.113
Chain 1: 2800 -16928.863 0.213 0.113
Chain 1: 2900 -9950.277 0.235 0.113
Chain 1: 3000 -16067.632 0.261 0.338
Chain 1: 3100 -8976.487 0.339 0.349
Chain 1: 3200 -12477.002 0.365 0.349
Chain 1: 3300 -10492.422 0.384 0.349
Chain 1: 3400 -9322.266 0.363 0.349
Chain 1: 3500 -10917.864 0.368 0.349
Chain 1: 3600 -9284.251 0.322 0.281
Chain 1: 3700 -9601.840 0.317 0.281
Chain 1: 3800 -10027.289 0.286 0.189
Chain 1: 3900 -9531.787 0.222 0.176
Chain 1: 4000 -13233.650 0.211 0.176
Chain 1: 4100 -14492.298 0.141 0.146
Chain 1: 4200 -9964.141 0.159 0.146
Chain 1: 4300 -10467.542 0.144 0.126
Chain 1: 4400 -9728.494 0.139 0.087
Chain 1: 4500 -10667.863 0.134 0.087
Chain 1: 4600 -8883.180 0.136 0.087
Chain 1: 4700 -13880.951 0.169 0.088
Chain 1: 4800 -9254.862 0.215 0.201
Chain 1: 4900 -19699.354 0.262 0.280
Chain 1: 5000 -11811.435 0.301 0.360
Chain 1: 5100 -8895.009 0.325 0.360
Chain 1: 5200 -9155.285 0.283 0.328
Chain 1: 5300 -11868.439 0.301 0.328
Chain 1: 5400 -12615.334 0.299 0.328
Chain 1: 5500 -8868.913 0.333 0.360
Chain 1: 5600 -13714.990 0.348 0.360
Chain 1: 5700 -9926.151 0.350 0.382
Chain 1: 5800 -9838.388 0.301 0.353
Chain 1: 5900 -16645.848 0.289 0.353
Chain 1: 6000 -11640.442 0.265 0.353
Chain 1: 6100 -9638.278 0.253 0.353
Chain 1: 6200 -8649.316 0.262 0.353
Chain 1: 6300 -9023.132 0.243 0.353
Chain 1: 6400 -8492.229 0.243 0.353
Chain 1: 6500 -9450.961 0.211 0.208
Chain 1: 6600 -14027.682 0.208 0.208
Chain 1: 6700 -14130.912 0.171 0.114
Chain 1: 6800 -8881.560 0.229 0.208
Chain 1: 6900 -8509.433 0.193 0.114
Chain 1: 7000 -9964.407 0.164 0.114
Chain 1: 7100 -8508.152 0.161 0.114
Chain 1: 7200 -11589.766 0.176 0.146
Chain 1: 7300 -8427.640 0.209 0.171
Chain 1: 7400 -10371.604 0.222 0.187
Chain 1: 7500 -9968.112 0.215 0.187
Chain 1: 7600 -9151.949 0.192 0.171
Chain 1: 7700 -12896.836 0.220 0.187
Chain 1: 7800 -8873.643 0.206 0.187
Chain 1: 7900 -8432.560 0.207 0.187
Chain 1: 8000 -8465.966 0.193 0.187
Chain 1: 8100 -8624.431 0.178 0.187
Chain 1: 8200 -9565.558 0.161 0.098
Chain 1: 8300 -12388.985 0.146 0.098
Chain 1: 8400 -9951.984 0.152 0.098
Chain 1: 8500 -8345.076 0.167 0.193
Chain 1: 8600 -8719.772 0.163 0.193
Chain 1: 8700 -9093.528 0.138 0.098
Chain 1: 8800 -8911.604 0.094 0.052
Chain 1: 8900 -9733.074 0.097 0.084
Chain 1: 9000 -8574.356 0.111 0.098
Chain 1: 9100 -8459.151 0.110 0.098
Chain 1: 9200 -9458.774 0.111 0.106
Chain 1: 9300 -9272.331 0.090 0.084
Chain 1: 9400 -8718.055 0.072 0.064
Chain 1: 9500 -11456.653 0.077 0.064
Chain 1: 9600 -11753.418 0.075 0.064
Chain 1: 9700 -8872.997 0.103 0.084
Chain 1: 9800 -11087.523 0.121 0.106
Chain 1: 9900 -8534.941 0.143 0.135
Chain 1: 10000 -8276.007 0.132 0.106
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001376 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.76 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -63462.180 1.000 1.000
Chain 1: 200 -18573.960 1.708 2.417
Chain 1: 300 -8919.542 1.500 1.082
Chain 1: 400 -8512.394 1.137 1.082
Chain 1: 500 -8505.488 0.910 1.000
Chain 1: 600 -8909.192 0.766 1.000
Chain 1: 700 -7948.379 0.673 0.121
Chain 1: 800 -7652.641 0.594 0.121
Chain 1: 900 -7767.014 0.530 0.048
Chain 1: 1000 -8016.635 0.480 0.048
Chain 1: 1100 -7596.637 0.385 0.048
Chain 1: 1200 -7942.416 0.148 0.045
Chain 1: 1300 -7714.198 0.043 0.044
Chain 1: 1400 -7861.378 0.040 0.039
Chain 1: 1500 -7471.880 0.045 0.044
Chain 1: 1600 -7691.244 0.043 0.039
Chain 1: 1700 -7549.879 0.033 0.031
Chain 1: 1800 -7661.566 0.031 0.030
Chain 1: 1900 -7540.334 0.031 0.030
Chain 1: 2000 -7627.737 0.029 0.029
Chain 1: 2100 -7489.277 0.025 0.019
Chain 1: 2200 -7705.473 0.024 0.019
Chain 1: 2300 -7471.064 0.024 0.019
Chain 1: 2400 -7498.301 0.022 0.019
Chain 1: 2500 -7538.730 0.018 0.018
Chain 1: 2600 -7462.870 0.016 0.016
Chain 1: 2700 -7532.325 0.015 0.015
Chain 1: 2800 -7470.311 0.014 0.011
Chain 1: 2900 -7355.784 0.014 0.011
Chain 1: 3000 -7498.711 0.015 0.016
Chain 1: 3100 -7467.016 0.013 0.010
Chain 1: 3200 -7672.207 0.013 0.010
Chain 1: 3300 -7386.809 0.014 0.010
Chain 1: 3400 -7623.870 0.017 0.016
Chain 1: 3500 -7370.102 0.020 0.019
Chain 1: 3600 -7436.043 0.020 0.019
Chain 1: 3700 -7386.616 0.019 0.019
Chain 1: 3800 -7387.307 0.019 0.019
Chain 1: 3900 -7344.335 0.018 0.019
Chain 1: 4000 -7337.086 0.016 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002629 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 26.29 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86699.910 1.000 1.000
Chain 1: 200 -13862.283 3.127 5.254
Chain 1: 300 -10135.439 2.207 1.000
Chain 1: 400 -11474.890 1.685 1.000
Chain 1: 500 -9025.201 1.402 0.368
Chain 1: 600 -9349.285 1.174 0.368
Chain 1: 700 -8738.162 1.016 0.271
Chain 1: 800 -8411.547 0.894 0.271
Chain 1: 900 -8432.660 0.795 0.117
Chain 1: 1000 -8735.428 0.719 0.117
Chain 1: 1100 -8865.830 0.621 0.070 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8527.447 0.099 0.040
Chain 1: 1300 -8802.557 0.065 0.039
Chain 1: 1400 -8715.796 0.055 0.035
Chain 1: 1500 -8636.208 0.029 0.035
Chain 1: 1600 -8744.362 0.026 0.031
Chain 1: 1700 -8805.708 0.020 0.015
Chain 1: 1800 -8365.026 0.021 0.015
Chain 1: 1900 -8469.865 0.022 0.015
Chain 1: 2000 -8451.397 0.019 0.012
Chain 1: 2100 -8574.951 0.019 0.012
Chain 1: 2200 -8369.592 0.018 0.012
Chain 1: 2300 -8462.642 0.016 0.012
Chain 1: 2400 -8529.669 0.015 0.012
Chain 1: 2500 -8478.538 0.015 0.012
Chain 1: 2600 -8490.266 0.014 0.011
Chain 1: 2700 -8399.204 0.014 0.011
Chain 1: 2800 -8348.214 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00329 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.9 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8410172.768 1.000 1.000
Chain 1: 200 -1584657.498 2.654 4.307
Chain 1: 300 -889977.897 2.029 1.000
Chain 1: 400 -457118.625 1.759 1.000
Chain 1: 500 -357573.771 1.463 0.947
Chain 1: 600 -232786.764 1.308 0.947
Chain 1: 700 -119352.080 1.257 0.947
Chain 1: 800 -86623.031 1.147 0.947
Chain 1: 900 -67026.318 1.052 0.781
Chain 1: 1000 -51871.135 0.976 0.781
Chain 1: 1100 -39386.792 0.908 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38573.558 0.479 0.378
Chain 1: 1300 -26561.835 0.446 0.378
Chain 1: 1400 -26286.433 0.353 0.317
Chain 1: 1500 -22880.953 0.340 0.317
Chain 1: 1600 -22100.363 0.290 0.292
Chain 1: 1700 -20977.373 0.200 0.292
Chain 1: 1800 -20922.670 0.163 0.149
Chain 1: 1900 -21249.332 0.135 0.054
Chain 1: 2000 -19761.178 0.113 0.054
Chain 1: 2100 -19999.687 0.083 0.035
Chain 1: 2200 -20226.097 0.082 0.035
Chain 1: 2300 -19843.192 0.038 0.019
Chain 1: 2400 -19615.119 0.039 0.019
Chain 1: 2500 -19416.911 0.025 0.015
Chain 1: 2600 -19046.868 0.023 0.015
Chain 1: 2700 -19003.783 0.018 0.012
Chain 1: 2800 -18720.301 0.019 0.015
Chain 1: 2900 -19001.729 0.019 0.015
Chain 1: 3000 -18987.983 0.012 0.012
Chain 1: 3100 -19073.015 0.011 0.012
Chain 1: 3200 -18763.446 0.011 0.015
Chain 1: 3300 -18968.368 0.011 0.012
Chain 1: 3400 -18442.727 0.012 0.015
Chain 1: 3500 -19055.368 0.014 0.015
Chain 1: 3600 -18361.027 0.016 0.015
Chain 1: 3700 -18748.551 0.018 0.016
Chain 1: 3800 -17706.613 0.023 0.021
Chain 1: 3900 -17702.665 0.021 0.021
Chain 1: 4000 -17820.027 0.022 0.021
Chain 1: 4100 -17733.669 0.022 0.021
Chain 1: 4200 -17549.563 0.021 0.021
Chain 1: 4300 -17688.253 0.021 0.021
Chain 1: 4400 -17644.792 0.018 0.010
Chain 1: 4500 -17547.221 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001311 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.11 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12245.194 1.000 1.000
Chain 1: 200 -9165.530 0.668 1.000
Chain 1: 300 -7885.490 0.499 0.336
Chain 1: 400 -8052.409 0.380 0.336
Chain 1: 500 -8009.044 0.305 0.162
Chain 1: 600 -7829.858 0.258 0.162
Chain 1: 700 -7751.301 0.222 0.023
Chain 1: 800 -7774.900 0.195 0.023
Chain 1: 900 -7804.190 0.174 0.021
Chain 1: 1000 -7807.429 0.156 0.021
Chain 1: 1100 -7847.512 0.057 0.010
Chain 1: 1200 -7751.866 0.025 0.010
Chain 1: 1300 -7748.379 0.008 0.005 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001378 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.78 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61032.511 1.000 1.000
Chain 1: 200 -17494.612 1.744 2.489
Chain 1: 300 -8742.313 1.497 1.001
Chain 1: 400 -8277.910 1.136 1.001
Chain 1: 500 -7948.130 0.917 1.000
Chain 1: 600 -8772.139 0.780 1.000
Chain 1: 700 -8332.287 0.676 0.094
Chain 1: 800 -8253.573 0.593 0.094
Chain 1: 900 -7948.557 0.531 0.056
Chain 1: 1000 -7772.601 0.480 0.056
Chain 1: 1100 -7562.864 0.383 0.053
Chain 1: 1200 -7895.899 0.139 0.042
Chain 1: 1300 -7543.085 0.043 0.042
Chain 1: 1400 -7767.222 0.040 0.041
Chain 1: 1500 -7538.569 0.039 0.038
Chain 1: 1600 -7729.439 0.032 0.030
Chain 1: 1700 -7495.160 0.030 0.030
Chain 1: 1800 -7583.561 0.030 0.030
Chain 1: 1900 -7596.409 0.027 0.029
Chain 1: 2000 -7623.188 0.025 0.029
Chain 1: 2100 -7552.060 0.023 0.029
Chain 1: 2200 -7679.810 0.020 0.025
Chain 1: 2300 -7585.300 0.017 0.017
Chain 1: 2400 -7616.484 0.015 0.012
Chain 1: 2500 -7559.983 0.012 0.012
Chain 1: 2600 -7525.321 0.010 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002595 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.95 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85537.720 1.000 1.000
Chain 1: 200 -13432.789 3.184 5.368
Chain 1: 300 -9767.941 2.248 1.000
Chain 1: 400 -10706.965 1.708 1.000
Chain 1: 500 -8747.455 1.411 0.375
Chain 1: 600 -8414.498 1.182 0.375
Chain 1: 700 -8332.457 1.015 0.224
Chain 1: 800 -8617.295 0.892 0.224
Chain 1: 900 -8646.480 0.793 0.088
Chain 1: 1000 -8340.560 0.718 0.088
Chain 1: 1100 -8613.120 0.621 0.040 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8221.191 0.089 0.040
Chain 1: 1300 -8445.560 0.054 0.037
Chain 1: 1400 -8461.902 0.045 0.033
Chain 1: 1500 -8309.371 0.025 0.032
Chain 1: 1600 -8424.567 0.022 0.027
Chain 1: 1700 -8499.552 0.022 0.027
Chain 1: 1800 -8074.383 0.024 0.027
Chain 1: 1900 -8176.454 0.025 0.027
Chain 1: 2000 -8151.046 0.022 0.018
Chain 1: 2100 -8277.470 0.020 0.015
Chain 1: 2200 -8077.855 0.018 0.015
Chain 1: 2300 -8171.418 0.016 0.014
Chain 1: 2400 -8239.782 0.017 0.014
Chain 1: 2500 -8185.981 0.016 0.012
Chain 1: 2600 -8187.929 0.014 0.011
Chain 1: 2700 -8104.412 0.015 0.011
Chain 1: 2800 -8063.500 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003225 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.25 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8380158.712 1.000 1.000
Chain 1: 200 -1579931.446 2.652 4.304
Chain 1: 300 -891751.445 2.025 1.000
Chain 1: 400 -458668.145 1.755 1.000
Chain 1: 500 -359175.962 1.459 0.944
Chain 1: 600 -233881.823 1.305 0.944
Chain 1: 700 -119650.329 1.255 0.944
Chain 1: 800 -86720.940 1.146 0.944
Chain 1: 900 -66972.866 1.051 0.772
Chain 1: 1000 -51702.778 0.976 0.772
Chain 1: 1100 -39114.526 0.908 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38286.828 0.480 0.380
Chain 1: 1300 -26178.367 0.449 0.380
Chain 1: 1400 -25891.968 0.355 0.322
Chain 1: 1500 -22462.012 0.343 0.322
Chain 1: 1600 -21673.443 0.293 0.295
Chain 1: 1700 -20539.559 0.203 0.295
Chain 1: 1800 -20482.136 0.165 0.153
Chain 1: 1900 -20808.357 0.138 0.055
Chain 1: 2000 -19315.365 0.116 0.055
Chain 1: 2100 -19553.880 0.085 0.036
Chain 1: 2200 -19781.031 0.084 0.036
Chain 1: 2300 -19397.654 0.039 0.020
Chain 1: 2400 -19169.643 0.040 0.020
Chain 1: 2500 -18971.846 0.025 0.016
Chain 1: 2600 -18601.578 0.024 0.016
Chain 1: 2700 -18558.521 0.018 0.012
Chain 1: 2800 -18275.305 0.020 0.015
Chain 1: 2900 -18556.813 0.020 0.015
Chain 1: 3000 -18542.935 0.012 0.012
Chain 1: 3100 -18627.890 0.011 0.012
Chain 1: 3200 -18318.429 0.012 0.015
Chain 1: 3300 -18523.325 0.011 0.012
Chain 1: 3400 -17997.974 0.013 0.015
Chain 1: 3500 -18610.257 0.015 0.015
Chain 1: 3600 -17916.596 0.017 0.015
Chain 1: 3700 -18303.637 0.019 0.017
Chain 1: 3800 -17262.722 0.023 0.021
Chain 1: 3900 -17258.930 0.022 0.021
Chain 1: 4000 -17376.198 0.022 0.021
Chain 1: 4100 -17289.847 0.022 0.021
Chain 1: 4200 -17106.068 0.022 0.021
Chain 1: 4300 -17244.460 0.021 0.021
Chain 1: 4400 -17201.180 0.019 0.011
Chain 1: 4500 -17103.777 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001284 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.84 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12737.958 1.000 1.000
Chain 1: 200 -9688.400 0.657 1.000
Chain 1: 300 -8410.786 0.489 0.315
Chain 1: 400 -8573.972 0.371 0.315
Chain 1: 500 -8444.043 0.300 0.152
Chain 1: 600 -8371.534 0.252 0.152
Chain 1: 700 -8283.379 0.217 0.019
Chain 1: 800 -8326.471 0.191 0.019
Chain 1: 900 -8446.337 0.171 0.015
Chain 1: 1000 -8348.490 0.155 0.015
Chain 1: 1100 -8358.883 0.055 0.014
Chain 1: 1200 -8295.319 0.025 0.012
Chain 1: 1300 -8257.818 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001384 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.84 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -46695.960 1.000 1.000
Chain 1: 200 -15885.488 1.470 1.940
Chain 1: 300 -8824.179 1.247 1.000
Chain 1: 400 -8194.045 0.954 1.000
Chain 1: 500 -9102.582 0.783 0.800
Chain 1: 600 -8316.074 0.669 0.800
Chain 1: 700 -7828.303 0.582 0.100
Chain 1: 800 -8147.484 0.514 0.100
Chain 1: 900 -8123.133 0.457 0.095
Chain 1: 1000 -7800.543 0.416 0.095
Chain 1: 1100 -7764.885 0.316 0.077
Chain 1: 1200 -7713.630 0.123 0.062
Chain 1: 1300 -8090.492 0.047 0.047
Chain 1: 1400 -8027.502 0.041 0.041
Chain 1: 1500 -7535.017 0.037 0.041
Chain 1: 1600 -7693.799 0.030 0.039
Chain 1: 1700 -7571.267 0.025 0.021
Chain 1: 1800 -7587.450 0.021 0.016
Chain 1: 1900 -7582.022 0.021 0.016
Chain 1: 2000 -7624.485 0.018 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002666 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 26.66 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86784.015 1.000 1.000
Chain 1: 200 -13837.750 3.136 5.272
Chain 1: 300 -10228.406 2.208 1.000
Chain 1: 400 -10960.776 1.673 1.000
Chain 1: 500 -9172.159 1.377 0.353
Chain 1: 600 -8829.347 1.154 0.353
Chain 1: 700 -8804.460 0.990 0.195
Chain 1: 800 -9290.868 0.873 0.195
Chain 1: 900 -8969.071 0.780 0.067
Chain 1: 1000 -8910.133 0.702 0.067
Chain 1: 1100 -9078.621 0.604 0.052 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8629.679 0.082 0.052
Chain 1: 1300 -8943.302 0.050 0.039
Chain 1: 1400 -8938.969 0.044 0.036
Chain 1: 1500 -8813.169 0.026 0.035
Chain 1: 1600 -8920.811 0.023 0.019
Chain 1: 1700 -9006.991 0.024 0.019
Chain 1: 1800 -8602.618 0.023 0.019
Chain 1: 1900 -8700.429 0.021 0.014
Chain 1: 2000 -8672.261 0.020 0.014
Chain 1: 2100 -8792.188 0.020 0.014
Chain 1: 2200 -8584.148 0.017 0.014
Chain 1: 2300 -8735.696 0.015 0.014
Chain 1: 2400 -8742.417 0.015 0.014
Chain 1: 2500 -8714.228 0.014 0.012
Chain 1: 2600 -8713.482 0.013 0.011
Chain 1: 2700 -8624.891 0.013 0.011
Chain 1: 2800 -8591.331 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003118 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.18 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8371106.381 1.000 1.000
Chain 1: 200 -1576583.652 2.655 4.310
Chain 1: 300 -889244.269 2.028 1.000
Chain 1: 400 -456848.083 1.757 1.000
Chain 1: 500 -357888.777 1.461 0.946
Chain 1: 600 -233259.668 1.307 0.946
Chain 1: 700 -119601.881 1.256 0.946
Chain 1: 800 -86801.742 1.146 0.946
Chain 1: 900 -67148.261 1.051 0.773
Chain 1: 1000 -51936.616 0.975 0.773
Chain 1: 1100 -39400.036 0.907 0.534 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38578.622 0.478 0.378
Chain 1: 1300 -26520.709 0.447 0.378
Chain 1: 1400 -26238.407 0.353 0.318
Chain 1: 1500 -22821.053 0.340 0.318
Chain 1: 1600 -22036.112 0.290 0.293
Chain 1: 1700 -20907.976 0.201 0.293
Chain 1: 1800 -20851.846 0.163 0.150
Chain 1: 1900 -21177.851 0.136 0.054
Chain 1: 2000 -19688.383 0.114 0.054
Chain 1: 2100 -19926.835 0.083 0.036
Chain 1: 2200 -20153.220 0.082 0.036
Chain 1: 2300 -19770.537 0.039 0.019
Chain 1: 2400 -19542.685 0.039 0.019
Chain 1: 2500 -19344.699 0.025 0.015
Chain 1: 2600 -18975.052 0.023 0.015
Chain 1: 2700 -18932.147 0.018 0.012
Chain 1: 2800 -18649.053 0.019 0.015
Chain 1: 2900 -18930.294 0.019 0.015
Chain 1: 3000 -18916.472 0.012 0.012
Chain 1: 3100 -19001.394 0.011 0.012
Chain 1: 3200 -18692.217 0.011 0.015
Chain 1: 3300 -18896.882 0.011 0.012
Chain 1: 3400 -18372.033 0.012 0.015
Chain 1: 3500 -18983.559 0.015 0.015
Chain 1: 3600 -18290.780 0.016 0.015
Chain 1: 3700 -18677.184 0.018 0.017
Chain 1: 3800 -17637.669 0.023 0.021
Chain 1: 3900 -17633.863 0.021 0.021
Chain 1: 4000 -17751.146 0.022 0.021
Chain 1: 4100 -17664.892 0.022 0.021
Chain 1: 4200 -17481.394 0.021 0.021
Chain 1: 4300 -17619.632 0.021 0.021
Chain 1: 4400 -17576.616 0.018 0.010
Chain 1: 4500 -17479.187 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001135 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 11.35 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -50444.865 1.000 1.000
Chain 1: 200 -14988.381 1.683 2.366
Chain 1: 300 -19979.418 1.205 1.000
Chain 1: 400 -21882.507 0.926 1.000
Chain 1: 500 -16109.538 0.812 0.358
Chain 1: 600 -20654.035 0.713 0.358
Chain 1: 700 -15743.844 0.656 0.312
Chain 1: 800 -16195.469 0.578 0.312
Chain 1: 900 -16084.293 0.514 0.250
Chain 1: 1000 -25521.634 0.500 0.312
Chain 1: 1100 -11800.046 0.516 0.312 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -12453.558 0.285 0.250
Chain 1: 1300 -13620.155 0.268 0.220
Chain 1: 1400 -13352.019 0.262 0.220
Chain 1: 1500 -12886.658 0.229 0.086
Chain 1: 1600 -10768.039 0.227 0.086
Chain 1: 1700 -17477.346 0.234 0.086
Chain 1: 1800 -12624.742 0.270 0.197
Chain 1: 1900 -12912.069 0.271 0.197
Chain 1: 2000 -10353.004 0.259 0.197
Chain 1: 2100 -11273.814 0.151 0.086
Chain 1: 2200 -12258.886 0.154 0.086
Chain 1: 2300 -17460.574 0.175 0.197
Chain 1: 2400 -10346.637 0.242 0.247
Chain 1: 2500 -10756.220 0.242 0.247
Chain 1: 2600 -13582.414 0.243 0.247
Chain 1: 2700 -14462.660 0.211 0.208
Chain 1: 2800 -10357.527 0.212 0.208
Chain 1: 2900 -10228.915 0.211 0.208
Chain 1: 3000 -10389.973 0.188 0.082
Chain 1: 3100 -11394.775 0.189 0.088
Chain 1: 3200 -9923.755 0.195 0.148
Chain 1: 3300 -11536.681 0.180 0.140
Chain 1: 3400 -11027.707 0.115 0.088
Chain 1: 3500 -10514.942 0.116 0.088
Chain 1: 3600 -14887.296 0.125 0.088
Chain 1: 3700 -16140.087 0.127 0.088
Chain 1: 3800 -10430.252 0.142 0.088
Chain 1: 3900 -9954.708 0.145 0.088
Chain 1: 4000 -19854.604 0.194 0.140
Chain 1: 4100 -10506.878 0.274 0.148
Chain 1: 4200 -9920.338 0.265 0.140
Chain 1: 4300 -11357.598 0.264 0.127
Chain 1: 4400 -10392.811 0.268 0.127
Chain 1: 4500 -9793.512 0.269 0.127
Chain 1: 4600 -13986.201 0.270 0.127
Chain 1: 4700 -14714.332 0.267 0.127
Chain 1: 4800 -14693.562 0.213 0.093
Chain 1: 4900 -10797.450 0.244 0.127
Chain 1: 5000 -17415.694 0.232 0.127
Chain 1: 5100 -9635.113 0.224 0.127
Chain 1: 5200 -10075.622 0.222 0.127
Chain 1: 5300 -9280.858 0.218 0.093
Chain 1: 5400 -11770.054 0.230 0.211
Chain 1: 5500 -10876.170 0.232 0.211
Chain 1: 5600 -10162.994 0.209 0.086
Chain 1: 5700 -10475.352 0.207 0.086
Chain 1: 5800 -9457.599 0.218 0.108
Chain 1: 5900 -12675.210 0.207 0.108
Chain 1: 6000 -9301.780 0.205 0.108
Chain 1: 6100 -16478.928 0.168 0.108
Chain 1: 6200 -9518.783 0.237 0.211
Chain 1: 6300 -13761.320 0.259 0.254
Chain 1: 6400 -14867.341 0.246 0.254
Chain 1: 6500 -10312.138 0.282 0.308
Chain 1: 6600 -9461.373 0.284 0.308
Chain 1: 6700 -9276.303 0.283 0.308
Chain 1: 6800 -9282.628 0.272 0.308
Chain 1: 6900 -11147.289 0.263 0.308
Chain 1: 7000 -9609.944 0.243 0.167
Chain 1: 7100 -9103.419 0.205 0.160
Chain 1: 7200 -9991.926 0.141 0.090
Chain 1: 7300 -9652.485 0.113 0.089
Chain 1: 7400 -12642.458 0.130 0.090
Chain 1: 7500 -12886.126 0.087 0.089
Chain 1: 7600 -14717.456 0.091 0.089
Chain 1: 7700 -9727.605 0.140 0.124
Chain 1: 7800 -16700.362 0.182 0.160
Chain 1: 7900 -11404.872 0.211 0.160
Chain 1: 8000 -9489.066 0.216 0.202
Chain 1: 8100 -12256.772 0.233 0.226
Chain 1: 8200 -10078.909 0.245 0.226
Chain 1: 8300 -9723.453 0.245 0.226
Chain 1: 8400 -13820.305 0.251 0.226
Chain 1: 8500 -10931.510 0.276 0.264
Chain 1: 8600 -8935.888 0.286 0.264
Chain 1: 8700 -10085.652 0.246 0.226
Chain 1: 8800 -9813.650 0.207 0.223
Chain 1: 8900 -14313.934 0.192 0.223
Chain 1: 9000 -9045.462 0.230 0.226
Chain 1: 9100 -9156.800 0.209 0.223
Chain 1: 9200 -14144.159 0.222 0.264
Chain 1: 9300 -11901.196 0.238 0.264
Chain 1: 9400 -9322.093 0.236 0.264
Chain 1: 9500 -10303.973 0.219 0.223
Chain 1: 9600 -9191.501 0.208 0.188
Chain 1: 9700 -9144.665 0.198 0.188
Chain 1: 9800 -12710.475 0.223 0.277
Chain 1: 9900 -10909.438 0.208 0.188
Chain 1: 10000 -9210.830 0.168 0.184
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001378 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.78 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -47316.161 1.000 1.000
Chain 1: 200 -16886.350 1.401 1.802
Chain 1: 300 -9516.029 1.192 1.000
Chain 1: 400 -8388.512 0.928 1.000
Chain 1: 500 -9127.470 0.758 0.775
Chain 1: 600 -9424.306 0.637 0.775
Chain 1: 700 -8068.040 0.570 0.168
Chain 1: 800 -8624.520 0.507 0.168
Chain 1: 900 -7758.741 0.463 0.134
Chain 1: 1000 -7876.270 0.418 0.134
Chain 1: 1100 -7814.527 0.319 0.112
Chain 1: 1200 -7674.664 0.141 0.081
Chain 1: 1300 -7731.100 0.064 0.065
Chain 1: 1400 -7881.056 0.052 0.031
Chain 1: 1500 -7792.701 0.045 0.019
Chain 1: 1600 -7819.444 0.043 0.018
Chain 1: 1700 -7652.598 0.028 0.018
Chain 1: 1800 -7673.020 0.022 0.015
Chain 1: 1900 -7786.481 0.012 0.015
Chain 1: 2000 -7792.345 0.011 0.011
Chain 1: 2100 -7652.884 0.012 0.015
Chain 1: 2200 -7884.591 0.013 0.015
Chain 1: 2300 -7704.186 0.014 0.018
Chain 1: 2400 -7692.124 0.013 0.015
Chain 1: 2500 -7668.011 0.012 0.015
Chain 1: 2600 -7595.463 0.013 0.015
Chain 1: 2700 -7562.691 0.011 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002511 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.11 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87900.975 1.000 1.000
Chain 1: 200 -14884.908 2.953 4.905
Chain 1: 300 -10941.576 2.089 1.000
Chain 1: 400 -13282.447 1.611 1.000
Chain 1: 500 -9318.855 1.373 0.425
Chain 1: 600 -9157.878 1.147 0.425
Chain 1: 700 -9200.771 0.984 0.360
Chain 1: 800 -9419.799 0.864 0.360
Chain 1: 900 -9712.657 0.771 0.176
Chain 1: 1000 -9962.704 0.697 0.176
Chain 1: 1100 -9573.731 0.601 0.041 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8976.486 0.117 0.041
Chain 1: 1300 -9645.716 0.088 0.041
Chain 1: 1400 -9214.367 0.075 0.041
Chain 1: 1500 -9216.402 0.032 0.030
Chain 1: 1600 -9134.235 0.032 0.030
Chain 1: 1700 -9000.455 0.033 0.030
Chain 1: 1800 -9042.290 0.031 0.030
Chain 1: 1900 -9075.445 0.028 0.025
Chain 1: 2000 -9250.436 0.027 0.019
Chain 1: 2100 -9033.070 0.026 0.019
Chain 1: 2200 -8986.694 0.020 0.015
Chain 1: 2300 -9224.880 0.015 0.015
Chain 1: 2400 -8943.139 0.014 0.015
Chain 1: 2500 -9015.379 0.015 0.015
Chain 1: 2600 -8920.153 0.015 0.015
Chain 1: 2700 -8932.810 0.013 0.011
Chain 1: 2800 -8799.258 0.014 0.015
Chain 1: 2900 -8985.051 0.016 0.019
Chain 1: 3000 -8883.612 0.015 0.015
Chain 1: 3100 -8988.976 0.014 0.012
Chain 1: 3200 -8850.925 0.015 0.015
Chain 1: 3300 -9112.430 0.015 0.015
Chain 1: 3400 -9076.551 0.013 0.012
Chain 1: 3500 -8947.090 0.013 0.014
Chain 1: 3600 -9034.719 0.013 0.014
Chain 1: 3700 -8882.737 0.015 0.015
Chain 1: 3800 -8786.319 0.014 0.014
Chain 1: 3900 -9030.093 0.015 0.014
Chain 1: 4000 -9046.325 0.014 0.014
Chain 1: 4100 -8818.971 0.016 0.016
Chain 1: 4200 -8803.745 0.014 0.014
Chain 1: 4300 -8804.584 0.011 0.011
Chain 1: 4400 -8757.753 0.011 0.011
Chain 1: 4500 -8902.799 0.012 0.011
Chain 1: 4600 -8931.601 0.011 0.011
Chain 1: 4700 -9055.415 0.011 0.011
Chain 1: 4800 -8868.473 0.012 0.014
Chain 1: 4900 -8901.323 0.009 0.005 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003428 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.28 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8401033.193 1.000 1.000
Chain 1: 200 -1585399.050 2.650 4.299
Chain 1: 300 -893057.617 2.025 1.000
Chain 1: 400 -459790.275 1.754 1.000
Chain 1: 500 -359779.029 1.459 0.942
Chain 1: 600 -234676.200 1.305 0.942
Chain 1: 700 -120782.916 1.253 0.942
Chain 1: 800 -87954.178 1.143 0.942
Chain 1: 900 -68290.257 1.048 0.775
Chain 1: 1000 -53095.609 0.972 0.775
Chain 1: 1100 -40564.466 0.903 0.533 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39755.808 0.475 0.373
Chain 1: 1300 -27679.929 0.441 0.373
Chain 1: 1400 -27403.933 0.348 0.309
Chain 1: 1500 -23980.492 0.334 0.309
Chain 1: 1600 -23196.105 0.284 0.288
Chain 1: 1700 -22064.367 0.195 0.286
Chain 1: 1800 -22008.334 0.158 0.143
Chain 1: 1900 -22335.876 0.131 0.051
Chain 1: 2000 -20841.356 0.109 0.051
Chain 1: 2100 -21080.318 0.079 0.034
Chain 1: 2200 -21308.050 0.079 0.034
Chain 1: 2300 -20923.731 0.037 0.018
Chain 1: 2400 -20695.201 0.037 0.018
Chain 1: 2500 -20497.207 0.024 0.015
Chain 1: 2600 -20125.880 0.022 0.015
Chain 1: 2700 -20082.461 0.017 0.011
Chain 1: 2800 -19798.579 0.018 0.014
Chain 1: 2900 -20080.609 0.018 0.014
Chain 1: 3000 -20066.754 0.011 0.011
Chain 1: 3100 -20151.903 0.010 0.011
Chain 1: 3200 -19841.622 0.011 0.014
Chain 1: 3300 -20047.131 0.010 0.011
Chain 1: 3400 -19520.251 0.012 0.014
Chain 1: 3500 -20134.813 0.014 0.014
Chain 1: 3600 -19438.056 0.015 0.014
Chain 1: 3700 -19827.366 0.017 0.016
Chain 1: 3800 -18781.682 0.021 0.020
Chain 1: 3900 -18777.668 0.020 0.020
Chain 1: 4000 -18895.021 0.021 0.020
Chain 1: 4100 -18808.436 0.021 0.020
Chain 1: 4200 -18623.532 0.020 0.020
Chain 1: 4300 -18762.761 0.020 0.020
Chain 1: 4400 -18718.621 0.017 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001248 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.48 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12381.320 1.000 1.000
Chain 1: 200 -9303.125 0.665 1.000
Chain 1: 300 -7995.752 0.498 0.331
Chain 1: 400 -8186.505 0.379 0.331
Chain 1: 500 -8052.047 0.307 0.164
Chain 1: 600 -7878.909 0.259 0.164
Chain 1: 700 -7824.114 0.223 0.023
Chain 1: 800 -7835.833 0.196 0.023
Chain 1: 900 -7885.211 0.175 0.022
Chain 1: 1000 -7925.388 0.158 0.022
Chain 1: 1100 -8005.889 0.059 0.017
Chain 1: 1200 -7860.359 0.027 0.017
Chain 1: 1300 -7783.007 0.012 0.010
Chain 1: 1400 -7804.532 0.010 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001372 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.72 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -45898.751 1.000 1.000
Chain 1: 200 -15585.800 1.472 1.945
Chain 1: 300 -8707.367 1.245 1.000
Chain 1: 400 -8492.461 0.940 1.000
Chain 1: 500 -8355.831 0.755 0.790
Chain 1: 600 -8262.517 0.631 0.790
Chain 1: 700 -7798.071 0.550 0.060
Chain 1: 800 -7938.521 0.483 0.060
Chain 1: 900 -7862.583 0.431 0.025
Chain 1: 1000 -7724.882 0.389 0.025
Chain 1: 1100 -7735.839 0.289 0.018
Chain 1: 1200 -7675.173 0.096 0.018
Chain 1: 1300 -7836.461 0.019 0.018
Chain 1: 1400 -7827.782 0.016 0.016
Chain 1: 1500 -7598.745 0.018 0.018
Chain 1: 1600 -7714.965 0.018 0.018
Chain 1: 1700 -7542.488 0.014 0.018
Chain 1: 1800 -7598.099 0.013 0.015
Chain 1: 1900 -7601.725 0.012 0.015
Chain 1: 2000 -7642.007 0.011 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00256 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.6 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86430.272 1.000 1.000
Chain 1: 200 -13544.729 3.191 5.381
Chain 1: 300 -9862.578 2.251 1.000
Chain 1: 400 -10753.461 1.709 1.000
Chain 1: 500 -8863.486 1.410 0.373
Chain 1: 600 -8519.715 1.182 0.373
Chain 1: 700 -8229.149 1.018 0.213
Chain 1: 800 -8747.405 0.898 0.213
Chain 1: 900 -8579.332 0.801 0.083
Chain 1: 1000 -8593.304 0.721 0.083
Chain 1: 1100 -8646.189 0.621 0.059 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8159.217 0.089 0.059
Chain 1: 1300 -8463.739 0.055 0.040
Chain 1: 1400 -8515.533 0.048 0.036
Chain 1: 1500 -8395.299 0.028 0.035
Chain 1: 1600 -8507.395 0.025 0.020
Chain 1: 1700 -8575.208 0.022 0.014
Chain 1: 1800 -8144.288 0.022 0.014
Chain 1: 1900 -8248.286 0.021 0.013
Chain 1: 2000 -8223.423 0.021 0.013
Chain 1: 2100 -8359.056 0.022 0.014
Chain 1: 2200 -8152.789 0.019 0.014
Chain 1: 2300 -8248.936 0.016 0.013
Chain 1: 2400 -8312.369 0.016 0.013
Chain 1: 2500 -8256.740 0.016 0.013
Chain 1: 2600 -8260.882 0.014 0.012
Chain 1: 2700 -8176.203 0.015 0.012
Chain 1: 2800 -8132.852 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003381 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.81 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8378214.347 1.000 1.000
Chain 1: 200 -1581036.305 2.650 4.299
Chain 1: 300 -891083.007 2.024 1.000
Chain 1: 400 -457910.712 1.755 1.000
Chain 1: 500 -358587.722 1.459 0.946
Chain 1: 600 -233416.042 1.305 0.946
Chain 1: 700 -119505.911 1.255 0.946
Chain 1: 800 -86652.206 1.146 0.946
Chain 1: 900 -66961.555 1.051 0.774
Chain 1: 1000 -51728.411 0.975 0.774
Chain 1: 1100 -39174.782 0.907 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38352.393 0.480 0.379
Chain 1: 1300 -26275.792 0.448 0.379
Chain 1: 1400 -25992.861 0.355 0.320
Chain 1: 1500 -22571.151 0.342 0.320
Chain 1: 1600 -21785.239 0.292 0.294
Chain 1: 1700 -20654.948 0.202 0.294
Chain 1: 1800 -20598.462 0.165 0.152
Chain 1: 1900 -20924.839 0.137 0.055
Chain 1: 2000 -19433.464 0.115 0.055
Chain 1: 2100 -19671.992 0.084 0.036
Chain 1: 2200 -19898.878 0.083 0.036
Chain 1: 2300 -19515.669 0.039 0.020
Chain 1: 2400 -19287.640 0.039 0.020
Chain 1: 2500 -19089.704 0.025 0.016
Chain 1: 2600 -18719.506 0.023 0.016
Chain 1: 2700 -18676.463 0.018 0.012
Chain 1: 2800 -18393.153 0.020 0.015
Chain 1: 2900 -18674.664 0.019 0.015
Chain 1: 3000 -18660.783 0.012 0.012
Chain 1: 3100 -18745.754 0.011 0.012
Chain 1: 3200 -18436.283 0.012 0.015
Chain 1: 3300 -18641.195 0.011 0.012
Chain 1: 3400 -18115.756 0.012 0.015
Chain 1: 3500 -18728.136 0.015 0.015
Chain 1: 3600 -18034.302 0.017 0.015
Chain 1: 3700 -18421.459 0.018 0.017
Chain 1: 3800 -17380.253 0.023 0.021
Chain 1: 3900 -17376.418 0.021 0.021
Chain 1: 4000 -17493.717 0.022 0.021
Chain 1: 4100 -17407.348 0.022 0.021
Chain 1: 4200 -17223.476 0.021 0.021
Chain 1: 4300 -17361.956 0.021 0.021
Chain 1: 4400 -17318.630 0.019 0.011
Chain 1: 4500 -17221.159 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001337 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.37 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12082.677 1.000 1.000
Chain 1: 200 -9009.930 0.671 1.000
Chain 1: 300 -7818.843 0.498 0.341
Chain 1: 400 -8017.684 0.380 0.341
Chain 1: 500 -7953.215 0.305 0.152
Chain 1: 600 -7919.437 0.255 0.152
Chain 1: 700 -7701.442 0.223 0.028
Chain 1: 800 -7656.855 0.196 0.028
Chain 1: 900 -7601.724 0.175 0.025
Chain 1: 1000 -7704.100 0.159 0.025
Chain 1: 1100 -7845.026 0.060 0.018
Chain 1: 1200 -7712.302 0.028 0.017
Chain 1: 1300 -7655.054 0.013 0.013
Chain 1: 1400 -7681.788 0.011 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00142 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.2 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56460.467 1.000 1.000
Chain 1: 200 -17103.001 1.651 2.301
Chain 1: 300 -8574.926 1.432 1.000
Chain 1: 400 -8257.359 1.084 1.000
Chain 1: 500 -8045.082 0.872 0.995
Chain 1: 600 -8561.227 0.737 0.995
Chain 1: 700 -7797.524 0.646 0.098
Chain 1: 800 -8190.937 0.571 0.098
Chain 1: 900 -7972.323 0.510 0.060
Chain 1: 1000 -7640.417 0.464 0.060
Chain 1: 1100 -7694.040 0.364 0.048
Chain 1: 1200 -7773.334 0.135 0.043
Chain 1: 1300 -7685.324 0.037 0.038
Chain 1: 1400 -7628.356 0.034 0.027
Chain 1: 1500 -7564.390 0.032 0.027
Chain 1: 1600 -7494.669 0.027 0.011
Chain 1: 1700 -7488.670 0.017 0.010
Chain 1: 1800 -7557.456 0.013 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003656 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 36.56 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86167.441 1.000 1.000
Chain 1: 200 -13178.924 3.269 5.538
Chain 1: 300 -9598.782 2.304 1.000
Chain 1: 400 -10388.394 1.747 1.000
Chain 1: 500 -8519.268 1.441 0.373
Chain 1: 600 -8231.526 1.207 0.373
Chain 1: 700 -8257.236 1.035 0.219
Chain 1: 800 -8532.080 0.910 0.219
Chain 1: 900 -8423.943 0.810 0.076
Chain 1: 1000 -8214.824 0.732 0.076
Chain 1: 1100 -8443.380 0.634 0.035 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8075.247 0.085 0.035
Chain 1: 1300 -8138.128 0.048 0.032
Chain 1: 1400 -8246.972 0.042 0.027
Chain 1: 1500 -8174.186 0.021 0.025
Chain 1: 1600 -8171.516 0.018 0.013
Chain 1: 1700 -8091.657 0.018 0.013
Chain 1: 1800 -7981.152 0.016 0.013
Chain 1: 1900 -8101.590 0.017 0.014
Chain 1: 2000 -8062.057 0.015 0.013
Chain 1: 2100 -8186.409 0.013 0.013
Chain 1: 2200 -7963.523 0.012 0.013
Chain 1: 2300 -8123.571 0.013 0.014
Chain 1: 2400 -8007.777 0.013 0.014
Chain 1: 2500 -8069.918 0.013 0.014
Chain 1: 2600 -8089.982 0.013 0.014
Chain 1: 2700 -8009.559 0.013 0.014
Chain 1: 2800 -7984.917 0.012 0.014
Chain 1: 2900 -8039.492 0.011 0.010
Chain 1: 3000 -7924.992 0.012 0.014
Chain 1: 3100 -8061.735 0.012 0.014
Chain 1: 3200 -7942.343 0.011 0.014
Chain 1: 3300 -7963.187 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002604 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 26.04 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8402285.469 1.000 1.000
Chain 1: 200 -1588182.626 2.645 4.291
Chain 1: 300 -891721.084 2.024 1.000
Chain 1: 400 -457423.094 1.755 1.000
Chain 1: 500 -357340.512 1.460 0.949
Chain 1: 600 -232237.271 1.307 0.949
Chain 1: 700 -118687.396 1.257 0.949
Chain 1: 800 -85921.946 1.147 0.949
Chain 1: 900 -66318.270 1.053 0.781
Chain 1: 1000 -51152.208 0.977 0.781
Chain 1: 1100 -38659.578 0.909 0.539 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37842.264 0.482 0.381
Chain 1: 1300 -25836.485 0.451 0.381
Chain 1: 1400 -25558.322 0.357 0.323
Chain 1: 1500 -22154.849 0.344 0.323
Chain 1: 1600 -21373.665 0.294 0.296
Chain 1: 1700 -20252.411 0.204 0.296
Chain 1: 1800 -20197.727 0.166 0.154
Chain 1: 1900 -20523.523 0.138 0.055
Chain 1: 2000 -19037.832 0.116 0.055
Chain 1: 2100 -19276.109 0.085 0.037
Chain 1: 2200 -19501.795 0.084 0.037
Chain 1: 2300 -19119.783 0.040 0.020
Chain 1: 2400 -18892.050 0.040 0.020
Chain 1: 2500 -18693.829 0.026 0.016
Chain 1: 2600 -18324.564 0.024 0.016
Chain 1: 2700 -18281.799 0.019 0.012
Chain 1: 2800 -17998.600 0.020 0.016
Chain 1: 2900 -18279.755 0.020 0.015
Chain 1: 3000 -18266.006 0.012 0.012
Chain 1: 3100 -18350.867 0.011 0.012
Chain 1: 3200 -18041.866 0.012 0.015
Chain 1: 3300 -18246.420 0.011 0.012
Chain 1: 3400 -17721.708 0.013 0.015
Chain 1: 3500 -18332.882 0.015 0.016
Chain 1: 3600 -17640.603 0.017 0.016
Chain 1: 3700 -18026.542 0.019 0.017
Chain 1: 3800 -16987.699 0.023 0.021
Chain 1: 3900 -16983.883 0.022 0.021
Chain 1: 4000 -17101.223 0.022 0.021
Chain 1: 4100 -17014.931 0.023 0.021
Chain 1: 4200 -16831.610 0.022 0.021
Chain 1: 4300 -16969.742 0.022 0.021
Chain 1: 4400 -16926.837 0.019 0.011
Chain 1: 4500 -16829.412 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00131 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.1 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49488.964 1.000 1.000
Chain 1: 200 -20214.651 1.224 1.448
Chain 1: 300 -24098.287 0.870 1.000
Chain 1: 400 -22951.851 0.665 1.000
Chain 1: 500 -13390.652 0.675 0.714
Chain 1: 600 -35648.257 0.666 0.714
Chain 1: 700 -14071.805 0.790 0.714
Chain 1: 800 -14425.204 0.694 0.714
Chain 1: 900 -12917.700 0.630 0.624
Chain 1: 1000 -15490.824 0.584 0.624
Chain 1: 1100 -15513.975 0.484 0.166
Chain 1: 1200 -12771.056 0.361 0.166
Chain 1: 1300 -11283.095 0.358 0.166
Chain 1: 1400 -15205.338 0.379 0.215
Chain 1: 1500 -10643.967 0.350 0.215
Chain 1: 1600 -11659.693 0.296 0.166
Chain 1: 1700 -21208.831 0.188 0.166
Chain 1: 1800 -13683.361 0.240 0.215
Chain 1: 1900 -10666.471 0.257 0.258
Chain 1: 2000 -11393.990 0.247 0.258
Chain 1: 2100 -10532.090 0.255 0.258
Chain 1: 2200 -11758.202 0.244 0.258
Chain 1: 2300 -9950.604 0.249 0.258
Chain 1: 2400 -10494.510 0.228 0.182
Chain 1: 2500 -10039.124 0.190 0.104
Chain 1: 2600 -10137.421 0.182 0.104
Chain 1: 2700 -9410.079 0.145 0.082
Chain 1: 2800 -11272.858 0.106 0.082
Chain 1: 2900 -10318.765 0.087 0.082
Chain 1: 3000 -9755.998 0.087 0.082
Chain 1: 3100 -9417.822 0.082 0.077
Chain 1: 3200 -9891.406 0.077 0.058
Chain 1: 3300 -9586.272 0.062 0.052
Chain 1: 3400 -9589.792 0.056 0.048
Chain 1: 3500 -9831.240 0.054 0.048
Chain 1: 3600 -12139.815 0.072 0.058
Chain 1: 3700 -10213.395 0.083 0.058
Chain 1: 3800 -10350.912 0.068 0.048
Chain 1: 3900 -14255.328 0.086 0.048
Chain 1: 4000 -9498.022 0.131 0.048
Chain 1: 4100 -9612.554 0.128 0.048
Chain 1: 4200 -9445.473 0.125 0.032
Chain 1: 4300 -10423.585 0.132 0.094
Chain 1: 4400 -9137.588 0.146 0.141
Chain 1: 4500 -9673.932 0.149 0.141
Chain 1: 4600 -9975.308 0.133 0.094
Chain 1: 4700 -9507.390 0.119 0.055
Chain 1: 4800 -9493.348 0.118 0.055
Chain 1: 4900 -11441.178 0.107 0.055
Chain 1: 5000 -13853.539 0.074 0.055
Chain 1: 5100 -9683.687 0.116 0.094
Chain 1: 5200 -9441.324 0.117 0.094
Chain 1: 5300 -9704.487 0.110 0.055
Chain 1: 5400 -16035.103 0.136 0.055
Chain 1: 5500 -12840.429 0.155 0.170
Chain 1: 5600 -14967.915 0.166 0.170
Chain 1: 5700 -9343.034 0.222 0.174
Chain 1: 5800 -14717.200 0.258 0.249
Chain 1: 5900 -17429.632 0.257 0.249
Chain 1: 6000 -9630.491 0.320 0.365
Chain 1: 6100 -12412.367 0.300 0.249
Chain 1: 6200 -12042.471 0.300 0.249
Chain 1: 6300 -10051.957 0.317 0.249
Chain 1: 6400 -9032.940 0.289 0.224
Chain 1: 6500 -10210.675 0.276 0.198
Chain 1: 6600 -9169.563 0.273 0.198
Chain 1: 6700 -9157.613 0.213 0.156
Chain 1: 6800 -12778.073 0.204 0.156
Chain 1: 6900 -9210.805 0.228 0.198
Chain 1: 7000 -12896.347 0.175 0.198
Chain 1: 7100 -10500.429 0.176 0.198
Chain 1: 7200 -8984.735 0.189 0.198
Chain 1: 7300 -12153.784 0.196 0.228
Chain 1: 7400 -8824.845 0.222 0.261
Chain 1: 7500 -10394.168 0.226 0.261
Chain 1: 7600 -9217.300 0.227 0.261
Chain 1: 7700 -9524.361 0.230 0.261
Chain 1: 7800 -10510.875 0.211 0.228
Chain 1: 7900 -9543.270 0.183 0.169
Chain 1: 8000 -9250.863 0.157 0.151
Chain 1: 8100 -8803.992 0.140 0.128
Chain 1: 8200 -11464.868 0.146 0.128
Chain 1: 8300 -8809.254 0.150 0.128
Chain 1: 8400 -9988.524 0.124 0.118
Chain 1: 8500 -8762.515 0.123 0.118
Chain 1: 8600 -8704.010 0.111 0.101
Chain 1: 8700 -9134.068 0.112 0.101
Chain 1: 8800 -8656.599 0.108 0.101
Chain 1: 8900 -9845.873 0.110 0.118
Chain 1: 9000 -9796.503 0.108 0.118
Chain 1: 9100 -11607.364 0.118 0.121
Chain 1: 9200 -9136.147 0.122 0.121
Chain 1: 9300 -10315.164 0.103 0.118
Chain 1: 9400 -8688.569 0.110 0.121
Chain 1: 9500 -11226.819 0.119 0.121
Chain 1: 9600 -8884.167 0.145 0.156
Chain 1: 9700 -13452.202 0.174 0.187
Chain 1: 9800 -9506.619 0.210 0.226
Chain 1: 9900 -11037.145 0.212 0.226
Chain 1: 10000 -10939.543 0.212 0.226
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001368 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.68 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57873.154 1.000 1.000
Chain 1: 200 -18129.796 1.596 2.192
Chain 1: 300 -9060.965 1.398 1.001
Chain 1: 400 -8223.357 1.074 1.001
Chain 1: 500 -8639.347 0.869 1.000
Chain 1: 600 -8052.382 0.736 1.000
Chain 1: 700 -8498.527 0.638 0.102
Chain 1: 800 -8405.059 0.560 0.102
Chain 1: 900 -8162.946 0.501 0.073
Chain 1: 1000 -7728.952 0.457 0.073
Chain 1: 1100 -8008.787 0.360 0.056
Chain 1: 1200 -7787.247 0.144 0.052
Chain 1: 1300 -7594.292 0.046 0.048
Chain 1: 1400 -7930.398 0.040 0.042
Chain 1: 1500 -7597.545 0.040 0.042
Chain 1: 1600 -7767.089 0.035 0.035
Chain 1: 1700 -7670.677 0.031 0.030
Chain 1: 1800 -7663.063 0.030 0.030
Chain 1: 1900 -7632.164 0.027 0.028
Chain 1: 2000 -7714.509 0.023 0.025
Chain 1: 2100 -7646.604 0.020 0.022
Chain 1: 2200 -8012.203 0.022 0.022
Chain 1: 2300 -7644.956 0.024 0.022
Chain 1: 2400 -7539.159 0.021 0.014
Chain 1: 2500 -7605.271 0.018 0.013
Chain 1: 2600 -7558.935 0.016 0.011
Chain 1: 2700 -7456.710 0.016 0.011
Chain 1: 2800 -7699.285 0.019 0.014
Chain 1: 2900 -7399.247 0.023 0.014
Chain 1: 3000 -7550.222 0.024 0.020
Chain 1: 3100 -7552.497 0.023 0.020
Chain 1: 3200 -7767.406 0.021 0.020
Chain 1: 3300 -7465.993 0.020 0.020
Chain 1: 3400 -7703.387 0.022 0.028
Chain 1: 3500 -7453.140 0.024 0.031
Chain 1: 3600 -7529.491 0.025 0.031
Chain 1: 3700 -7475.993 0.024 0.031
Chain 1: 3800 -7587.367 0.023 0.028
Chain 1: 3900 -7440.277 0.020 0.020
Chain 1: 4000 -7422.574 0.019 0.020
Chain 1: 4100 -7438.918 0.019 0.020
Chain 1: 4200 -7517.575 0.017 0.015
Chain 1: 4300 -7419.728 0.014 0.013
Chain 1: 4400 -7458.952 0.012 0.010
Chain 1: 4500 -7563.324 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003122 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.22 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86218.803 1.000 1.000
Chain 1: 200 -14139.649 3.049 5.098
Chain 1: 300 -10479.543 2.149 1.000
Chain 1: 400 -11482.825 1.634 1.000
Chain 1: 500 -9466.922 1.349 0.349
Chain 1: 600 -9494.455 1.125 0.349
Chain 1: 700 -8919.615 0.974 0.213
Chain 1: 800 -9482.861 0.859 0.213
Chain 1: 900 -9311.731 0.766 0.087
Chain 1: 1000 -8980.388 0.693 0.087
Chain 1: 1100 -9305.019 0.596 0.064 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8814.093 0.092 0.059
Chain 1: 1300 -9053.854 0.060 0.056
Chain 1: 1400 -9189.248 0.053 0.037
Chain 1: 1500 -9030.219 0.033 0.035
Chain 1: 1600 -9140.919 0.034 0.035
Chain 1: 1700 -9215.657 0.028 0.026
Chain 1: 1800 -8792.575 0.027 0.026
Chain 1: 1900 -8893.010 0.027 0.026
Chain 1: 2000 -8867.639 0.023 0.018
Chain 1: 2100 -8993.471 0.021 0.015
Chain 1: 2200 -8795.386 0.018 0.015
Chain 1: 2300 -8887.908 0.016 0.014
Chain 1: 2400 -8956.578 0.015 0.012
Chain 1: 2500 -8902.904 0.014 0.011
Chain 1: 2600 -8904.469 0.013 0.010
Chain 1: 2700 -8821.045 0.013 0.010
Chain 1: 2800 -8780.682 0.009 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002945 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.45 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8381214.541 1.000 1.000
Chain 1: 200 -1582513.063 2.648 4.296
Chain 1: 300 -892322.125 2.023 1.000
Chain 1: 400 -459541.828 1.753 1.000
Chain 1: 500 -360105.827 1.458 0.942
Chain 1: 600 -234735.201 1.304 0.942
Chain 1: 700 -120394.838 1.253 0.942
Chain 1: 800 -87494.773 1.143 0.942
Chain 1: 900 -67726.461 1.049 0.773
Chain 1: 1000 -52442.478 0.973 0.773
Chain 1: 1100 -39847.210 0.905 0.534 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39014.427 0.477 0.376
Chain 1: 1300 -26894.086 0.445 0.376
Chain 1: 1400 -26606.407 0.352 0.316
Chain 1: 1500 -23174.716 0.339 0.316
Chain 1: 1600 -22385.848 0.289 0.292
Chain 1: 1700 -21250.137 0.200 0.291
Chain 1: 1800 -21192.292 0.162 0.148
Chain 1: 1900 -21518.491 0.135 0.053
Chain 1: 2000 -20024.712 0.113 0.053
Chain 1: 2100 -20263.197 0.082 0.035
Chain 1: 2200 -20490.685 0.081 0.035
Chain 1: 2300 -20106.966 0.038 0.019
Chain 1: 2400 -19878.922 0.038 0.019
Chain 1: 2500 -19681.402 0.024 0.015
Chain 1: 2600 -19311.044 0.023 0.015
Chain 1: 2700 -19267.776 0.018 0.012
Chain 1: 2800 -18984.808 0.019 0.015
Chain 1: 2900 -19266.226 0.019 0.015
Chain 1: 3000 -19252.241 0.012 0.012
Chain 1: 3100 -19337.322 0.011 0.011
Chain 1: 3200 -19027.802 0.011 0.015
Chain 1: 3300 -19232.659 0.010 0.011
Chain 1: 3400 -18707.449 0.012 0.015
Chain 1: 3500 -19319.708 0.014 0.015
Chain 1: 3600 -18625.884 0.016 0.015
Chain 1: 3700 -19013.155 0.018 0.016
Chain 1: 3800 -17972.217 0.022 0.020
Chain 1: 3900 -17968.410 0.021 0.020
Chain 1: 4000 -18085.641 0.021 0.020
Chain 1: 4100 -17999.443 0.021 0.020
Chain 1: 4200 -17815.500 0.021 0.020
Chain 1: 4300 -17953.957 0.020 0.020
Chain 1: 4400 -17910.653 0.018 0.010
Chain 1: 4500 -17813.233 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001281 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.81 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12108.124 1.000 1.000
Chain 1: 200 -8932.600 0.678 1.000
Chain 1: 300 -7858.792 0.497 0.355
Chain 1: 400 -7950.060 0.376 0.355
Chain 1: 500 -7923.273 0.301 0.137
Chain 1: 600 -7785.681 0.254 0.137
Chain 1: 700 -7707.903 0.219 0.018
Chain 1: 800 -7719.650 0.192 0.018
Chain 1: 900 -7686.234 0.171 0.011
Chain 1: 1000 -7751.919 0.155 0.011
Chain 1: 1100 -7831.673 0.056 0.010
Chain 1: 1200 -7742.622 0.022 0.010
Chain 1: 1300 -7683.110 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001435 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.35 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57855.287 1.000 1.000
Chain 1: 200 -17332.009 1.669 2.338
Chain 1: 300 -8543.757 1.456 1.029
Chain 1: 400 -8142.880 1.104 1.029
Chain 1: 500 -8363.505 0.888 1.000
Chain 1: 600 -8708.784 0.747 1.000
Chain 1: 700 -8052.964 0.652 0.081
Chain 1: 800 -8226.728 0.573 0.081
Chain 1: 900 -7877.866 0.514 0.049
Chain 1: 1000 -7889.871 0.463 0.049
Chain 1: 1100 -7755.926 0.365 0.044
Chain 1: 1200 -7758.424 0.131 0.040
Chain 1: 1300 -7604.599 0.030 0.026
Chain 1: 1400 -7859.294 0.028 0.026
Chain 1: 1500 -7599.795 0.029 0.032
Chain 1: 1600 -7501.969 0.027 0.021
Chain 1: 1700 -7502.019 0.018 0.020
Chain 1: 1800 -7548.360 0.017 0.017
Chain 1: 1900 -7559.711 0.013 0.013
Chain 1: 2000 -7556.875 0.013 0.013
Chain 1: 2100 -7580.023 0.011 0.006 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002568 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.68 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85938.915 1.000 1.000
Chain 1: 200 -13172.522 3.262 5.524
Chain 1: 300 -9618.347 2.298 1.000
Chain 1: 400 -10661.230 1.748 1.000
Chain 1: 500 -8544.794 1.448 0.370
Chain 1: 600 -8475.280 1.208 0.370
Chain 1: 700 -8134.410 1.041 0.248
Chain 1: 800 -8639.651 0.918 0.248
Chain 1: 900 -8581.248 0.817 0.098
Chain 1: 1000 -8339.407 0.738 0.098
Chain 1: 1100 -8498.549 0.640 0.058 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8058.343 0.093 0.055
Chain 1: 1300 -8314.143 0.059 0.042
Chain 1: 1400 -8339.327 0.050 0.031
Chain 1: 1500 -8222.405 0.027 0.029
Chain 1: 1600 -8327.428 0.027 0.029
Chain 1: 1700 -8413.639 0.024 0.019
Chain 1: 1800 -8020.700 0.023 0.019
Chain 1: 1900 -8122.705 0.023 0.019
Chain 1: 2000 -8093.092 0.021 0.014
Chain 1: 2100 -8218.100 0.021 0.014
Chain 1: 2200 -8002.790 0.018 0.014
Chain 1: 2300 -8151.452 0.017 0.014
Chain 1: 2400 -8166.553 0.016 0.014
Chain 1: 2500 -8133.924 0.015 0.013
Chain 1: 2600 -8136.050 0.014 0.013
Chain 1: 2700 -8042.710 0.014 0.013
Chain 1: 2800 -8015.127 0.010 0.012 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003312 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.12 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8425083.892 1.000 1.000
Chain 1: 200 -1585931.368 2.656 4.312
Chain 1: 300 -889471.718 2.032 1.000
Chain 1: 400 -456802.919 1.761 1.000
Chain 1: 500 -356980.499 1.464 0.947
Chain 1: 600 -231947.427 1.310 0.947
Chain 1: 700 -118502.469 1.260 0.947
Chain 1: 800 -85839.467 1.150 0.947
Chain 1: 900 -66242.282 1.055 0.783
Chain 1: 1000 -51091.595 0.979 0.783
Chain 1: 1100 -38624.796 0.911 0.539 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37802.117 0.482 0.381
Chain 1: 1300 -25816.987 0.450 0.381
Chain 1: 1400 -25539.176 0.357 0.323
Chain 1: 1500 -22143.498 0.344 0.323
Chain 1: 1600 -21364.918 0.294 0.297
Chain 1: 1700 -20245.888 0.204 0.296
Chain 1: 1800 -20191.656 0.166 0.153
Chain 1: 1900 -20517.423 0.138 0.055
Chain 1: 2000 -19033.513 0.116 0.055
Chain 1: 2100 -19271.284 0.085 0.036
Chain 1: 2200 -19497.078 0.084 0.036
Chain 1: 2300 -19115.026 0.040 0.020
Chain 1: 2400 -18887.388 0.040 0.020
Chain 1: 2500 -18689.316 0.025 0.016
Chain 1: 2600 -18319.930 0.024 0.016
Chain 1: 2700 -18277.080 0.019 0.012
Chain 1: 2800 -17994.132 0.020 0.016
Chain 1: 2900 -18275.096 0.020 0.015
Chain 1: 3000 -18261.248 0.012 0.012
Chain 1: 3100 -18346.209 0.011 0.012
Chain 1: 3200 -18037.164 0.012 0.015
Chain 1: 3300 -18241.698 0.011 0.012
Chain 1: 3400 -17717.106 0.013 0.015
Chain 1: 3500 -18328.168 0.015 0.016
Chain 1: 3600 -17635.883 0.017 0.016
Chain 1: 3700 -18021.896 0.019 0.017
Chain 1: 3800 -16983.184 0.023 0.021
Chain 1: 3900 -16979.384 0.022 0.021
Chain 1: 4000 -17096.679 0.022 0.021
Chain 1: 4100 -17010.533 0.023 0.021
Chain 1: 4200 -16827.148 0.022 0.021
Chain 1: 4300 -16965.281 0.022 0.021
Chain 1: 4400 -16922.367 0.019 0.011
Chain 1: 4500 -16824.980 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001283 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.83 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -11996.548 1.000 1.000
Chain 1: 200 -8937.644 0.671 1.000
Chain 1: 300 -7946.911 0.489 0.342
Chain 1: 400 -8028.599 0.369 0.342
Chain 1: 500 -7958.498 0.297 0.125
Chain 1: 600 -7952.670 0.248 0.125
Chain 1: 700 -7739.838 0.216 0.027
Chain 1: 800 -7756.074 0.190 0.027
Chain 1: 900 -7935.694 0.171 0.023
Chain 1: 1000 -7763.226 0.156 0.023
Chain 1: 1100 -7825.630 0.057 0.022
Chain 1: 1200 -7760.565 0.024 0.010
Chain 1: 1300 -7721.084 0.012 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001558 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.58 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61137.556 1.000 1.000
Chain 1: 200 -17361.721 1.761 2.521
Chain 1: 300 -8638.130 1.510 1.010
Chain 1: 400 -8124.823 1.149 1.010
Chain 1: 500 -8304.478 0.923 1.000
Chain 1: 600 -8410.731 0.771 1.000
Chain 1: 700 -7837.827 0.672 0.073
Chain 1: 800 -7937.162 0.589 0.073
Chain 1: 900 -7906.801 0.524 0.063
Chain 1: 1000 -7835.273 0.473 0.063
Chain 1: 1100 -7757.470 0.374 0.022
Chain 1: 1200 -7591.793 0.124 0.022
Chain 1: 1300 -7627.898 0.023 0.013
Chain 1: 1400 -7871.783 0.020 0.013
Chain 1: 1500 -7591.016 0.022 0.013
Chain 1: 1600 -7500.050 0.022 0.013
Chain 1: 1700 -7480.906 0.014 0.012
Chain 1: 1800 -7513.213 0.014 0.010
Chain 1: 1900 -7552.509 0.014 0.010
Chain 1: 2000 -7554.781 0.013 0.010
Chain 1: 2100 -7570.674 0.012 0.005 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002607 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 26.07 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86156.288 1.000 1.000
Chain 1: 200 -13057.232 3.299 5.598
Chain 1: 300 -9554.166 2.322 1.000
Chain 1: 400 -10405.174 1.762 1.000
Chain 1: 500 -8405.953 1.457 0.367
Chain 1: 600 -8168.617 1.219 0.367
Chain 1: 700 -8416.572 1.049 0.238
Chain 1: 800 -8444.778 0.918 0.238
Chain 1: 900 -8437.191 0.816 0.082
Chain 1: 1000 -8295.909 0.736 0.082
Chain 1: 1100 -8485.995 0.639 0.029 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8177.913 0.083 0.029
Chain 1: 1300 -8327.813 0.048 0.029
Chain 1: 1400 -8321.585 0.040 0.022
Chain 1: 1500 -8227.163 0.017 0.018
Chain 1: 1600 -8311.298 0.015 0.017
Chain 1: 1700 -8416.025 0.013 0.012
Chain 1: 1800 -8036.462 0.018 0.017
Chain 1: 1900 -8134.304 0.019 0.017
Chain 1: 2000 -8104.985 0.018 0.012
Chain 1: 2100 -8250.709 0.017 0.012
Chain 1: 2200 -8028.017 0.016 0.012
Chain 1: 2300 -8158.096 0.016 0.012
Chain 1: 2400 -8056.668 0.017 0.013
Chain 1: 2500 -8111.796 0.017 0.013
Chain 1: 2600 -8124.422 0.016 0.013
Chain 1: 2700 -8045.477 0.015 0.013
Chain 1: 2800 -8031.053 0.011 0.012
Chain 1: 2900 -8019.528 0.010 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003919 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 39.19 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8426774.144 1.000 1.000
Chain 1: 200 -1587593.858 2.654 4.308
Chain 1: 300 -890732.592 2.030 1.000
Chain 1: 400 -457263.389 1.760 1.000
Chain 1: 500 -357256.570 1.464 0.948
Chain 1: 600 -232296.258 1.309 0.948
Chain 1: 700 -118599.838 1.259 0.948
Chain 1: 800 -85844.107 1.150 0.948
Chain 1: 900 -66208.150 1.055 0.782
Chain 1: 1000 -51017.011 0.979 0.782
Chain 1: 1100 -38514.754 0.912 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37684.589 0.483 0.382
Chain 1: 1300 -25676.405 0.451 0.382
Chain 1: 1400 -25395.047 0.358 0.325
Chain 1: 1500 -21992.709 0.345 0.325
Chain 1: 1600 -21210.942 0.295 0.298
Chain 1: 1700 -20089.989 0.205 0.297
Chain 1: 1800 -20034.719 0.167 0.155
Chain 1: 1900 -20360.040 0.139 0.056
Chain 1: 2000 -18875.535 0.117 0.056
Chain 1: 2100 -19113.561 0.086 0.037
Chain 1: 2200 -19339.084 0.085 0.037
Chain 1: 2300 -18957.340 0.040 0.020
Chain 1: 2400 -18729.788 0.040 0.020
Chain 1: 2500 -18531.712 0.026 0.016
Chain 1: 2600 -18162.877 0.024 0.016
Chain 1: 2700 -18120.118 0.019 0.012
Chain 1: 2800 -17837.358 0.020 0.016
Chain 1: 2900 -18118.138 0.020 0.015
Chain 1: 3000 -18104.415 0.012 0.012
Chain 1: 3100 -18189.268 0.011 0.012
Chain 1: 3200 -17880.532 0.012 0.015
Chain 1: 3300 -18084.779 0.011 0.012
Chain 1: 3400 -17560.748 0.013 0.015
Chain 1: 3500 -18171.008 0.015 0.016
Chain 1: 3600 -17479.774 0.017 0.016
Chain 1: 3700 -17865.025 0.019 0.017
Chain 1: 3800 -16827.928 0.024 0.022
Chain 1: 3900 -16824.132 0.022 0.022
Chain 1: 4000 -16941.445 0.023 0.022
Chain 1: 4100 -16855.378 0.023 0.022
Chain 1: 4200 -16672.301 0.022 0.022
Chain 1: 4300 -16810.218 0.022 0.022
Chain 1: 4400 -16767.610 0.019 0.011
Chain 1: 4500 -16670.238 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001242 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.42 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48890.782 1.000 1.000
Chain 1: 200 -15118.039 1.617 2.234
Chain 1: 300 -15660.528 1.090 1.000
Chain 1: 400 -13604.445 0.855 1.000
Chain 1: 500 -22917.842 0.765 0.406
Chain 1: 600 -12576.285 0.775 0.822
Chain 1: 700 -13138.139 0.670 0.406
Chain 1: 800 -21105.914 0.634 0.406
Chain 1: 900 -16783.319 0.592 0.378
Chain 1: 1000 -18642.729 0.543 0.378
Chain 1: 1100 -10933.473 0.513 0.378 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -11510.219 0.295 0.258
Chain 1: 1300 -12626.375 0.300 0.258
Chain 1: 1400 -11651.175 0.293 0.258
Chain 1: 1500 -13551.148 0.267 0.140
Chain 1: 1600 -11470.870 0.203 0.140
Chain 1: 1700 -13054.426 0.210 0.140
Chain 1: 1800 -17492.574 0.198 0.140
Chain 1: 1900 -10394.083 0.241 0.140
Chain 1: 2000 -9880.190 0.236 0.140
Chain 1: 2100 -15500.022 0.202 0.140
Chain 1: 2200 -11545.860 0.231 0.181
Chain 1: 2300 -9332.201 0.246 0.237
Chain 1: 2400 -22052.546 0.295 0.254
Chain 1: 2500 -13174.564 0.348 0.342
Chain 1: 2600 -9403.371 0.370 0.363
Chain 1: 2700 -8948.328 0.363 0.363
Chain 1: 2800 -10369.992 0.352 0.363
Chain 1: 2900 -9196.197 0.296 0.342
Chain 1: 3000 -9301.885 0.292 0.342
Chain 1: 3100 -8837.895 0.261 0.237
Chain 1: 3200 -16495.852 0.273 0.237
Chain 1: 3300 -17526.711 0.255 0.137
Chain 1: 3400 -8990.170 0.293 0.137
Chain 1: 3500 -12500.796 0.253 0.137
Chain 1: 3600 -20094.656 0.251 0.137
Chain 1: 3700 -8732.617 0.376 0.281
Chain 1: 3800 -8476.464 0.365 0.281
Chain 1: 3900 -8704.788 0.355 0.281
Chain 1: 4000 -8990.341 0.357 0.281
Chain 1: 4100 -9241.603 0.355 0.281
Chain 1: 4200 -10117.169 0.317 0.087
Chain 1: 4300 -11967.715 0.327 0.155
Chain 1: 4400 -11139.776 0.239 0.087
Chain 1: 4500 -10250.587 0.220 0.087
Chain 1: 4600 -12560.731 0.200 0.087
Chain 1: 4700 -14674.594 0.085 0.087
Chain 1: 4800 -8718.915 0.150 0.087
Chain 1: 4900 -9698.825 0.157 0.101
Chain 1: 5000 -14222.600 0.186 0.144
Chain 1: 5100 -8402.827 0.252 0.155
Chain 1: 5200 -10364.624 0.263 0.184
Chain 1: 5300 -13393.471 0.270 0.189
Chain 1: 5400 -8489.670 0.320 0.226
Chain 1: 5500 -11924.531 0.340 0.288
Chain 1: 5600 -10738.180 0.333 0.288
Chain 1: 5700 -8620.005 0.343 0.288
Chain 1: 5800 -8443.299 0.277 0.246
Chain 1: 5900 -9884.364 0.281 0.246
Chain 1: 6000 -8773.217 0.262 0.226
Chain 1: 6100 -13202.151 0.227 0.226
Chain 1: 6200 -9163.518 0.252 0.246
Chain 1: 6300 -8334.597 0.239 0.246
Chain 1: 6400 -10513.375 0.202 0.207
Chain 1: 6500 -8564.926 0.196 0.207
Chain 1: 6600 -8931.875 0.189 0.207
Chain 1: 6700 -12555.292 0.193 0.207
Chain 1: 6800 -8385.509 0.241 0.227
Chain 1: 6900 -8176.926 0.229 0.227
Chain 1: 7000 -8616.857 0.221 0.227
Chain 1: 7100 -8239.544 0.192 0.207
Chain 1: 7200 -8543.002 0.152 0.099
Chain 1: 7300 -9551.492 0.153 0.106
Chain 1: 7400 -8582.030 0.143 0.106
Chain 1: 7500 -10786.720 0.141 0.106
Chain 1: 7600 -10263.853 0.142 0.106
Chain 1: 7700 -8583.076 0.132 0.106
Chain 1: 7800 -13401.319 0.119 0.106
Chain 1: 7900 -9200.105 0.162 0.113
Chain 1: 8000 -8162.346 0.169 0.127
Chain 1: 8100 -10502.813 0.187 0.196
Chain 1: 8200 -9278.222 0.197 0.196
Chain 1: 8300 -8512.469 0.195 0.196
Chain 1: 8400 -9191.864 0.191 0.196
Chain 1: 8500 -11040.909 0.188 0.167
Chain 1: 8600 -8024.038 0.220 0.196
Chain 1: 8700 -8454.407 0.206 0.167
Chain 1: 8800 -8538.479 0.171 0.132
Chain 1: 8900 -8705.872 0.127 0.127
Chain 1: 9000 -8838.571 0.116 0.090
Chain 1: 9100 -8197.000 0.101 0.078
Chain 1: 9200 -9780.797 0.104 0.078
Chain 1: 9300 -8211.115 0.114 0.078
Chain 1: 9400 -8150.054 0.108 0.078
Chain 1: 9500 -7972.214 0.093 0.051
Chain 1: 9600 -8269.441 0.059 0.036
Chain 1: 9700 -9299.123 0.065 0.036
Chain 1: 9800 -8658.230 0.072 0.074
Chain 1: 9900 -8116.655 0.076 0.074
Chain 1: 10000 -7970.089 0.077 0.074
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001411 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.11 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57091.896 1.000 1.000
Chain 1: 200 -17489.926 1.632 2.264
Chain 1: 300 -8774.182 1.419 1.000
Chain 1: 400 -8236.953 1.081 1.000
Chain 1: 500 -8910.360 0.880 0.993
Chain 1: 600 -8051.053 0.751 0.993
Chain 1: 700 -9056.323 0.659 0.111
Chain 1: 800 -8440.086 0.586 0.111
Chain 1: 900 -8130.306 0.525 0.107
Chain 1: 1000 -7748.375 0.478 0.107
Chain 1: 1100 -7704.853 0.378 0.076
Chain 1: 1200 -7831.261 0.153 0.073
Chain 1: 1300 -7766.228 0.055 0.065
Chain 1: 1400 -7857.413 0.050 0.049
Chain 1: 1500 -7638.554 0.045 0.038
Chain 1: 1600 -7804.699 0.036 0.029
Chain 1: 1700 -7556.870 0.028 0.029
Chain 1: 1800 -7620.698 0.022 0.021
Chain 1: 1900 -7624.266 0.018 0.016
Chain 1: 2000 -7626.248 0.013 0.012
Chain 1: 2100 -7638.082 0.013 0.012
Chain 1: 2200 -7752.370 0.013 0.012
Chain 1: 2300 -7642.336 0.013 0.014
Chain 1: 2400 -7696.838 0.013 0.014
Chain 1: 2500 -7643.429 0.011 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002845 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 28.45 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85813.242 1.000 1.000
Chain 1: 200 -13521.780 3.173 5.346
Chain 1: 300 -9824.316 2.241 1.000
Chain 1: 400 -10906.950 1.705 1.000
Chain 1: 500 -8765.630 1.413 0.376
Chain 1: 600 -8679.364 1.179 0.376
Chain 1: 700 -8486.369 1.014 0.244
Chain 1: 800 -8188.642 0.892 0.244
Chain 1: 900 -8166.246 0.793 0.099
Chain 1: 1000 -8828.034 0.721 0.099
Chain 1: 1100 -8401.277 0.626 0.075 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8758.153 0.096 0.051
Chain 1: 1300 -8329.807 0.063 0.051
Chain 1: 1400 -8371.659 0.054 0.041
Chain 1: 1500 -8283.448 0.031 0.036
Chain 1: 1600 -8293.437 0.030 0.036
Chain 1: 1700 -8183.852 0.029 0.036
Chain 1: 1800 -8240.172 0.026 0.013
Chain 1: 1900 -8119.320 0.027 0.015
Chain 1: 2000 -8178.011 0.020 0.013
Chain 1: 2100 -8316.475 0.017 0.013
Chain 1: 2200 -8119.009 0.015 0.013
Chain 1: 2300 -8266.357 0.012 0.013
Chain 1: 2400 -8111.006 0.013 0.015
Chain 1: 2500 -8180.226 0.013 0.015
Chain 1: 2600 -8094.712 0.014 0.015
Chain 1: 2700 -8126.842 0.013 0.015
Chain 1: 2800 -8088.028 0.013 0.015
Chain 1: 2900 -8179.824 0.012 0.011
Chain 1: 3000 -8004.933 0.014 0.017
Chain 1: 3100 -8170.160 0.014 0.018
Chain 1: 3200 -8043.051 0.013 0.016
Chain 1: 3300 -8052.636 0.012 0.011
Chain 1: 3400 -8203.338 0.012 0.011
Chain 1: 3500 -8190.010 0.011 0.011
Chain 1: 3600 -8000.561 0.012 0.016
Chain 1: 3700 -8143.152 0.014 0.018
Chain 1: 3800 -8007.457 0.015 0.018
Chain 1: 3900 -7942.796 0.015 0.018
Chain 1: 4000 -8017.226 0.013 0.017
Chain 1: 4100 -8007.898 0.011 0.016
Chain 1: 4200 -7996.409 0.010 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003455 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.55 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8372622.389 1.000 1.000
Chain 1: 200 -1580375.379 2.649 4.298
Chain 1: 300 -890362.486 2.024 1.000
Chain 1: 400 -457867.110 1.754 1.000
Chain 1: 500 -358535.209 1.459 0.945
Chain 1: 600 -233553.979 1.305 0.945
Chain 1: 700 -119551.573 1.255 0.945
Chain 1: 800 -86704.373 1.145 0.945
Chain 1: 900 -67001.240 1.051 0.775
Chain 1: 1000 -51756.726 0.975 0.775
Chain 1: 1100 -39190.311 0.907 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38364.915 0.479 0.379
Chain 1: 1300 -26272.003 0.448 0.379
Chain 1: 1400 -25988.129 0.355 0.321
Chain 1: 1500 -22562.557 0.342 0.321
Chain 1: 1600 -21775.720 0.292 0.295
Chain 1: 1700 -20643.215 0.202 0.294
Chain 1: 1800 -20586.257 0.165 0.152
Chain 1: 1900 -20912.670 0.137 0.055
Chain 1: 2000 -19419.892 0.115 0.055
Chain 1: 2100 -19658.564 0.084 0.036
Chain 1: 2200 -19885.793 0.083 0.036
Chain 1: 2300 -19502.213 0.039 0.020
Chain 1: 2400 -19274.118 0.039 0.020
Chain 1: 2500 -19076.361 0.025 0.016
Chain 1: 2600 -18706.077 0.024 0.016
Chain 1: 2700 -18662.815 0.018 0.012
Chain 1: 2800 -18379.649 0.020 0.015
Chain 1: 2900 -18661.109 0.019 0.015
Chain 1: 3000 -18647.214 0.012 0.012
Chain 1: 3100 -18732.298 0.011 0.012
Chain 1: 3200 -18422.708 0.012 0.015
Chain 1: 3300 -18627.620 0.011 0.012
Chain 1: 3400 -18102.179 0.013 0.015
Chain 1: 3500 -18714.717 0.015 0.015
Chain 1: 3600 -18020.503 0.017 0.015
Chain 1: 3700 -18408.039 0.018 0.017
Chain 1: 3800 -17366.451 0.023 0.021
Chain 1: 3900 -17362.581 0.021 0.021
Chain 1: 4000 -17479.857 0.022 0.021
Chain 1: 4100 -17393.598 0.022 0.021
Chain 1: 4200 -17209.518 0.021 0.021
Chain 1: 4300 -17348.118 0.021 0.021
Chain 1: 4400 -17304.708 0.019 0.011
Chain 1: 4500 -17207.215 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001274 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.74 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49072.574 1.000 1.000
Chain 1: 200 -33990.953 0.722 1.000
Chain 1: 300 -18435.580 0.762 0.844
Chain 1: 400 -21510.659 0.608 0.844
Chain 1: 500 -13210.431 0.612 0.628
Chain 1: 600 -16508.674 0.543 0.628
Chain 1: 700 -15620.374 0.474 0.444
Chain 1: 800 -12550.847 0.445 0.444
Chain 1: 900 -19171.344 0.434 0.345
Chain 1: 1000 -20297.995 0.396 0.345
Chain 1: 1100 -23061.164 0.308 0.245
Chain 1: 1200 -13086.324 0.340 0.245
Chain 1: 1300 -12483.585 0.260 0.200
Chain 1: 1400 -10948.422 0.260 0.200
Chain 1: 1500 -10310.019 0.203 0.140
Chain 1: 1600 -9890.818 0.188 0.120
Chain 1: 1700 -11165.683 0.193 0.120
Chain 1: 1800 -13398.273 0.186 0.120
Chain 1: 1900 -11966.751 0.163 0.120
Chain 1: 2000 -11062.069 0.166 0.120
Chain 1: 2100 -9380.670 0.172 0.120
Chain 1: 2200 -11380.588 0.113 0.120
Chain 1: 2300 -9544.889 0.127 0.140
Chain 1: 2400 -10959.391 0.126 0.129
Chain 1: 2500 -9790.082 0.132 0.129
Chain 1: 2600 -9291.284 0.133 0.129
Chain 1: 2700 -9290.886 0.122 0.129
Chain 1: 2800 -8994.611 0.108 0.120
Chain 1: 2900 -10008.465 0.107 0.119
Chain 1: 3000 -8940.073 0.110 0.120
Chain 1: 3100 -15096.636 0.133 0.120
Chain 1: 3200 -9060.489 0.182 0.120
Chain 1: 3300 -9471.225 0.167 0.119
Chain 1: 3400 -12422.524 0.178 0.119
Chain 1: 3500 -9449.414 0.198 0.120
Chain 1: 3600 -8773.863 0.200 0.120
Chain 1: 3700 -9148.640 0.204 0.120
Chain 1: 3800 -16716.833 0.246 0.238
Chain 1: 3900 -10136.313 0.301 0.315
Chain 1: 4000 -8750.508 0.305 0.315
Chain 1: 4100 -9126.274 0.268 0.238
Chain 1: 4200 -10260.638 0.213 0.158
Chain 1: 4300 -9036.940 0.222 0.158
Chain 1: 4400 -9010.983 0.198 0.135
Chain 1: 4500 -13973.645 0.202 0.135
Chain 1: 4600 -12532.905 0.206 0.135
Chain 1: 4700 -8833.470 0.244 0.158
Chain 1: 4800 -8744.038 0.200 0.135
Chain 1: 4900 -8570.248 0.137 0.115
Chain 1: 5000 -10277.200 0.138 0.115
Chain 1: 5100 -8557.443 0.154 0.135
Chain 1: 5200 -9922.374 0.156 0.138
Chain 1: 5300 -14149.010 0.173 0.166
Chain 1: 5400 -8779.038 0.233 0.201
Chain 1: 5500 -14072.675 0.236 0.201
Chain 1: 5600 -9559.153 0.271 0.299
Chain 1: 5700 -10717.629 0.240 0.201
Chain 1: 5800 -8464.353 0.266 0.266
Chain 1: 5900 -13540.886 0.301 0.299
Chain 1: 6000 -8998.531 0.335 0.375
Chain 1: 6100 -9690.022 0.322 0.375
Chain 1: 6200 -9085.625 0.315 0.375
Chain 1: 6300 -10156.283 0.296 0.375
Chain 1: 6400 -14698.467 0.265 0.309
Chain 1: 6500 -9346.849 0.285 0.309
Chain 1: 6600 -8669.912 0.246 0.266
Chain 1: 6700 -12773.739 0.267 0.309
Chain 1: 6800 -12884.370 0.241 0.309
Chain 1: 6900 -12382.485 0.208 0.105
Chain 1: 7000 -10147.189 0.179 0.105
Chain 1: 7100 -8772.175 0.188 0.157
Chain 1: 7200 -9862.134 0.192 0.157
Chain 1: 7300 -8569.573 0.197 0.157
Chain 1: 7400 -8262.759 0.170 0.151
Chain 1: 7500 -8288.628 0.113 0.111
Chain 1: 7600 -8441.681 0.107 0.111
Chain 1: 7700 -12010.812 0.104 0.111
Chain 1: 7800 -8330.490 0.148 0.151
Chain 1: 7900 -8356.497 0.144 0.151
Chain 1: 8000 -8283.532 0.123 0.111
Chain 1: 8100 -8199.487 0.108 0.037
Chain 1: 8200 -8600.686 0.102 0.037
Chain 1: 8300 -9555.872 0.097 0.037
Chain 1: 8400 -8656.631 0.103 0.047
Chain 1: 8500 -8290.127 0.107 0.047
Chain 1: 8600 -8452.010 0.107 0.047
Chain 1: 8700 -9092.544 0.085 0.047
Chain 1: 8800 -9880.656 0.049 0.047
Chain 1: 8900 -9727.185 0.050 0.047
Chain 1: 9000 -11026.196 0.061 0.070
Chain 1: 9100 -8087.559 0.096 0.080
Chain 1: 9200 -8266.604 0.094 0.080
Chain 1: 9300 -8472.054 0.086 0.070
Chain 1: 9400 -11863.416 0.104 0.070
Chain 1: 9500 -8214.383 0.144 0.080
Chain 1: 9600 -8618.685 0.147 0.080
Chain 1: 9700 -10893.795 0.161 0.118
Chain 1: 9800 -8611.151 0.179 0.209
Chain 1: 9900 -10777.327 0.198 0.209
Chain 1: 10000 -10587.168 0.188 0.209
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001397 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.97 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58210.242 1.000 1.000
Chain 1: 200 -17725.458 1.642 2.284
Chain 1: 300 -8695.376 1.441 1.038
Chain 1: 400 -8136.318 1.098 1.038
Chain 1: 500 -8347.409 0.883 1.000
Chain 1: 600 -8467.365 0.738 1.000
Chain 1: 700 -8650.158 0.636 0.069
Chain 1: 800 -8648.937 0.556 0.069
Chain 1: 900 -7944.671 0.505 0.069
Chain 1: 1000 -7917.119 0.454 0.069
Chain 1: 1100 -7646.634 0.358 0.035
Chain 1: 1200 -7584.824 0.130 0.025
Chain 1: 1300 -7706.109 0.028 0.021
Chain 1: 1400 -7818.880 0.023 0.016
Chain 1: 1500 -7563.155 0.024 0.016
Chain 1: 1600 -7741.224 0.024 0.021
Chain 1: 1700 -7489.543 0.026 0.023
Chain 1: 1800 -7540.249 0.026 0.023
Chain 1: 1900 -7569.865 0.018 0.016
Chain 1: 2000 -7580.638 0.018 0.016
Chain 1: 2100 -7567.385 0.014 0.014
Chain 1: 2200 -7675.362 0.015 0.014
Chain 1: 2300 -7548.659 0.015 0.014
Chain 1: 2400 -7603.851 0.014 0.014
Chain 1: 2500 -7564.483 0.011 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003395 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.95 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86411.048 1.000 1.000
Chain 1: 200 -13538.646 3.191 5.383
Chain 1: 300 -9876.948 2.251 1.000
Chain 1: 400 -10766.736 1.709 1.000
Chain 1: 500 -8873.030 1.410 0.371
Chain 1: 600 -8861.087 1.175 0.371
Chain 1: 700 -8577.705 1.012 0.213
Chain 1: 800 -8745.294 0.888 0.213
Chain 1: 900 -8711.424 0.790 0.083
Chain 1: 1000 -8609.251 0.712 0.083
Chain 1: 1100 -8664.618 0.613 0.033 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8207.620 0.080 0.033
Chain 1: 1300 -8579.497 0.047 0.033
Chain 1: 1400 -8570.056 0.039 0.019
Chain 1: 1500 -8436.676 0.019 0.016
Chain 1: 1600 -8546.355 0.020 0.016
Chain 1: 1700 -8626.036 0.018 0.013
Chain 1: 1800 -8204.931 0.021 0.013
Chain 1: 1900 -8304.215 0.022 0.013
Chain 1: 2000 -8278.547 0.021 0.013
Chain 1: 2100 -8403.456 0.022 0.015
Chain 1: 2200 -8209.939 0.019 0.015
Chain 1: 2300 -8299.010 0.015 0.013
Chain 1: 2400 -8368.117 0.016 0.013
Chain 1: 2500 -8314.262 0.015 0.012
Chain 1: 2600 -8315.176 0.014 0.011
Chain 1: 2700 -8232.116 0.014 0.011
Chain 1: 2800 -8192.649 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003173 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.73 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8406971.746 1.000 1.000
Chain 1: 200 -1585812.607 2.651 4.301
Chain 1: 300 -890923.297 2.027 1.000
Chain 1: 400 -457318.579 1.757 1.000
Chain 1: 500 -357411.971 1.462 0.948
Chain 1: 600 -232579.380 1.308 0.948
Chain 1: 700 -119061.465 1.257 0.948
Chain 1: 800 -86312.872 1.147 0.948
Chain 1: 900 -66708.966 1.052 0.780
Chain 1: 1000 -51542.098 0.977 0.780
Chain 1: 1100 -39049.134 0.909 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38233.360 0.481 0.379
Chain 1: 1300 -26219.335 0.448 0.379
Chain 1: 1400 -25942.779 0.355 0.320
Chain 1: 1500 -22536.072 0.342 0.320
Chain 1: 1600 -21754.748 0.292 0.294
Chain 1: 1700 -20631.778 0.202 0.294
Chain 1: 1800 -20576.928 0.164 0.151
Chain 1: 1900 -20903.145 0.136 0.054
Chain 1: 2000 -19415.683 0.115 0.054
Chain 1: 2100 -19654.242 0.084 0.036
Chain 1: 2200 -19880.286 0.083 0.036
Chain 1: 2300 -19497.791 0.039 0.020
Chain 1: 2400 -19269.849 0.039 0.020
Chain 1: 2500 -19071.646 0.025 0.016
Chain 1: 2600 -18702.031 0.023 0.016
Chain 1: 2700 -18659.082 0.018 0.012
Chain 1: 2800 -18375.737 0.020 0.015
Chain 1: 2900 -18657.020 0.019 0.015
Chain 1: 3000 -18643.327 0.012 0.012
Chain 1: 3100 -18728.276 0.011 0.012
Chain 1: 3200 -18418.980 0.012 0.015
Chain 1: 3300 -18623.704 0.011 0.012
Chain 1: 3400 -18098.540 0.013 0.015
Chain 1: 3500 -18710.462 0.015 0.015
Chain 1: 3600 -18017.099 0.017 0.015
Chain 1: 3700 -18403.864 0.018 0.017
Chain 1: 3800 -17363.450 0.023 0.021
Chain 1: 3900 -17359.545 0.021 0.021
Chain 1: 4000 -17476.902 0.022 0.021
Chain 1: 4100 -17390.585 0.022 0.021
Chain 1: 4200 -17206.845 0.021 0.021
Chain 1: 4300 -17345.285 0.021 0.021
Chain 1: 4400 -17302.100 0.019 0.011
Chain 1: 4500 -17204.572 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001299 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.99 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49481.999 1.000 1.000
Chain 1: 200 -21402.638 1.156 1.312
Chain 1: 300 -16651.691 0.866 1.000
Chain 1: 400 -16062.146 0.658 1.000
Chain 1: 500 -14799.368 0.544 0.285
Chain 1: 600 -15411.791 0.460 0.285
Chain 1: 700 -19967.813 0.427 0.228
Chain 1: 800 -15604.209 0.408 0.280
Chain 1: 900 -12016.953 0.396 0.280
Chain 1: 1000 -12273.863 0.359 0.280
Chain 1: 1100 -11680.461 0.264 0.228
Chain 1: 1200 -13975.056 0.149 0.164
Chain 1: 1300 -12773.456 0.130 0.094
Chain 1: 1400 -11035.985 0.142 0.157
Chain 1: 1500 -19020.612 0.175 0.164
Chain 1: 1600 -21480.513 0.183 0.164
Chain 1: 1700 -18850.481 0.174 0.157
Chain 1: 1800 -14497.254 0.176 0.157
Chain 1: 1900 -16630.728 0.159 0.140
Chain 1: 2000 -10393.399 0.217 0.157
Chain 1: 2100 -12772.202 0.230 0.164
Chain 1: 2200 -12023.808 0.220 0.157
Chain 1: 2300 -9550.450 0.237 0.186
Chain 1: 2400 -10123.760 0.227 0.186
Chain 1: 2500 -10922.009 0.192 0.140
Chain 1: 2600 -10280.919 0.187 0.140
Chain 1: 2700 -10561.379 0.175 0.128
Chain 1: 2800 -22548.433 0.199 0.128
Chain 1: 2900 -11395.883 0.284 0.186
Chain 1: 3000 -9451.336 0.244 0.186
Chain 1: 3100 -9383.386 0.226 0.073
Chain 1: 3200 -9659.698 0.223 0.073
Chain 1: 3300 -10887.143 0.208 0.073
Chain 1: 3400 -9815.486 0.214 0.109
Chain 1: 3500 -9699.985 0.207 0.109
Chain 1: 3600 -10313.951 0.207 0.109
Chain 1: 3700 -9132.541 0.217 0.113
Chain 1: 3800 -9237.651 0.165 0.109
Chain 1: 3900 -9413.619 0.069 0.060
Chain 1: 4000 -17615.424 0.095 0.060
Chain 1: 4100 -9260.182 0.185 0.109
Chain 1: 4200 -13605.021 0.214 0.113
Chain 1: 4300 -9890.547 0.240 0.129
Chain 1: 4400 -14998.405 0.263 0.319
Chain 1: 4500 -11881.081 0.288 0.319
Chain 1: 4600 -9454.425 0.308 0.319
Chain 1: 4700 -10047.200 0.301 0.319
Chain 1: 4800 -8985.114 0.312 0.319
Chain 1: 4900 -9455.765 0.315 0.319
Chain 1: 5000 -15257.747 0.306 0.319
Chain 1: 5100 -8939.265 0.287 0.319
Chain 1: 5200 -11088.842 0.274 0.262
Chain 1: 5300 -8837.218 0.262 0.257
Chain 1: 5400 -15361.252 0.271 0.257
Chain 1: 5500 -9353.005 0.309 0.257
Chain 1: 5600 -10357.624 0.293 0.255
Chain 1: 5700 -14447.973 0.315 0.283
Chain 1: 5800 -10759.786 0.338 0.343
Chain 1: 5900 -15357.059 0.363 0.343
Chain 1: 6000 -13384.446 0.339 0.299
Chain 1: 6100 -14382.819 0.275 0.283
Chain 1: 6200 -9197.087 0.312 0.299
Chain 1: 6300 -9065.860 0.288 0.299
Chain 1: 6400 -9115.582 0.247 0.283
Chain 1: 6500 -9116.529 0.182 0.147
Chain 1: 6600 -9891.692 0.180 0.147
Chain 1: 6700 -8766.585 0.165 0.128
Chain 1: 6800 -14296.217 0.169 0.128
Chain 1: 6900 -9269.222 0.194 0.128
Chain 1: 7000 -9661.736 0.183 0.078
Chain 1: 7100 -8846.242 0.185 0.092
Chain 1: 7200 -11681.571 0.153 0.092
Chain 1: 7300 -9599.792 0.173 0.128
Chain 1: 7400 -8941.127 0.180 0.128
Chain 1: 7500 -10780.483 0.197 0.171
Chain 1: 7600 -8880.827 0.211 0.214
Chain 1: 7700 -10349.883 0.212 0.214
Chain 1: 7800 -11926.730 0.187 0.171
Chain 1: 7900 -8763.375 0.169 0.171
Chain 1: 8000 -12592.015 0.195 0.214
Chain 1: 8100 -10360.978 0.207 0.215
Chain 1: 8200 -10787.542 0.187 0.214
Chain 1: 8300 -11853.913 0.174 0.171
Chain 1: 8400 -8467.977 0.207 0.214
Chain 1: 8500 -10465.560 0.209 0.214
Chain 1: 8600 -10394.850 0.188 0.191
Chain 1: 8700 -9575.389 0.183 0.191
Chain 1: 8800 -8554.265 0.181 0.191
Chain 1: 8900 -14885.222 0.188 0.191
Chain 1: 9000 -12505.700 0.176 0.190
Chain 1: 9100 -8850.510 0.196 0.190
Chain 1: 9200 -9702.175 0.201 0.190
Chain 1: 9300 -12886.239 0.217 0.191
Chain 1: 9400 -8639.913 0.226 0.191
Chain 1: 9500 -12823.167 0.239 0.247
Chain 1: 9600 -8600.113 0.288 0.326
Chain 1: 9700 -8989.497 0.283 0.326
Chain 1: 9800 -10385.369 0.285 0.326
Chain 1: 9900 -8434.031 0.266 0.247
Chain 1: 10000 -8319.361 0.248 0.247
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001444 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.44 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57768.665 1.000 1.000
Chain 1: 200 -18116.608 1.594 2.189
Chain 1: 300 -9015.840 1.399 1.009
Chain 1: 400 -8187.373 1.075 1.009
Chain 1: 500 -8501.853 0.867 1.000
Chain 1: 600 -8506.476 0.723 1.000
Chain 1: 700 -8198.530 0.625 0.101
Chain 1: 800 -8379.321 0.549 0.101
Chain 1: 900 -8017.127 0.493 0.045
Chain 1: 1000 -7919.541 0.445 0.045
Chain 1: 1100 -7712.091 0.348 0.038
Chain 1: 1200 -7783.339 0.130 0.037
Chain 1: 1300 -7923.293 0.031 0.027
Chain 1: 1400 -7731.547 0.023 0.025
Chain 1: 1500 -7537.773 0.022 0.025
Chain 1: 1600 -7692.182 0.024 0.025
Chain 1: 1700 -7640.188 0.021 0.022
Chain 1: 1800 -7630.844 0.019 0.020
Chain 1: 1900 -7632.301 0.014 0.018
Chain 1: 2000 -7721.166 0.014 0.018
Chain 1: 2100 -7579.760 0.014 0.018
Chain 1: 2200 -7869.415 0.016 0.019
Chain 1: 2300 -7588.353 0.018 0.020
Chain 1: 2400 -7549.032 0.016 0.019
Chain 1: 2500 -7580.555 0.014 0.012
Chain 1: 2600 -7526.584 0.013 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003184 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.84 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86987.364 1.000 1.000
Chain 1: 200 -14107.490 3.083 5.166
Chain 1: 300 -10312.266 2.178 1.000
Chain 1: 400 -12417.035 1.676 1.000
Chain 1: 500 -8729.956 1.425 0.422
Chain 1: 600 -8843.493 1.190 0.422
Chain 1: 700 -9413.226 1.028 0.368
Chain 1: 800 -8516.798 0.913 0.368
Chain 1: 900 -8539.290 0.812 0.170
Chain 1: 1000 -9208.235 0.738 0.170
Chain 1: 1100 -8844.723 0.642 0.105 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8984.001 0.127 0.073
Chain 1: 1300 -8556.150 0.095 0.061
Chain 1: 1400 -8747.068 0.080 0.050
Chain 1: 1500 -8640.350 0.039 0.041
Chain 1: 1600 -8646.427 0.038 0.041
Chain 1: 1700 -8520.067 0.034 0.022
Chain 1: 1800 -8575.933 0.024 0.016
Chain 1: 1900 -8598.313 0.024 0.016
Chain 1: 2000 -8684.809 0.018 0.015
Chain 1: 2100 -8522.693 0.015 0.015
Chain 1: 2200 -8519.869 0.014 0.012
Chain 1: 2300 -8663.347 0.010 0.012
Chain 1: 2400 -8452.751 0.011 0.012
Chain 1: 2500 -8520.784 0.010 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002585 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.85 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8410444.048 1.000 1.000
Chain 1: 200 -1586028.250 2.651 4.303
Chain 1: 300 -892155.056 2.027 1.000
Chain 1: 400 -458645.561 1.756 1.000
Chain 1: 500 -359116.552 1.461 0.945
Chain 1: 600 -233738.075 1.307 0.945
Chain 1: 700 -119892.442 1.256 0.945
Chain 1: 800 -87140.789 1.146 0.945
Chain 1: 900 -67473.391 1.051 0.778
Chain 1: 1000 -52281.513 0.975 0.778
Chain 1: 1100 -39760.886 0.906 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38944.888 0.478 0.376
Chain 1: 1300 -26876.707 0.445 0.376
Chain 1: 1400 -26598.146 0.352 0.315
Chain 1: 1500 -23179.650 0.339 0.315
Chain 1: 1600 -22396.629 0.289 0.291
Chain 1: 1700 -21265.697 0.199 0.291
Chain 1: 1800 -21209.526 0.162 0.147
Chain 1: 1900 -21536.693 0.134 0.053
Chain 1: 2000 -20044.174 0.112 0.053
Chain 1: 2100 -20282.479 0.082 0.035
Chain 1: 2200 -20510.242 0.081 0.035
Chain 1: 2300 -20126.099 0.038 0.019
Chain 1: 2400 -19897.829 0.038 0.019
Chain 1: 2500 -19700.169 0.024 0.015
Chain 1: 2600 -19329.060 0.023 0.015
Chain 1: 2700 -19285.696 0.018 0.012
Chain 1: 2800 -19002.306 0.019 0.015
Chain 1: 2900 -19284.015 0.019 0.015
Chain 1: 3000 -19269.978 0.012 0.012
Chain 1: 3100 -19355.166 0.011 0.011
Chain 1: 3200 -19045.163 0.011 0.015
Chain 1: 3300 -19250.465 0.010 0.011
Chain 1: 3400 -18724.272 0.012 0.015
Chain 1: 3500 -19337.920 0.014 0.015
Chain 1: 3600 -18642.298 0.016 0.015
Chain 1: 3700 -19030.839 0.018 0.016
Chain 1: 3800 -17987.076 0.022 0.020
Chain 1: 3900 -17983.196 0.021 0.020
Chain 1: 4000 -18100.437 0.021 0.020
Chain 1: 4100 -18014.088 0.021 0.020
Chain 1: 4200 -17829.561 0.021 0.020
Chain 1: 4300 -17968.444 0.021 0.020
Chain 1: 4400 -17924.629 0.018 0.010
Chain 1: 4500 -17827.108 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001249 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.49 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12631.196 1.000 1.000
Chain 1: 200 -9367.664 0.674 1.000
Chain 1: 300 -8239.897 0.495 0.348
Chain 1: 400 -8392.623 0.376 0.348
Chain 1: 500 -8337.330 0.302 0.137
Chain 1: 600 -8148.394 0.256 0.137
Chain 1: 700 -8224.465 0.220 0.023
Chain 1: 800 -8124.464 0.194 0.023
Chain 1: 900 -7963.999 0.175 0.020
Chain 1: 1000 -8170.159 0.160 0.023
Chain 1: 1100 -8113.903 0.061 0.020
Chain 1: 1200 -8072.048 0.026 0.018
Chain 1: 1300 -8037.562 0.013 0.012
Chain 1: 1400 -8044.262 0.011 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001374 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.74 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -55788.381 1.000 1.000
Chain 1: 200 -17454.364 1.598 2.196
Chain 1: 300 -8800.443 1.393 1.000
Chain 1: 400 -8539.376 1.053 1.000
Chain 1: 500 -8131.264 0.852 0.983
Chain 1: 600 -8598.828 0.719 0.983
Chain 1: 700 -8404.565 0.620 0.054
Chain 1: 800 -8200.541 0.545 0.054
Chain 1: 900 -8216.046 0.485 0.050
Chain 1: 1000 -7753.422 0.442 0.054
Chain 1: 1100 -7701.555 0.343 0.050
Chain 1: 1200 -7790.452 0.125 0.031
Chain 1: 1300 -7698.784 0.027 0.025
Chain 1: 1400 -7762.795 0.025 0.023
Chain 1: 1500 -7589.929 0.022 0.023
Chain 1: 1600 -7746.169 0.019 0.020
Chain 1: 1700 -7670.791 0.018 0.012
Chain 1: 1800 -7665.604 0.015 0.011
Chain 1: 1900 -7560.838 0.017 0.012
Chain 1: 2000 -7663.274 0.012 0.012
Chain 1: 2100 -7580.345 0.012 0.012
Chain 1: 2200 -7773.237 0.014 0.013
Chain 1: 2300 -7552.466 0.015 0.014
Chain 1: 2400 -7675.500 0.016 0.016
Chain 1: 2500 -7585.393 0.015 0.014
Chain 1: 2600 -7536.107 0.014 0.013
Chain 1: 2700 -7534.212 0.013 0.013
Chain 1: 2800 -7533.119 0.013 0.013
Chain 1: 2900 -7443.928 0.013 0.012
Chain 1: 3000 -7554.452 0.013 0.012
Chain 1: 3100 -7541.474 0.012 0.012
Chain 1: 3200 -7739.365 0.012 0.012
Chain 1: 3300 -7462.103 0.013 0.012
Chain 1: 3400 -7683.847 0.014 0.012
Chain 1: 3500 -7446.208 0.016 0.015
Chain 1: 3600 -7512.184 0.016 0.015
Chain 1: 3700 -7461.022 0.017 0.015
Chain 1: 3800 -7459.759 0.017 0.015
Chain 1: 3900 -7427.068 0.016 0.015
Chain 1: 4000 -7422.048 0.015 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002523 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.23 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86409.035 1.000 1.000
Chain 1: 200 -13773.699 3.137 5.273
Chain 1: 300 -10093.186 2.213 1.000
Chain 1: 400 -11178.047 1.684 1.000
Chain 1: 500 -8972.384 1.396 0.365
Chain 1: 600 -9252.058 1.169 0.365
Chain 1: 700 -9364.585 1.003 0.246
Chain 1: 800 -8537.070 0.890 0.246
Chain 1: 900 -8471.534 0.792 0.097
Chain 1: 1000 -8774.022 0.716 0.097
Chain 1: 1100 -8924.237 0.618 0.097 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8452.469 0.096 0.056
Chain 1: 1300 -8728.579 0.063 0.034
Chain 1: 1400 -8726.218 0.053 0.032
Chain 1: 1500 -8633.063 0.030 0.030
Chain 1: 1600 -8731.240 0.028 0.017
Chain 1: 1700 -8810.071 0.027 0.017
Chain 1: 1800 -8377.405 0.023 0.017
Chain 1: 1900 -8481.385 0.023 0.017
Chain 1: 2000 -8456.839 0.020 0.012
Chain 1: 2100 -8592.619 0.020 0.012
Chain 1: 2200 -8386.135 0.017 0.012
Chain 1: 2300 -8482.596 0.015 0.011
Chain 1: 2400 -8545.917 0.016 0.011
Chain 1: 2500 -8490.146 0.015 0.011
Chain 1: 2600 -8494.321 0.014 0.011
Chain 1: 2700 -8409.607 0.014 0.011
Chain 1: 2800 -8366.272 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003183 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.83 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8426640.312 1.000 1.000
Chain 1: 200 -1590527.788 2.649 4.298
Chain 1: 300 -891964.147 2.027 1.000
Chain 1: 400 -458061.241 1.757 1.000
Chain 1: 500 -357712.666 1.462 0.947
Chain 1: 600 -232566.687 1.308 0.947
Chain 1: 700 -119099.449 1.257 0.947
Chain 1: 800 -86406.627 1.147 0.947
Chain 1: 900 -66830.236 1.052 0.783
Chain 1: 1000 -51702.239 0.976 0.783
Chain 1: 1100 -39245.896 0.908 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38433.316 0.480 0.378
Chain 1: 1300 -26449.301 0.447 0.378
Chain 1: 1400 -26176.068 0.354 0.317
Chain 1: 1500 -22778.472 0.341 0.317
Chain 1: 1600 -22000.096 0.290 0.293
Chain 1: 1700 -20880.539 0.200 0.293
Chain 1: 1800 -20826.615 0.163 0.149
Chain 1: 1900 -21152.980 0.135 0.054
Chain 1: 2000 -19667.257 0.113 0.054
Chain 1: 2100 -19905.507 0.083 0.035
Chain 1: 2200 -20131.537 0.082 0.035
Chain 1: 2300 -19749.066 0.038 0.019
Chain 1: 2400 -19521.113 0.039 0.019
Chain 1: 2500 -19322.914 0.025 0.015
Chain 1: 2600 -18953.080 0.023 0.015
Chain 1: 2700 -18910.113 0.018 0.012
Chain 1: 2800 -18626.718 0.019 0.015
Chain 1: 2900 -18908.002 0.019 0.015
Chain 1: 3000 -18894.272 0.012 0.012
Chain 1: 3100 -18979.253 0.011 0.012
Chain 1: 3200 -18669.861 0.011 0.015
Chain 1: 3300 -18874.672 0.011 0.012
Chain 1: 3400 -18349.296 0.012 0.015
Chain 1: 3500 -18961.507 0.015 0.015
Chain 1: 3600 -18267.759 0.016 0.015
Chain 1: 3700 -18654.757 0.018 0.017
Chain 1: 3800 -17613.773 0.023 0.021
Chain 1: 3900 -17609.869 0.021 0.021
Chain 1: 4000 -17727.214 0.022 0.021
Chain 1: 4100 -17640.878 0.022 0.021
Chain 1: 4200 -17457.021 0.021 0.021
Chain 1: 4300 -17595.525 0.021 0.021
Chain 1: 4400 -17552.196 0.018 0.011
Chain 1: 4500 -17454.691 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001294 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.94 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49208.804 1.000 1.000
Chain 1: 200 -14773.858 1.665 2.331
Chain 1: 300 -19812.281 1.195 1.000
Chain 1: 400 -18307.288 0.917 1.000
Chain 1: 500 -14255.949 0.790 0.284
Chain 1: 600 -14930.102 0.666 0.284
Chain 1: 700 -17983.052 0.595 0.254
Chain 1: 800 -15150.964 0.544 0.254
Chain 1: 900 -12130.398 0.511 0.249
Chain 1: 1000 -30154.586 0.520 0.254
Chain 1: 1100 -18923.314 0.479 0.254
Chain 1: 1200 -10895.104 0.320 0.254
Chain 1: 1300 -11734.495 0.302 0.249
Chain 1: 1400 -17680.849 0.327 0.284
Chain 1: 1500 -12077.641 0.345 0.336
Chain 1: 1600 -12709.919 0.346 0.336
Chain 1: 1700 -19847.813 0.365 0.360
Chain 1: 1800 -13018.544 0.398 0.464
Chain 1: 1900 -11876.180 0.383 0.464
Chain 1: 2000 -19001.813 0.361 0.375
Chain 1: 2100 -12890.732 0.349 0.375
Chain 1: 2200 -11108.776 0.291 0.360
Chain 1: 2300 -11189.397 0.285 0.360
Chain 1: 2400 -9209.996 0.273 0.360
Chain 1: 2500 -10041.677 0.234 0.215
Chain 1: 2600 -9726.200 0.233 0.215
Chain 1: 2700 -9377.687 0.200 0.160
Chain 1: 2800 -10506.338 0.159 0.107
Chain 1: 2900 -9670.948 0.158 0.107
Chain 1: 3000 -11124.330 0.133 0.107
Chain 1: 3100 -9006.715 0.109 0.107
Chain 1: 3200 -9972.724 0.103 0.097
Chain 1: 3300 -9760.425 0.105 0.097
Chain 1: 3400 -9479.051 0.086 0.086
Chain 1: 3500 -9542.523 0.078 0.086
Chain 1: 3600 -14313.506 0.109 0.097
Chain 1: 3700 -19587.918 0.132 0.107
Chain 1: 3800 -9268.363 0.232 0.131
Chain 1: 3900 -9996.947 0.231 0.131
Chain 1: 4000 -9088.630 0.228 0.100
Chain 1: 4100 -8899.958 0.206 0.097
Chain 1: 4200 -11104.974 0.217 0.100
Chain 1: 4300 -9111.894 0.236 0.199
Chain 1: 4400 -9126.006 0.234 0.199
Chain 1: 4500 -10117.825 0.243 0.199
Chain 1: 4600 -8810.448 0.224 0.148
Chain 1: 4700 -9964.194 0.209 0.116
Chain 1: 4800 -8762.117 0.111 0.116
Chain 1: 4900 -8859.013 0.105 0.116
Chain 1: 5000 -9829.982 0.105 0.116
Chain 1: 5100 -8883.110 0.113 0.116
Chain 1: 5200 -14619.217 0.133 0.116
Chain 1: 5300 -13000.750 0.123 0.116
Chain 1: 5400 -9311.527 0.163 0.124
Chain 1: 5500 -9712.298 0.157 0.124
Chain 1: 5600 -14930.833 0.177 0.124
Chain 1: 5700 -10981.862 0.202 0.137
Chain 1: 5800 -8835.520 0.212 0.243
Chain 1: 5900 -9506.415 0.218 0.243
Chain 1: 6000 -11844.977 0.228 0.243
Chain 1: 6100 -9221.971 0.246 0.284
Chain 1: 6200 -9028.408 0.209 0.243
Chain 1: 6300 -9093.071 0.197 0.243
Chain 1: 6400 -9501.057 0.162 0.197
Chain 1: 6500 -9729.323 0.160 0.197
Chain 1: 6600 -10496.141 0.132 0.073
Chain 1: 6700 -9375.004 0.108 0.073
Chain 1: 6800 -10072.071 0.091 0.071
Chain 1: 6900 -12014.326 0.100 0.073
Chain 1: 7000 -8931.978 0.115 0.073
Chain 1: 7100 -10996.549 0.105 0.073
Chain 1: 7200 -8860.621 0.127 0.120
Chain 1: 7300 -9740.313 0.135 0.120
Chain 1: 7400 -8490.622 0.146 0.147
Chain 1: 7500 -9876.465 0.158 0.147
Chain 1: 7600 -8787.050 0.163 0.147
Chain 1: 7700 -8809.551 0.151 0.147
Chain 1: 7800 -8605.729 0.146 0.147
Chain 1: 7900 -9472.729 0.139 0.140
Chain 1: 8000 -10091.640 0.111 0.124
Chain 1: 8100 -8462.831 0.111 0.124
Chain 1: 8200 -9302.940 0.096 0.092
Chain 1: 8300 -8497.284 0.097 0.095
Chain 1: 8400 -9028.020 0.088 0.092
Chain 1: 8500 -9466.140 0.079 0.090
Chain 1: 8600 -10834.362 0.079 0.090
Chain 1: 8700 -8482.055 0.106 0.092
Chain 1: 8800 -9557.873 0.115 0.095
Chain 1: 8900 -8536.665 0.118 0.113
Chain 1: 9000 -8857.598 0.115 0.113
Chain 1: 9100 -8767.296 0.097 0.095
Chain 1: 9200 -9015.148 0.091 0.095
Chain 1: 9300 -11369.228 0.102 0.113
Chain 1: 9400 -9856.922 0.112 0.120
Chain 1: 9500 -9028.395 0.116 0.120
Chain 1: 9600 -10982.424 0.121 0.120
Chain 1: 9700 -10483.247 0.098 0.113
Chain 1: 9800 -9926.092 0.093 0.092
Chain 1: 9900 -8838.187 0.093 0.092
Chain 1: 10000 -9389.485 0.095 0.092
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00143 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.3 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58346.508 1.000 1.000
Chain 1: 200 -17761.581 1.642 2.285
Chain 1: 300 -8766.474 1.437 1.026
Chain 1: 400 -8109.428 1.098 1.026
Chain 1: 500 -8786.782 0.894 1.000
Chain 1: 600 -8741.660 0.746 1.000
Chain 1: 700 -7837.359 0.656 0.115
Chain 1: 800 -8106.443 0.578 0.115
Chain 1: 900 -8036.761 0.515 0.081
Chain 1: 1000 -7814.310 0.466 0.081
Chain 1: 1100 -7780.855 0.366 0.077
Chain 1: 1200 -7708.175 0.139 0.033
Chain 1: 1300 -7709.851 0.036 0.028
Chain 1: 1400 -7806.249 0.029 0.012
Chain 1: 1500 -7624.649 0.024 0.012
Chain 1: 1600 -7936.911 0.028 0.024
Chain 1: 1700 -7592.712 0.021 0.024
Chain 1: 1800 -7691.568 0.018 0.013
Chain 1: 1900 -7577.597 0.019 0.015
Chain 1: 2000 -7667.254 0.017 0.013
Chain 1: 2100 -7631.918 0.017 0.013
Chain 1: 2200 -7759.931 0.018 0.015
Chain 1: 2300 -7702.302 0.019 0.015
Chain 1: 2400 -7679.053 0.018 0.015
Chain 1: 2500 -7613.565 0.016 0.013
Chain 1: 2600 -7560.514 0.013 0.012
Chain 1: 2700 -7563.698 0.009 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002555 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.55 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86682.621 1.000 1.000
Chain 1: 200 -13699.322 3.164 5.328
Chain 1: 300 -10104.125 2.228 1.000
Chain 1: 400 -10785.345 1.687 1.000
Chain 1: 500 -9042.414 1.388 0.356
Chain 1: 600 -8573.713 1.166 0.356
Chain 1: 700 -8772.548 1.002 0.193
Chain 1: 800 -8844.552 0.878 0.193
Chain 1: 900 -8959.625 0.782 0.063
Chain 1: 1000 -8663.516 0.707 0.063
Chain 1: 1100 -8995.407 0.611 0.055 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8612.133 0.083 0.045
Chain 1: 1300 -8728.008 0.048 0.037
Chain 1: 1400 -8821.430 0.043 0.034
Chain 1: 1500 -8686.307 0.025 0.023
Chain 1: 1600 -8792.664 0.021 0.016
Chain 1: 1700 -8888.561 0.020 0.013
Chain 1: 1800 -8483.387 0.024 0.016
Chain 1: 1900 -8581.807 0.024 0.016
Chain 1: 2000 -8553.244 0.021 0.013
Chain 1: 2100 -8673.058 0.018 0.013
Chain 1: 2200 -8478.597 0.016 0.013
Chain 1: 2300 -8616.605 0.016 0.014
Chain 1: 2400 -8627.494 0.016 0.014
Chain 1: 2500 -8593.966 0.014 0.012
Chain 1: 2600 -8595.957 0.013 0.011
Chain 1: 2700 -8504.994 0.013 0.011
Chain 1: 2800 -8472.776 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003228 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.28 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8414082.089 1.000 1.000
Chain 1: 200 -1587117.655 2.651 4.301
Chain 1: 300 -890977.890 2.028 1.000
Chain 1: 400 -457312.780 1.758 1.000
Chain 1: 500 -357275.750 1.462 0.948
Chain 1: 600 -232292.986 1.308 0.948
Chain 1: 700 -118978.251 1.257 0.948
Chain 1: 800 -86278.267 1.148 0.948
Chain 1: 900 -66714.567 1.053 0.781
Chain 1: 1000 -51584.230 0.977 0.781
Chain 1: 1100 -39126.211 0.909 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38311.793 0.481 0.379
Chain 1: 1300 -26337.518 0.448 0.379
Chain 1: 1400 -26062.594 0.354 0.318
Chain 1: 1500 -22667.183 0.341 0.318
Chain 1: 1600 -21888.733 0.291 0.293
Chain 1: 1700 -20771.058 0.201 0.293
Chain 1: 1800 -20717.214 0.163 0.150
Chain 1: 1900 -21043.169 0.136 0.054
Chain 1: 2000 -19559.062 0.114 0.054
Chain 1: 2100 -19797.256 0.083 0.036
Chain 1: 2200 -20022.776 0.082 0.036
Chain 1: 2300 -19640.854 0.039 0.019
Chain 1: 2400 -19413.137 0.039 0.019
Chain 1: 2500 -19214.762 0.025 0.015
Chain 1: 2600 -18845.568 0.023 0.015
Chain 1: 2700 -18802.763 0.018 0.012
Chain 1: 2800 -18519.544 0.019 0.015
Chain 1: 2900 -18800.595 0.019 0.015
Chain 1: 3000 -18786.906 0.012 0.012
Chain 1: 3100 -18871.830 0.011 0.012
Chain 1: 3200 -18562.752 0.012 0.015
Chain 1: 3300 -18767.302 0.011 0.012
Chain 1: 3400 -18242.501 0.012 0.015
Chain 1: 3500 -18853.812 0.015 0.015
Chain 1: 3600 -18161.236 0.016 0.015
Chain 1: 3700 -18547.428 0.018 0.017
Chain 1: 3800 -17508.174 0.023 0.021
Chain 1: 3900 -17504.303 0.021 0.021
Chain 1: 4000 -17621.659 0.022 0.021
Chain 1: 4100 -17535.427 0.022 0.021
Chain 1: 4200 -17351.950 0.021 0.021
Chain 1: 4300 -17490.203 0.021 0.021
Chain 1: 4400 -17447.229 0.018 0.011
Chain 1: 4500 -17349.762 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001227 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.27 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48365.929 1.000 1.000
Chain 1: 200 -20437.036 1.183 1.367
Chain 1: 300 -12463.922 1.002 1.000
Chain 1: 400 -13296.564 0.767 1.000
Chain 1: 500 -12549.385 0.626 0.640
Chain 1: 600 -20020.653 0.584 0.640
Chain 1: 700 -11416.958 0.608 0.640
Chain 1: 800 -10550.273 0.542 0.640
Chain 1: 900 -12602.857 0.500 0.373
Chain 1: 1000 -11206.259 0.462 0.373
Chain 1: 1100 -10430.115 0.370 0.163
Chain 1: 1200 -11549.623 0.243 0.125
Chain 1: 1300 -12923.096 0.190 0.106
Chain 1: 1400 -10215.538 0.210 0.125
Chain 1: 1500 -17945.337 0.247 0.163
Chain 1: 1600 -9895.979 0.291 0.163
Chain 1: 1700 -25124.443 0.276 0.163
Chain 1: 1800 -15704.449 0.328 0.265
Chain 1: 1900 -10575.382 0.360 0.431
Chain 1: 2000 -21157.139 0.398 0.485
Chain 1: 2100 -9505.440 0.513 0.500 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2200 -11275.638 0.519 0.500 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2300 -11734.966 0.512 0.500 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2400 -9855.807 0.505 0.500 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2500 -17742.189 0.506 0.500 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2600 -9596.960 0.510 0.500 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2700 -10782.596 0.460 0.485
Chain 1: 2800 -14615.360 0.426 0.444
Chain 1: 2900 -9264.871 0.436 0.444
Chain 1: 3000 -8499.199 0.395 0.262
Chain 1: 3100 -9109.653 0.279 0.191
Chain 1: 3200 -8696.762 0.268 0.191
Chain 1: 3300 -9719.003 0.274 0.191
Chain 1: 3400 -8443.636 0.270 0.151
Chain 1: 3500 -10196.545 0.243 0.151
Chain 1: 3600 -11963.434 0.173 0.148
Chain 1: 3700 -9184.100 0.192 0.151
Chain 1: 3800 -8525.260 0.174 0.148
Chain 1: 3900 -8565.674 0.117 0.105
Chain 1: 4000 -14317.095 0.148 0.148
Chain 1: 4100 -10917.693 0.172 0.151
Chain 1: 4200 -13977.716 0.189 0.172
Chain 1: 4300 -9649.151 0.224 0.219
Chain 1: 4400 -8891.756 0.217 0.219
Chain 1: 4500 -8686.112 0.202 0.219
Chain 1: 4600 -11350.900 0.211 0.235
Chain 1: 4700 -12200.974 0.188 0.219
Chain 1: 4800 -8214.176 0.228 0.235
Chain 1: 4900 -8354.358 0.230 0.235
Chain 1: 5000 -8353.315 0.189 0.219
Chain 1: 5100 -8670.433 0.162 0.085
Chain 1: 5200 -8554.712 0.141 0.070
Chain 1: 5300 -9739.885 0.109 0.070
Chain 1: 5400 -9709.834 0.101 0.037
Chain 1: 5500 -8253.710 0.116 0.070
Chain 1: 5600 -12361.674 0.126 0.070
Chain 1: 5700 -12157.880 0.120 0.037
Chain 1: 5800 -8296.408 0.118 0.037
Chain 1: 5900 -11800.667 0.146 0.122
Chain 1: 6000 -9161.169 0.175 0.176
Chain 1: 6100 -8514.398 0.179 0.176
Chain 1: 6200 -8361.228 0.180 0.176
Chain 1: 6300 -13034.291 0.203 0.288
Chain 1: 6400 -8601.069 0.254 0.297
Chain 1: 6500 -8213.960 0.241 0.297
Chain 1: 6600 -9674.898 0.223 0.288
Chain 1: 6700 -8718.694 0.233 0.288
Chain 1: 6800 -12183.691 0.215 0.284
Chain 1: 6900 -8872.873 0.222 0.284
Chain 1: 7000 -9321.085 0.198 0.151
Chain 1: 7100 -7974.132 0.207 0.169
Chain 1: 7200 -10753.960 0.231 0.258
Chain 1: 7300 -8316.615 0.225 0.258
Chain 1: 7400 -10285.591 0.193 0.191
Chain 1: 7500 -9999.222 0.191 0.191
Chain 1: 7600 -8297.246 0.196 0.205
Chain 1: 7700 -8147.340 0.187 0.205
Chain 1: 7800 -8556.289 0.163 0.191
Chain 1: 7900 -10464.016 0.144 0.182
Chain 1: 8000 -11624.109 0.149 0.182
Chain 1: 8100 -8044.704 0.177 0.191
Chain 1: 8200 -7719.013 0.155 0.182
Chain 1: 8300 -11979.194 0.162 0.182
Chain 1: 8400 -11239.608 0.149 0.100
Chain 1: 8500 -8439.183 0.179 0.182
Chain 1: 8600 -9646.694 0.171 0.125
Chain 1: 8700 -8488.458 0.183 0.136
Chain 1: 8800 -9103.394 0.185 0.136
Chain 1: 8900 -8966.599 0.168 0.125
Chain 1: 9000 -10360.674 0.172 0.135
Chain 1: 9100 -9479.008 0.137 0.125
Chain 1: 9200 -8737.411 0.141 0.125
Chain 1: 9300 -7928.210 0.116 0.102
Chain 1: 9400 -8599.146 0.117 0.102
Chain 1: 9500 -8368.398 0.086 0.093
Chain 1: 9600 -8245.830 0.075 0.085
Chain 1: 9700 -9116.863 0.071 0.085
Chain 1: 9800 -8575.971 0.071 0.085
Chain 1: 9900 -8465.037 0.071 0.085
Chain 1: 10000 -8444.781 0.057 0.078
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001751 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 17.51 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56421.156 1.000 1.000
Chain 1: 200 -16922.512 1.667 2.334
Chain 1: 300 -8499.302 1.442 1.000
Chain 1: 400 -8630.341 1.085 1.000
Chain 1: 500 -7972.172 0.885 0.991
Chain 1: 600 -8324.617 0.744 0.991
Chain 1: 700 -7815.652 0.647 0.083
Chain 1: 800 -8223.833 0.572 0.083
Chain 1: 900 -7816.165 0.515 0.065
Chain 1: 1000 -7643.091 0.465 0.065
Chain 1: 1100 -7602.833 0.366 0.052
Chain 1: 1200 -7703.118 0.134 0.050
Chain 1: 1300 -7622.927 0.036 0.042
Chain 1: 1400 -7612.324 0.034 0.042
Chain 1: 1500 -7551.914 0.027 0.023
Chain 1: 1600 -7485.312 0.024 0.013
Chain 1: 1700 -7469.113 0.017 0.011
Chain 1: 1800 -7487.663 0.013 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003652 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 36.52 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86229.043 1.000 1.000
Chain 1: 200 -12991.409 3.319 5.637
Chain 1: 300 -9450.211 2.337 1.000
Chain 1: 400 -10447.624 1.777 1.000
Chain 1: 500 -8322.376 1.473 0.375
Chain 1: 600 -8032.289 1.233 0.375
Chain 1: 700 -8325.048 1.062 0.255
Chain 1: 800 -8554.631 0.933 0.255
Chain 1: 900 -8352.881 0.832 0.095
Chain 1: 1000 -8074.963 0.752 0.095
Chain 1: 1100 -8336.353 0.655 0.036 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8109.867 0.094 0.035
Chain 1: 1300 -8197.990 0.058 0.034
Chain 1: 1400 -8196.535 0.048 0.031
Chain 1: 1500 -8107.111 0.024 0.028
Chain 1: 1600 -8201.369 0.021 0.027
Chain 1: 1700 -8298.883 0.019 0.024
Chain 1: 1800 -7909.288 0.021 0.024
Chain 1: 1900 -8010.462 0.020 0.013
Chain 1: 2000 -7980.409 0.017 0.012
Chain 1: 2100 -8119.215 0.016 0.012
Chain 1: 2200 -7900.704 0.016 0.012
Chain 1: 2300 -8042.892 0.016 0.013
Chain 1: 2400 -8045.573 0.016 0.013
Chain 1: 2500 -8019.308 0.015 0.013
Chain 1: 2600 -8016.037 0.014 0.013
Chain 1: 2700 -7925.432 0.014 0.013
Chain 1: 2800 -7905.895 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003677 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 36.77 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8432398.599 1.000 1.000
Chain 1: 200 -1590694.014 2.651 4.301
Chain 1: 300 -892092.603 2.028 1.000
Chain 1: 400 -457673.900 1.758 1.000
Chain 1: 500 -357565.115 1.463 0.949
Chain 1: 600 -232203.207 1.309 0.949
Chain 1: 700 -118508.715 1.259 0.949
Chain 1: 800 -85752.246 1.149 0.949
Chain 1: 900 -66119.967 1.055 0.783
Chain 1: 1000 -50939.643 0.979 0.783
Chain 1: 1100 -38445.525 0.911 0.540 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37618.234 0.484 0.382
Chain 1: 1300 -25617.959 0.452 0.382
Chain 1: 1400 -25337.151 0.358 0.325
Chain 1: 1500 -21936.912 0.346 0.325
Chain 1: 1600 -21155.893 0.295 0.298
Chain 1: 1700 -20035.843 0.205 0.297
Chain 1: 1800 -19980.920 0.167 0.155
Chain 1: 1900 -20306.378 0.139 0.056
Chain 1: 2000 -18822.011 0.117 0.056
Chain 1: 2100 -19060.042 0.086 0.037
Chain 1: 2200 -19285.601 0.085 0.037
Chain 1: 2300 -18903.777 0.040 0.020
Chain 1: 2400 -18676.201 0.040 0.020
Chain 1: 2500 -18478.025 0.026 0.016
Chain 1: 2600 -18109.102 0.024 0.016
Chain 1: 2700 -18066.294 0.019 0.012
Chain 1: 2800 -17783.427 0.020 0.016
Chain 1: 2900 -18064.266 0.020 0.016
Chain 1: 3000 -18050.524 0.012 0.012
Chain 1: 3100 -18135.439 0.011 0.012
Chain 1: 3200 -17826.586 0.012 0.016
Chain 1: 3300 -18030.912 0.011 0.012
Chain 1: 3400 -17506.671 0.013 0.016
Chain 1: 3500 -18117.239 0.015 0.016
Chain 1: 3600 -17425.592 0.017 0.016
Chain 1: 3700 -17811.145 0.019 0.017
Chain 1: 3800 -16773.413 0.024 0.022
Chain 1: 3900 -16769.595 0.022 0.022
Chain 1: 4000 -16886.917 0.023 0.022
Chain 1: 4100 -16800.834 0.023 0.022
Chain 1: 4200 -16617.612 0.022 0.022
Chain 1: 4300 -16755.635 0.022 0.022
Chain 1: 4400 -16712.916 0.019 0.011
Chain 1: 4500 -16615.525 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001327 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.27 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49033.212 1.000 1.000
Chain 1: 200 -23441.377 1.046 1.092
Chain 1: 300 -15365.220 0.872 1.000
Chain 1: 400 -12542.305 0.711 1.000
Chain 1: 500 -18999.028 0.636 0.526
Chain 1: 600 -14683.905 0.579 0.526
Chain 1: 700 -19491.889 0.532 0.340
Chain 1: 800 -13110.412 0.526 0.487
Chain 1: 900 -18035.859 0.498 0.340
Chain 1: 1000 -30773.747 0.490 0.414
Chain 1: 1100 -10627.532 0.579 0.414 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -13639.569 0.492 0.340
Chain 1: 1300 -11704.973 0.456 0.294
Chain 1: 1400 -10613.285 0.444 0.294
Chain 1: 1500 -10625.821 0.410 0.273
Chain 1: 1600 -10575.931 0.381 0.247
Chain 1: 1700 -11290.535 0.363 0.221
Chain 1: 1800 -10082.473 0.326 0.165
Chain 1: 1900 -9968.307 0.300 0.120
Chain 1: 2000 -12750.431 0.280 0.120
Chain 1: 2100 -10784.479 0.109 0.120
Chain 1: 2200 -14368.979 0.112 0.120
Chain 1: 2300 -9798.215 0.142 0.120
Chain 1: 2400 -14651.214 0.165 0.182
Chain 1: 2500 -9885.096 0.213 0.218
Chain 1: 2600 -10044.817 0.214 0.218
Chain 1: 2700 -10316.217 0.210 0.218
Chain 1: 2800 -9378.162 0.208 0.218
Chain 1: 2900 -14178.699 0.241 0.249
Chain 1: 3000 -9290.283 0.272 0.331
Chain 1: 3100 -9852.438 0.259 0.331
Chain 1: 3200 -9787.109 0.235 0.331
Chain 1: 3300 -9296.731 0.194 0.100
Chain 1: 3400 -9222.013 0.161 0.057
Chain 1: 3500 -9262.237 0.114 0.053
Chain 1: 3600 -17723.264 0.160 0.057
Chain 1: 3700 -10210.436 0.231 0.100
Chain 1: 3800 -11354.466 0.231 0.101
Chain 1: 3900 -9048.575 0.222 0.101
Chain 1: 4000 -9003.052 0.170 0.057
Chain 1: 4100 -8916.113 0.166 0.053
Chain 1: 4200 -10092.870 0.177 0.101
Chain 1: 4300 -16198.889 0.209 0.117
Chain 1: 4400 -8888.962 0.290 0.255
Chain 1: 4500 -9294.583 0.294 0.255
Chain 1: 4600 -8690.218 0.254 0.117
Chain 1: 4700 -10308.201 0.196 0.117
Chain 1: 4800 -8872.767 0.202 0.157
Chain 1: 4900 -9643.519 0.184 0.117
Chain 1: 5000 -8637.778 0.195 0.117
Chain 1: 5100 -11227.994 0.217 0.157
Chain 1: 5200 -9067.917 0.230 0.162
Chain 1: 5300 -14470.925 0.229 0.162
Chain 1: 5400 -8975.555 0.208 0.162
Chain 1: 5500 -9776.217 0.212 0.162
Chain 1: 5600 -9496.143 0.208 0.162
Chain 1: 5700 -16322.069 0.234 0.231
Chain 1: 5800 -12834.599 0.245 0.238
Chain 1: 5900 -8556.670 0.287 0.272
Chain 1: 6000 -9000.815 0.281 0.272
Chain 1: 6100 -11966.037 0.282 0.272
Chain 1: 6200 -9020.897 0.291 0.326
Chain 1: 6300 -11278.158 0.274 0.272
Chain 1: 6400 -9548.553 0.231 0.248
Chain 1: 6500 -9665.639 0.224 0.248
Chain 1: 6600 -8717.374 0.232 0.248
Chain 1: 6700 -14240.770 0.229 0.248
Chain 1: 6800 -8963.707 0.260 0.248
Chain 1: 6900 -8499.634 0.216 0.200
Chain 1: 7000 -8379.786 0.212 0.200
Chain 1: 7100 -9935.524 0.203 0.181
Chain 1: 7200 -8755.793 0.184 0.157
Chain 1: 7300 -11525.439 0.188 0.157
Chain 1: 7400 -10208.679 0.183 0.135
Chain 1: 7500 -10558.957 0.185 0.135
Chain 1: 7600 -8954.493 0.192 0.157
Chain 1: 7700 -9194.875 0.156 0.135
Chain 1: 7800 -8888.887 0.100 0.129
Chain 1: 7900 -8656.057 0.097 0.129
Chain 1: 8000 -8318.639 0.100 0.129
Chain 1: 8100 -11372.572 0.111 0.129
Chain 1: 8200 -12168.478 0.104 0.065
Chain 1: 8300 -8449.244 0.124 0.065
Chain 1: 8400 -11503.413 0.138 0.065
Chain 1: 8500 -10736.195 0.142 0.071
Chain 1: 8600 -8254.297 0.154 0.071
Chain 1: 8700 -9350.694 0.163 0.117
Chain 1: 8800 -8419.068 0.171 0.117
Chain 1: 8900 -8540.356 0.169 0.117
Chain 1: 9000 -8274.415 0.169 0.117
Chain 1: 9100 -9375.628 0.153 0.117
Chain 1: 9200 -10301.016 0.156 0.117
Chain 1: 9300 -8394.834 0.135 0.117
Chain 1: 9400 -9138.129 0.116 0.111
Chain 1: 9500 -8451.346 0.117 0.111
Chain 1: 9600 -10078.486 0.103 0.111
Chain 1: 9700 -8296.956 0.113 0.111
Chain 1: 9800 -8304.834 0.102 0.090
Chain 1: 9900 -9709.328 0.115 0.117
Chain 1: 10000 -8818.712 0.122 0.117
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001506 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.06 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57569.800 1.000 1.000
Chain 1: 200 -17595.838 1.636 2.272
Chain 1: 300 -8838.349 1.421 1.000
Chain 1: 400 -8180.489 1.086 1.000
Chain 1: 500 -7676.407 0.882 0.991
Chain 1: 600 -8484.017 0.751 0.991
Chain 1: 700 -7804.377 0.656 0.095
Chain 1: 800 -8257.900 0.581 0.095
Chain 1: 900 -8001.279 0.520 0.087
Chain 1: 1000 -7823.609 0.470 0.087
Chain 1: 1100 -7621.976 0.373 0.080
Chain 1: 1200 -7819.067 0.148 0.066
Chain 1: 1300 -7889.746 0.050 0.055
Chain 1: 1400 -7669.839 0.045 0.032
Chain 1: 1500 -7567.922 0.039 0.029
Chain 1: 1600 -7891.654 0.034 0.029
Chain 1: 1700 -7484.343 0.031 0.029
Chain 1: 1800 -7672.169 0.028 0.026
Chain 1: 1900 -7679.515 0.025 0.025
Chain 1: 2000 -7685.929 0.022 0.025
Chain 1: 2100 -7593.746 0.021 0.024
Chain 1: 2200 -7741.776 0.020 0.019
Chain 1: 2300 -7622.112 0.021 0.019
Chain 1: 2400 -7689.356 0.019 0.016
Chain 1: 2500 -7660.506 0.018 0.016
Chain 1: 2600 -7546.485 0.016 0.015
Chain 1: 2700 -7649.512 0.011 0.013
Chain 1: 2800 -7536.274 0.010 0.013
Chain 1: 2900 -7440.358 0.012 0.013
Chain 1: 3000 -7560.088 0.013 0.015
Chain 1: 3100 -7558.234 0.012 0.015
Chain 1: 3200 -7748.927 0.013 0.015
Chain 1: 3300 -7485.181 0.014 0.015
Chain 1: 3400 -7693.632 0.016 0.015
Chain 1: 3500 -7463.002 0.019 0.016
Chain 1: 3600 -7527.562 0.018 0.016
Chain 1: 3700 -7477.039 0.018 0.016
Chain 1: 3800 -7480.289 0.016 0.016
Chain 1: 3900 -7445.887 0.015 0.016
Chain 1: 4000 -7440.892 0.014 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003089 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.89 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86170.851 1.000 1.000
Chain 1: 200 -13697.821 3.145 5.291
Chain 1: 300 -10041.530 2.218 1.000
Chain 1: 400 -10987.072 1.685 1.000
Chain 1: 500 -9025.240 1.392 0.364
Chain 1: 600 -8563.809 1.169 0.364
Chain 1: 700 -8751.849 1.005 0.217
Chain 1: 800 -9498.124 0.889 0.217
Chain 1: 900 -8794.805 0.799 0.086
Chain 1: 1000 -8734.944 0.720 0.086
Chain 1: 1100 -8722.010 0.620 0.080 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8473.979 0.094 0.079
Chain 1: 1300 -8710.501 0.060 0.054
Chain 1: 1400 -8739.596 0.052 0.029
Chain 1: 1500 -8583.101 0.032 0.027
Chain 1: 1600 -8697.448 0.028 0.021
Chain 1: 1700 -8773.332 0.027 0.018
Chain 1: 1800 -8349.589 0.024 0.018
Chain 1: 1900 -8450.850 0.017 0.013
Chain 1: 2000 -8425.392 0.017 0.013
Chain 1: 2100 -8551.162 0.018 0.015
Chain 1: 2200 -8353.308 0.017 0.015
Chain 1: 2300 -8445.711 0.016 0.013
Chain 1: 2400 -8514.383 0.016 0.013
Chain 1: 2500 -8460.638 0.015 0.012
Chain 1: 2600 -8462.168 0.014 0.011
Chain 1: 2700 -8378.816 0.014 0.011
Chain 1: 2800 -8338.465 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003354 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.54 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8387749.884 1.000 1.000
Chain 1: 200 -1578523.825 2.657 4.314
Chain 1: 300 -889111.467 2.030 1.000
Chain 1: 400 -456667.300 1.759 1.000
Chain 1: 500 -357575.299 1.463 0.947
Chain 1: 600 -232834.283 1.308 0.947
Chain 1: 700 -119311.760 1.257 0.947
Chain 1: 800 -86585.312 1.147 0.947
Chain 1: 900 -66956.310 1.052 0.775
Chain 1: 1000 -51770.528 0.976 0.775
Chain 1: 1100 -39261.505 0.908 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38443.509 0.479 0.378
Chain 1: 1300 -26400.447 0.447 0.378
Chain 1: 1400 -26121.818 0.354 0.319
Chain 1: 1500 -22708.585 0.341 0.319
Chain 1: 1600 -21926.082 0.291 0.293
Chain 1: 1700 -20798.866 0.201 0.293
Chain 1: 1800 -20743.347 0.164 0.150
Chain 1: 1900 -21069.662 0.136 0.054
Chain 1: 2000 -19580.301 0.114 0.054
Chain 1: 2100 -19818.685 0.083 0.036
Chain 1: 2200 -20045.329 0.082 0.036
Chain 1: 2300 -19662.328 0.039 0.019
Chain 1: 2400 -19434.342 0.039 0.019
Chain 1: 2500 -19236.467 0.025 0.015
Chain 1: 2600 -18866.411 0.023 0.015
Chain 1: 2700 -18823.384 0.018 0.012
Chain 1: 2800 -18540.207 0.019 0.015
Chain 1: 2900 -18821.496 0.019 0.015
Chain 1: 3000 -18807.665 0.012 0.012
Chain 1: 3100 -18892.697 0.011 0.012
Chain 1: 3200 -18583.272 0.012 0.015
Chain 1: 3300 -18788.111 0.011 0.012
Chain 1: 3400 -18262.899 0.012 0.015
Chain 1: 3500 -18875.033 0.015 0.015
Chain 1: 3600 -18181.336 0.016 0.015
Chain 1: 3700 -18568.400 0.018 0.017
Chain 1: 3800 -17527.643 0.023 0.021
Chain 1: 3900 -17523.803 0.021 0.021
Chain 1: 4000 -17641.063 0.022 0.021
Chain 1: 4100 -17554.800 0.022 0.021
Chain 1: 4200 -17371.005 0.021 0.021
Chain 1: 4300 -17509.444 0.021 0.021
Chain 1: 4400 -17466.164 0.018 0.011
Chain 1: 4500 -17368.703 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001324 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.24 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49169.608 1.000 1.000
Chain 1: 200 -20270.127 1.213 1.426
Chain 1: 300 -17189.230 0.868 1.000
Chain 1: 400 -18796.121 0.673 1.000
Chain 1: 500 -13386.262 0.619 0.404
Chain 1: 600 -14722.793 0.531 0.404
Chain 1: 700 -11195.491 0.500 0.315
Chain 1: 800 -10907.457 0.441 0.315
Chain 1: 900 -11217.198 0.395 0.179
Chain 1: 1000 -13126.944 0.370 0.179
Chain 1: 1100 -14818.920 0.281 0.145
Chain 1: 1200 -13438.203 0.149 0.114
Chain 1: 1300 -11287.070 0.150 0.114
Chain 1: 1400 -12667.533 0.153 0.114
Chain 1: 1500 -11163.063 0.126 0.114
Chain 1: 1600 -21709.806 0.165 0.135
Chain 1: 1700 -12289.279 0.210 0.135
Chain 1: 1800 -14242.573 0.221 0.137
Chain 1: 1900 -17434.537 0.237 0.145
Chain 1: 2000 -15850.062 0.232 0.137
Chain 1: 2100 -10418.986 0.273 0.183
Chain 1: 2200 -12329.345 0.278 0.183
Chain 1: 2300 -13046.991 0.265 0.155
Chain 1: 2400 -15280.572 0.268 0.155
Chain 1: 2500 -9241.048 0.320 0.183
Chain 1: 2600 -9212.014 0.272 0.155
Chain 1: 2700 -12041.817 0.219 0.155
Chain 1: 2800 -10202.166 0.223 0.180
Chain 1: 2900 -9326.255 0.214 0.155
Chain 1: 3000 -9669.473 0.208 0.155
Chain 1: 3100 -9568.568 0.157 0.146
Chain 1: 3200 -8937.900 0.148 0.094
Chain 1: 3300 -9573.573 0.150 0.094
Chain 1: 3400 -14070.417 0.167 0.094
Chain 1: 3500 -9394.279 0.151 0.094
Chain 1: 3600 -13453.861 0.181 0.180
Chain 1: 3700 -16857.989 0.178 0.180
Chain 1: 3800 -12770.183 0.192 0.202
Chain 1: 3900 -9483.449 0.217 0.302
Chain 1: 4000 -8739.224 0.222 0.302
Chain 1: 4100 -9001.601 0.224 0.302
Chain 1: 4200 -10499.015 0.231 0.302
Chain 1: 4300 -10215.770 0.227 0.302
Chain 1: 4400 -9088.637 0.208 0.202
Chain 1: 4500 -9159.255 0.159 0.143
Chain 1: 4600 -13989.494 0.163 0.143
Chain 1: 4700 -8618.496 0.205 0.143
Chain 1: 4800 -8900.487 0.176 0.124
Chain 1: 4900 -9474.702 0.148 0.085
Chain 1: 5000 -8753.209 0.147 0.082
Chain 1: 5100 -9714.331 0.154 0.099
Chain 1: 5200 -14773.074 0.174 0.099
Chain 1: 5300 -9009.871 0.236 0.124
Chain 1: 5400 -9738.293 0.231 0.099
Chain 1: 5500 -8639.512 0.243 0.127
Chain 1: 5600 -11325.570 0.232 0.127
Chain 1: 5700 -9405.214 0.190 0.127
Chain 1: 5800 -11217.384 0.203 0.162
Chain 1: 5900 -8862.364 0.223 0.204
Chain 1: 6000 -9560.050 0.222 0.204
Chain 1: 6100 -9606.507 0.213 0.204
Chain 1: 6200 -9133.499 0.184 0.162
Chain 1: 6300 -14334.157 0.156 0.162
Chain 1: 6400 -9788.823 0.195 0.204
Chain 1: 6500 -13290.820 0.209 0.237
Chain 1: 6600 -12237.331 0.194 0.204
Chain 1: 6700 -11174.179 0.183 0.162
Chain 1: 6800 -8489.431 0.198 0.263
Chain 1: 6900 -11475.327 0.198 0.260
Chain 1: 7000 -8462.620 0.226 0.263
Chain 1: 7100 -8342.671 0.227 0.263
Chain 1: 7200 -8457.230 0.223 0.263
Chain 1: 7300 -9852.337 0.201 0.260
Chain 1: 7400 -10192.722 0.158 0.142
Chain 1: 7500 -8760.795 0.148 0.142
Chain 1: 7600 -9043.976 0.143 0.142
Chain 1: 7700 -8760.952 0.136 0.142
Chain 1: 7800 -9561.473 0.113 0.084
Chain 1: 7900 -9138.125 0.092 0.046
Chain 1: 8000 -8470.830 0.064 0.046
Chain 1: 8100 -8624.415 0.064 0.046
Chain 1: 8200 -8271.530 0.067 0.046
Chain 1: 8300 -10102.839 0.071 0.046
Chain 1: 8400 -8722.289 0.084 0.079
Chain 1: 8500 -8587.881 0.069 0.046
Chain 1: 8600 -8981.730 0.070 0.046
Chain 1: 8700 -8882.802 0.068 0.046
Chain 1: 8800 -9541.587 0.066 0.046
Chain 1: 8900 -9137.108 0.066 0.044
Chain 1: 9000 -8699.292 0.063 0.044
Chain 1: 9100 -8514.711 0.064 0.044
Chain 1: 9200 -8625.367 0.061 0.044
Chain 1: 9300 -8702.064 0.044 0.044
Chain 1: 9400 -8948.947 0.031 0.028
Chain 1: 9500 -9184.735 0.032 0.028
Chain 1: 9600 -9484.280 0.030 0.028
Chain 1: 9700 -9233.549 0.032 0.028
Chain 1: 9800 -8963.252 0.028 0.028
Chain 1: 9900 -10478.860 0.038 0.028
Chain 1: 10000 -8293.307 0.059 0.028
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001423 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.23 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58626.136 1.000 1.000
Chain 1: 200 -17825.930 1.644 2.289
Chain 1: 300 -8655.509 1.449 1.059
Chain 1: 400 -8118.041 1.104 1.059
Chain 1: 500 -8027.922 0.885 1.000
Chain 1: 600 -8641.609 0.749 1.000
Chain 1: 700 -7738.929 0.659 0.117
Chain 1: 800 -8031.297 0.581 0.117
Chain 1: 900 -7786.388 0.520 0.071
Chain 1: 1000 -7835.980 0.469 0.071
Chain 1: 1100 -7656.726 0.371 0.066
Chain 1: 1200 -7806.471 0.144 0.036
Chain 1: 1300 -7587.515 0.041 0.031
Chain 1: 1400 -7831.195 0.038 0.031
Chain 1: 1500 -7454.936 0.041 0.031
Chain 1: 1600 -7694.304 0.037 0.031
Chain 1: 1700 -7453.908 0.029 0.031
Chain 1: 1800 -7498.572 0.026 0.031
Chain 1: 1900 -7513.518 0.023 0.029
Chain 1: 2000 -7562.829 0.023 0.029
Chain 1: 2100 -7500.747 0.022 0.029
Chain 1: 2200 -7627.912 0.021 0.029
Chain 1: 2300 -7472.031 0.021 0.021
Chain 1: 2400 -7569.825 0.019 0.017
Chain 1: 2500 -7538.209 0.014 0.013
Chain 1: 2600 -7443.256 0.012 0.013
Chain 1: 2700 -7465.026 0.009 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003347 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.47 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85789.564 1.000 1.000
Chain 1: 200 -13574.284 3.160 5.320
Chain 1: 300 -9958.147 2.228 1.000
Chain 1: 400 -10976.824 1.694 1.000
Chain 1: 500 -8809.497 1.404 0.363
Chain 1: 600 -8549.965 1.175 0.363
Chain 1: 700 -8700.241 1.010 0.246
Chain 1: 800 -9326.088 0.892 0.246
Chain 1: 900 -8798.628 0.800 0.093
Chain 1: 1000 -8564.784 0.722 0.093
Chain 1: 1100 -8847.707 0.626 0.067 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8451.013 0.098 0.060
Chain 1: 1300 -8656.366 0.064 0.047
Chain 1: 1400 -8668.361 0.055 0.032
Chain 1: 1500 -8520.546 0.032 0.030
Chain 1: 1600 -8633.518 0.031 0.027
Chain 1: 1700 -8718.266 0.030 0.027
Chain 1: 1800 -8307.957 0.028 0.027
Chain 1: 1900 -8403.652 0.023 0.024
Chain 1: 2000 -8376.583 0.021 0.017
Chain 1: 2100 -8498.424 0.019 0.014
Chain 1: 2200 -8320.502 0.016 0.014
Chain 1: 2300 -8399.909 0.015 0.013
Chain 1: 2400 -8468.412 0.016 0.013
Chain 1: 2500 -8413.922 0.015 0.011
Chain 1: 2600 -8412.624 0.013 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003244 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.44 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8393158.703 1.000 1.000
Chain 1: 200 -1582680.559 2.652 4.303
Chain 1: 300 -890759.304 2.027 1.000
Chain 1: 400 -458107.093 1.756 1.000
Chain 1: 500 -358769.680 1.460 0.944
Chain 1: 600 -233554.133 1.306 0.944
Chain 1: 700 -119521.789 1.256 0.944
Chain 1: 800 -86695.639 1.146 0.944
Chain 1: 900 -66978.432 1.052 0.777
Chain 1: 1000 -51739.911 0.976 0.777
Chain 1: 1100 -39187.678 0.908 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38357.580 0.480 0.379
Chain 1: 1300 -26285.773 0.448 0.379
Chain 1: 1400 -26000.950 0.355 0.320
Chain 1: 1500 -22582.153 0.342 0.320
Chain 1: 1600 -21796.733 0.292 0.295
Chain 1: 1700 -20667.109 0.202 0.294
Chain 1: 1800 -20610.599 0.165 0.151
Chain 1: 1900 -20936.593 0.137 0.055
Chain 1: 2000 -19446.425 0.115 0.055
Chain 1: 2100 -19684.694 0.084 0.036
Chain 1: 2200 -19911.490 0.083 0.036
Chain 1: 2300 -19528.462 0.039 0.020
Chain 1: 2400 -19300.557 0.039 0.020
Chain 1: 2500 -19102.790 0.025 0.016
Chain 1: 2600 -18732.835 0.023 0.016
Chain 1: 2700 -18689.746 0.018 0.012
Chain 1: 2800 -18406.742 0.019 0.015
Chain 1: 2900 -18688.021 0.019 0.015
Chain 1: 3000 -18674.081 0.012 0.012
Chain 1: 3100 -18759.110 0.011 0.012
Chain 1: 3200 -18449.789 0.012 0.015
Chain 1: 3300 -18654.514 0.011 0.012
Chain 1: 3400 -18129.542 0.012 0.015
Chain 1: 3500 -18741.337 0.015 0.015
Chain 1: 3600 -18048.119 0.017 0.015
Chain 1: 3700 -18434.878 0.018 0.017
Chain 1: 3800 -17394.820 0.023 0.021
Chain 1: 3900 -17391.011 0.021 0.021
Chain 1: 4000 -17508.271 0.022 0.021
Chain 1: 4100 -17422.076 0.022 0.021
Chain 1: 4200 -17238.363 0.021 0.021
Chain 1: 4300 -17376.696 0.021 0.021
Chain 1: 4400 -17333.543 0.018 0.011
Chain 1: 4500 -17236.131 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003908 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 39.08 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12870.010 1.000 1.000
Chain 1: 200 -9694.827 0.664 1.000
Chain 1: 300 -8311.072 0.498 0.328
Chain 1: 400 -8523.254 0.380 0.328
Chain 1: 500 -8506.551 0.304 0.166
Chain 1: 600 -8231.281 0.259 0.166
Chain 1: 700 -8121.362 0.224 0.033
Chain 1: 800 -8126.399 0.196 0.033
Chain 1: 900 -8182.343 0.175 0.025
Chain 1: 1000 -8206.104 0.158 0.025
Chain 1: 1100 -8223.663 0.058 0.014
Chain 1: 1200 -8149.472 0.026 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001822 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 18.22 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58843.695 1.000 1.000
Chain 1: 200 -18231.672 1.614 2.228
Chain 1: 300 -8935.899 1.423 1.040
Chain 1: 400 -8078.515 1.093 1.040
Chain 1: 500 -8745.052 0.890 1.000
Chain 1: 600 -8889.685 0.744 1.000
Chain 1: 700 -7994.235 0.654 0.112
Chain 1: 800 -8169.649 0.575 0.112
Chain 1: 900 -7907.759 0.515 0.106
Chain 1: 1000 -7935.471 0.464 0.106
Chain 1: 1100 -7805.536 0.365 0.076
Chain 1: 1200 -7944.543 0.144 0.033
Chain 1: 1300 -7732.647 0.043 0.027
Chain 1: 1400 -7634.912 0.034 0.021
Chain 1: 1500 -7535.450 0.027 0.017
Chain 1: 1600 -7748.564 0.029 0.021
Chain 1: 1700 -7590.912 0.019 0.021
Chain 1: 1800 -7588.459 0.017 0.017
Chain 1: 1900 -7586.961 0.014 0.017
Chain 1: 2000 -7709.391 0.015 0.017
Chain 1: 2100 -7536.859 0.016 0.017
Chain 1: 2200 -7775.565 0.017 0.021
Chain 1: 2300 -7531.960 0.018 0.021
Chain 1: 2400 -7555.014 0.017 0.021
Chain 1: 2500 -7590.374 0.016 0.021
Chain 1: 2600 -7518.153 0.014 0.016
Chain 1: 2700 -7491.126 0.012 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003654 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 36.54 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86953.357 1.000 1.000
Chain 1: 200 -14035.100 3.098 5.195
Chain 1: 300 -10286.493 2.187 1.000
Chain 1: 400 -11936.105 1.675 1.000
Chain 1: 500 -8828.682 1.410 0.364
Chain 1: 600 -8673.473 1.178 0.364
Chain 1: 700 -9063.397 1.016 0.352
Chain 1: 800 -9229.113 0.891 0.352
Chain 1: 900 -8991.538 0.795 0.138
Chain 1: 1000 -9244.552 0.718 0.138
Chain 1: 1100 -9028.471 0.621 0.043 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8585.243 0.106 0.043
Chain 1: 1300 -8901.926 0.073 0.036
Chain 1: 1400 -8772.756 0.061 0.027
Chain 1: 1500 -8761.338 0.026 0.026
Chain 1: 1600 -8861.532 0.025 0.026
Chain 1: 1700 -8916.747 0.022 0.024
Chain 1: 1800 -8462.861 0.025 0.026
Chain 1: 1900 -8573.175 0.024 0.024
Chain 1: 2000 -8573.195 0.021 0.015
Chain 1: 2100 -8729.789 0.021 0.015
Chain 1: 2200 -8470.898 0.018 0.015
Chain 1: 2300 -8649.687 0.017 0.015
Chain 1: 2400 -8470.027 0.018 0.018
Chain 1: 2500 -8546.094 0.018 0.018
Chain 1: 2600 -8457.485 0.018 0.018
Chain 1: 2700 -8490.345 0.018 0.018
Chain 1: 2800 -8442.232 0.013 0.013
Chain 1: 2900 -8553.304 0.013 0.013
Chain 1: 3000 -8493.906 0.014 0.013
Chain 1: 3100 -8434.451 0.013 0.010
Chain 1: 3200 -8407.578 0.010 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004427 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 44.27 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8443108.053 1.000 1.000
Chain 1: 200 -1587104.485 2.660 4.320
Chain 1: 300 -889946.170 2.034 1.000
Chain 1: 400 -457761.715 1.762 1.000
Chain 1: 500 -357703.079 1.465 0.944
Chain 1: 600 -232779.335 1.311 0.944
Chain 1: 700 -119366.303 1.259 0.944
Chain 1: 800 -86718.302 1.149 0.944
Chain 1: 900 -67135.459 1.054 0.783
Chain 1: 1000 -52004.877 0.977 0.783
Chain 1: 1100 -39550.764 0.909 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38738.970 0.479 0.376
Chain 1: 1300 -26745.249 0.445 0.376
Chain 1: 1400 -26472.861 0.352 0.315
Chain 1: 1500 -23073.761 0.339 0.315
Chain 1: 1600 -22295.970 0.289 0.292
Chain 1: 1700 -21174.314 0.199 0.291
Chain 1: 1800 -21120.241 0.161 0.147
Chain 1: 1900 -21447.134 0.134 0.053
Chain 1: 2000 -19959.864 0.112 0.053
Chain 1: 2100 -20197.939 0.082 0.035
Chain 1: 2200 -20424.675 0.081 0.035
Chain 1: 2300 -20041.484 0.038 0.019
Chain 1: 2400 -19813.364 0.038 0.019
Chain 1: 2500 -19615.356 0.024 0.015
Chain 1: 2600 -19244.805 0.023 0.015
Chain 1: 2700 -19201.627 0.018 0.012
Chain 1: 2800 -18918.171 0.019 0.015
Chain 1: 2900 -19199.668 0.019 0.015
Chain 1: 3000 -19185.770 0.012 0.012
Chain 1: 3100 -19270.883 0.011 0.012
Chain 1: 3200 -18961.119 0.011 0.015
Chain 1: 3300 -19166.215 0.010 0.012
Chain 1: 3400 -18640.307 0.012 0.015
Chain 1: 3500 -19253.377 0.014 0.015
Chain 1: 3600 -18558.438 0.016 0.015
Chain 1: 3700 -18946.406 0.018 0.016
Chain 1: 3800 -17903.610 0.022 0.020
Chain 1: 3900 -17899.680 0.021 0.020
Chain 1: 4000 -18017.004 0.021 0.020
Chain 1: 4100 -17930.647 0.021 0.020
Chain 1: 4200 -17746.337 0.021 0.020
Chain 1: 4300 -17885.124 0.021 0.020
Chain 1: 4400 -17841.476 0.018 0.010
Chain 1: 4500 -17743.921 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002448 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 24.48 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13033.449 1.000 1.000
Chain 1: 200 -9443.559 0.690 1.000
Chain 1: 300 -8361.799 0.503 0.380
Chain 1: 400 -8545.462 0.383 0.380
Chain 1: 500 -8413.475 0.309 0.129
Chain 1: 600 -8602.184 0.261 0.129
Chain 1: 700 -8162.325 0.232 0.054
Chain 1: 800 -8166.123 0.203 0.054
Chain 1: 900 -8234.259 0.181 0.022
Chain 1: 1000 -8248.778 0.163 0.022
Chain 1: 1100 -8281.558 0.064 0.021
Chain 1: 1200 -8182.549 0.027 0.016
Chain 1: 1300 -8137.728 0.015 0.012
Chain 1: 1400 -8102.296 0.013 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001499 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.99 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61929.628 1.000 1.000
Chain 1: 200 -18782.202 1.649 2.297
Chain 1: 300 -9143.785 1.450 1.054
Chain 1: 400 -9283.486 1.092 1.054
Chain 1: 500 -8142.273 0.901 1.000
Chain 1: 600 -8848.490 0.764 1.000
Chain 1: 700 -8145.227 0.668 0.140
Chain 1: 800 -8506.045 0.589 0.140
Chain 1: 900 -7929.377 0.532 0.086
Chain 1: 1000 -8344.755 0.484 0.086
Chain 1: 1100 -7595.677 0.394 0.086
Chain 1: 1200 -7971.528 0.169 0.080
Chain 1: 1300 -8056.240 0.064 0.073
Chain 1: 1400 -7724.413 0.067 0.073
Chain 1: 1500 -7604.612 0.055 0.050
Chain 1: 1600 -7799.715 0.049 0.047
Chain 1: 1700 -7741.130 0.041 0.043
Chain 1: 1800 -7692.919 0.038 0.043
Chain 1: 1900 -7628.420 0.031 0.025
Chain 1: 2000 -7756.853 0.028 0.017
Chain 1: 2100 -7647.150 0.019 0.016
Chain 1: 2200 -7909.952 0.018 0.016
Chain 1: 2300 -7691.854 0.020 0.017
Chain 1: 2400 -7644.084 0.016 0.016
Chain 1: 2500 -7693.833 0.015 0.014
Chain 1: 2600 -7612.718 0.014 0.011
Chain 1: 2700 -7515.290 0.014 0.013
Chain 1: 2800 -7749.081 0.017 0.014
Chain 1: 2900 -7455.973 0.020 0.017
Chain 1: 3000 -7606.490 0.020 0.020
Chain 1: 3100 -7604.895 0.019 0.020
Chain 1: 3200 -7803.498 0.018 0.020
Chain 1: 3300 -7489.440 0.019 0.020
Chain 1: 3400 -7727.245 0.022 0.025
Chain 1: 3500 -7522.267 0.024 0.027
Chain 1: 3600 -7590.169 0.024 0.027
Chain 1: 3700 -7547.513 0.023 0.027
Chain 1: 3800 -7517.083 0.020 0.025
Chain 1: 3900 -7486.414 0.017 0.020
Chain 1: 4000 -7480.976 0.015 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003277 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.77 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86863.486 1.000 1.000
Chain 1: 200 -14268.763 3.044 5.088
Chain 1: 300 -10418.522 2.152 1.000
Chain 1: 400 -12809.042 1.661 1.000
Chain 1: 500 -8801.656 1.420 0.455
Chain 1: 600 -8749.998 1.184 0.455
Chain 1: 700 -9017.310 1.019 0.370
Chain 1: 800 -9033.670 0.892 0.370
Chain 1: 900 -9121.506 0.794 0.187
Chain 1: 1000 -9037.777 0.716 0.187
Chain 1: 1100 -9190.522 0.617 0.030 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8623.577 0.115 0.030
Chain 1: 1300 -8937.874 0.082 0.030
Chain 1: 1400 -8849.659 0.064 0.017
Chain 1: 1500 -8860.767 0.019 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003717 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 37.17 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8432994.094 1.000 1.000
Chain 1: 200 -1591610.446 2.649 4.298
Chain 1: 300 -893586.636 2.027 1.000
Chain 1: 400 -459512.059 1.756 1.000
Chain 1: 500 -359300.435 1.461 0.945
Chain 1: 600 -233677.168 1.307 0.945
Chain 1: 700 -119891.901 1.256 0.945
Chain 1: 800 -87163.418 1.146 0.945
Chain 1: 900 -67518.036 1.051 0.781
Chain 1: 1000 -52367.586 0.975 0.781
Chain 1: 1100 -39881.208 0.906 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39072.571 0.478 0.375
Chain 1: 1300 -27039.069 0.444 0.375
Chain 1: 1400 -26765.491 0.351 0.313
Chain 1: 1500 -23355.500 0.338 0.313
Chain 1: 1600 -22575.413 0.287 0.291
Chain 1: 1700 -21448.484 0.198 0.289
Chain 1: 1800 -21393.554 0.160 0.146
Chain 1: 1900 -21720.898 0.133 0.053
Chain 1: 2000 -20229.956 0.111 0.053
Chain 1: 2100 -20468.287 0.081 0.035
Chain 1: 2200 -20695.794 0.080 0.035
Chain 1: 2300 -20311.823 0.038 0.019
Chain 1: 2400 -20083.498 0.038 0.019
Chain 1: 2500 -19885.646 0.024 0.015
Chain 1: 2600 -19514.467 0.023 0.015
Chain 1: 2700 -19471.076 0.018 0.012
Chain 1: 2800 -19187.431 0.019 0.015
Chain 1: 2900 -19469.240 0.019 0.014
Chain 1: 3000 -19455.275 0.011 0.012
Chain 1: 3100 -19540.487 0.011 0.011
Chain 1: 3200 -19230.331 0.011 0.014
Chain 1: 3300 -19435.749 0.010 0.011
Chain 1: 3400 -18909.207 0.012 0.014
Chain 1: 3500 -19523.265 0.014 0.015
Chain 1: 3600 -18827.103 0.016 0.015
Chain 1: 3700 -19215.981 0.018 0.016
Chain 1: 3800 -18171.335 0.022 0.020
Chain 1: 3900 -18167.395 0.021 0.020
Chain 1: 4000 -18284.690 0.021 0.020
Chain 1: 4100 -18198.251 0.021 0.020
Chain 1: 4200 -18013.554 0.021 0.020
Chain 1: 4300 -18152.584 0.020 0.020
Chain 1: 4400 -18108.599 0.018 0.010
Chain 1: 4500 -18011.037 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002727 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 27.27 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49095.064 1.000 1.000
Chain 1: 200 -24173.903 1.015 1.031
Chain 1: 300 -21108.994 0.725 1.000
Chain 1: 400 -18540.400 0.579 1.000
Chain 1: 500 -14245.365 0.523 0.302
Chain 1: 600 -12297.580 0.462 0.302
Chain 1: 700 -16802.151 0.435 0.268
Chain 1: 800 -14479.255 0.400 0.268
Chain 1: 900 -15170.074 0.361 0.160
Chain 1: 1000 -11901.688 0.352 0.268
Chain 1: 1100 -10491.565 0.266 0.160
Chain 1: 1200 -11449.003 0.171 0.158
Chain 1: 1300 -15776.170 0.184 0.160
Chain 1: 1400 -16259.229 0.173 0.160
Chain 1: 1500 -10281.205 0.201 0.160
Chain 1: 1600 -11407.677 0.195 0.160
Chain 1: 1700 -10292.964 0.179 0.134
Chain 1: 1800 -10987.516 0.169 0.108
Chain 1: 1900 -10570.812 0.169 0.108
Chain 1: 2000 -10087.277 0.146 0.099
Chain 1: 2100 -10220.429 0.134 0.084
Chain 1: 2200 -10306.727 0.126 0.063
Chain 1: 2300 -11414.857 0.109 0.063
Chain 1: 2400 -19569.664 0.147 0.097
Chain 1: 2500 -9749.436 0.190 0.097
Chain 1: 2600 -15656.425 0.218 0.097
Chain 1: 2700 -9265.987 0.276 0.097
Chain 1: 2800 -9759.378 0.275 0.097
Chain 1: 2900 -9777.412 0.271 0.097
Chain 1: 3000 -9517.181 0.269 0.097
Chain 1: 3100 -8924.860 0.274 0.097
Chain 1: 3200 -9857.884 0.283 0.097
Chain 1: 3300 -15594.673 0.310 0.368
Chain 1: 3400 -17087.028 0.277 0.095
Chain 1: 3500 -9230.277 0.261 0.095
Chain 1: 3600 -9234.299 0.224 0.087
Chain 1: 3700 -9739.552 0.160 0.066
Chain 1: 3800 -9570.973 0.157 0.066
Chain 1: 3900 -9610.272 0.157 0.066
Chain 1: 4000 -8860.668 0.163 0.085
Chain 1: 4100 -9414.574 0.162 0.085
Chain 1: 4200 -10221.954 0.160 0.079
Chain 1: 4300 -10254.166 0.124 0.059
Chain 1: 4400 -9781.075 0.120 0.052
Chain 1: 4500 -16000.603 0.074 0.052
Chain 1: 4600 -8603.217 0.160 0.059
Chain 1: 4700 -10499.067 0.172 0.079
Chain 1: 4800 -13825.367 0.195 0.085
Chain 1: 4900 -13597.741 0.196 0.085
Chain 1: 5000 -9337.037 0.233 0.181
Chain 1: 5100 -13487.206 0.258 0.241
Chain 1: 5200 -13921.043 0.253 0.241
Chain 1: 5300 -10722.694 0.283 0.298
Chain 1: 5400 -8697.316 0.301 0.298
Chain 1: 5500 -8871.337 0.264 0.241
Chain 1: 5600 -8698.991 0.180 0.233
Chain 1: 5700 -9514.299 0.171 0.233
Chain 1: 5800 -12898.459 0.173 0.233
Chain 1: 5900 -13866.114 0.178 0.233
Chain 1: 6000 -8790.031 0.190 0.233
Chain 1: 6100 -9191.237 0.164 0.086
Chain 1: 6200 -8876.282 0.165 0.086
Chain 1: 6300 -8623.364 0.138 0.070
Chain 1: 6400 -12367.198 0.145 0.070
Chain 1: 6500 -8608.888 0.186 0.086
Chain 1: 6600 -9123.991 0.190 0.086
Chain 1: 6700 -8515.534 0.189 0.071
Chain 1: 6800 -10181.281 0.179 0.071
Chain 1: 6900 -12311.220 0.189 0.164
Chain 1: 7000 -9970.260 0.155 0.164
Chain 1: 7100 -8465.654 0.168 0.173
Chain 1: 7200 -8977.340 0.170 0.173
Chain 1: 7300 -10913.571 0.185 0.177
Chain 1: 7400 -9793.520 0.166 0.173
Chain 1: 7500 -9815.618 0.123 0.164
Chain 1: 7600 -9642.343 0.119 0.164
Chain 1: 7700 -9059.825 0.118 0.164
Chain 1: 7800 -9049.661 0.102 0.114
Chain 1: 7900 -8659.644 0.089 0.064
Chain 1: 8000 -8493.112 0.068 0.057
Chain 1: 8100 -9621.912 0.062 0.057
Chain 1: 8200 -8321.773 0.072 0.064
Chain 1: 8300 -9935.985 0.070 0.064
Chain 1: 8400 -8340.328 0.078 0.064
Chain 1: 8500 -10114.609 0.095 0.117
Chain 1: 8600 -8785.519 0.108 0.151
Chain 1: 8700 -8423.193 0.106 0.151
Chain 1: 8800 -8679.564 0.109 0.151
Chain 1: 8900 -10424.743 0.121 0.156
Chain 1: 9000 -9085.925 0.134 0.156
Chain 1: 9100 -9089.197 0.122 0.156
Chain 1: 9200 -8515.137 0.114 0.151
Chain 1: 9300 -9988.700 0.112 0.148
Chain 1: 9400 -11175.463 0.104 0.147
Chain 1: 9500 -10579.717 0.092 0.106
Chain 1: 9600 -8426.442 0.102 0.106
Chain 1: 9700 -8351.188 0.099 0.106
Chain 1: 9800 -10010.349 0.112 0.147
Chain 1: 9900 -11675.185 0.110 0.143
Chain 1: 10000 -10064.039 0.111 0.143
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001488 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.88 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58258.090 1.000 1.000
Chain 1: 200 -17880.482 1.629 2.258
Chain 1: 300 -8774.848 1.432 1.038
Chain 1: 400 -8226.214 1.091 1.038
Chain 1: 500 -8196.111 0.873 1.000
Chain 1: 600 -8104.645 0.730 1.000
Chain 1: 700 -7744.566 0.632 0.067
Chain 1: 800 -8257.580 0.561 0.067
Chain 1: 900 -7983.059 0.502 0.062
Chain 1: 1000 -7601.711 0.457 0.062
Chain 1: 1100 -7826.021 0.360 0.050
Chain 1: 1200 -7685.191 0.136 0.046
Chain 1: 1300 -7793.940 0.034 0.034
Chain 1: 1400 -7906.333 0.028 0.029
Chain 1: 1500 -7613.058 0.032 0.034
Chain 1: 1600 -7522.894 0.032 0.034
Chain 1: 1700 -7612.524 0.028 0.029
Chain 1: 1800 -7660.049 0.023 0.018
Chain 1: 1900 -7695.167 0.020 0.014
Chain 1: 2000 -7684.022 0.015 0.014
Chain 1: 2100 -7562.546 0.014 0.014
Chain 1: 2200 -7803.683 0.015 0.014
Chain 1: 2300 -7568.819 0.017 0.014
Chain 1: 2400 -7577.220 0.015 0.012
Chain 1: 2500 -7601.827 0.012 0.012
Chain 1: 2600 -7590.489 0.011 0.006 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003895 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 38.95 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87271.191 1.000 1.000
Chain 1: 200 -13672.708 3.191 5.383
Chain 1: 300 -10012.741 2.249 1.000
Chain 1: 400 -10898.047 1.707 1.000
Chain 1: 500 -8998.694 1.408 0.366
Chain 1: 600 -8670.578 1.180 0.366
Chain 1: 700 -8812.814 1.014 0.211
Chain 1: 800 -9298.350 0.893 0.211
Chain 1: 900 -8846.181 0.800 0.081
Chain 1: 1000 -8788.484 0.720 0.081
Chain 1: 1100 -8768.172 0.621 0.052 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8520.610 0.085 0.051
Chain 1: 1300 -8731.931 0.051 0.038
Chain 1: 1400 -8706.334 0.043 0.029
Chain 1: 1500 -8560.581 0.024 0.024
Chain 1: 1600 -8672.209 0.021 0.017
Chain 1: 1700 -8753.402 0.021 0.017
Chain 1: 1800 -8330.858 0.021 0.017
Chain 1: 1900 -8431.554 0.017 0.013
Chain 1: 2000 -8405.866 0.016 0.013
Chain 1: 2100 -8530.848 0.018 0.015
Chain 1: 2200 -8336.214 0.017 0.015
Chain 1: 2300 -8426.228 0.016 0.013
Chain 1: 2400 -8495.248 0.016 0.013
Chain 1: 2500 -8441.434 0.015 0.012
Chain 1: 2600 -8442.483 0.014 0.011
Chain 1: 2700 -8359.350 0.014 0.011
Chain 1: 2800 -8319.692 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.005543 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 55.43 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8445623.820 1.000 1.000
Chain 1: 200 -1590713.004 2.655 4.309
Chain 1: 300 -891583.481 2.031 1.000
Chain 1: 400 -457977.318 1.760 1.000
Chain 1: 500 -357990.969 1.464 0.947
Chain 1: 600 -232679.136 1.310 0.947
Chain 1: 700 -119156.736 1.259 0.947
Chain 1: 800 -86407.139 1.149 0.947
Chain 1: 900 -66796.823 1.054 0.784
Chain 1: 1000 -51633.901 0.978 0.784
Chain 1: 1100 -39153.693 0.910 0.539 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38333.698 0.481 0.379
Chain 1: 1300 -26338.921 0.448 0.379
Chain 1: 1400 -26061.188 0.354 0.319
Chain 1: 1500 -22661.430 0.341 0.319
Chain 1: 1600 -21881.788 0.291 0.294
Chain 1: 1700 -20761.861 0.201 0.294
Chain 1: 1800 -20707.336 0.164 0.150
Chain 1: 1900 -21033.637 0.136 0.054
Chain 1: 2000 -19547.714 0.114 0.054
Chain 1: 2100 -19785.956 0.083 0.036
Chain 1: 2200 -20011.970 0.082 0.036
Chain 1: 2300 -19629.516 0.039 0.019
Chain 1: 2400 -19401.651 0.039 0.019
Chain 1: 2500 -19203.352 0.025 0.016
Chain 1: 2600 -18833.767 0.023 0.016
Chain 1: 2700 -18790.732 0.018 0.012
Chain 1: 2800 -18507.465 0.019 0.015
Chain 1: 2900 -18788.684 0.019 0.015
Chain 1: 3000 -18774.858 0.012 0.012
Chain 1: 3100 -18859.902 0.011 0.012
Chain 1: 3200 -18550.569 0.012 0.015
Chain 1: 3300 -18755.288 0.011 0.012
Chain 1: 3400 -18230.083 0.012 0.015
Chain 1: 3500 -18842.037 0.015 0.015
Chain 1: 3600 -18148.557 0.016 0.015
Chain 1: 3700 -18535.501 0.018 0.017
Chain 1: 3800 -17494.875 0.023 0.021
Chain 1: 3900 -17490.963 0.021 0.021
Chain 1: 4000 -17608.320 0.022 0.021
Chain 1: 4100 -17522.070 0.022 0.021
Chain 1: 4200 -17338.208 0.021 0.021
Chain 1: 4300 -17476.718 0.021 0.021
Chain 1: 4400 -17433.516 0.018 0.011
Chain 1: 4500 -17335.968 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001309 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.09 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48660.720 1.000 1.000
Chain 1: 200 -12895.900 1.887 2.773
Chain 1: 300 -21606.011 1.392 1.000
Chain 1: 400 -12612.412 1.222 1.000
Chain 1: 500 -13265.813 0.988 0.713
Chain 1: 600 -16308.853 0.854 0.713
Chain 1: 700 -12962.424 0.769 0.403
Chain 1: 800 -12851.503 0.674 0.403
Chain 1: 900 -10988.796 0.618 0.258
Chain 1: 1000 -12678.739 0.569 0.258
Chain 1: 1100 -11380.435 0.481 0.187
Chain 1: 1200 -16897.420 0.236 0.187
Chain 1: 1300 -15011.717 0.208 0.170
Chain 1: 1400 -19867.186 0.162 0.170
Chain 1: 1500 -12057.784 0.221 0.187
Chain 1: 1600 -10534.447 0.217 0.170
Chain 1: 1700 -15188.995 0.222 0.170
Chain 1: 1800 -9262.588 0.285 0.244
Chain 1: 1900 -10306.218 0.278 0.244
Chain 1: 2000 -18896.345 0.310 0.306
Chain 1: 2100 -9736.730 0.393 0.326
Chain 1: 2200 -9331.842 0.365 0.306
Chain 1: 2300 -9770.844 0.357 0.306
Chain 1: 2400 -9452.090 0.336 0.306
Chain 1: 2500 -12068.146 0.293 0.217
Chain 1: 2600 -16738.776 0.306 0.279
Chain 1: 2700 -9489.558 0.352 0.279
Chain 1: 2800 -9555.345 0.289 0.217
Chain 1: 2900 -8727.097 0.288 0.217
Chain 1: 3000 -8727.036 0.242 0.095
Chain 1: 3100 -8553.658 0.150 0.045
Chain 1: 3200 -13988.926 0.185 0.095
Chain 1: 3300 -10000.795 0.220 0.217
Chain 1: 3400 -8852.192 0.230 0.217
Chain 1: 3500 -8996.098 0.210 0.130
Chain 1: 3600 -17206.476 0.230 0.130
Chain 1: 3700 -9310.195 0.238 0.130
Chain 1: 3800 -8711.304 0.244 0.130
Chain 1: 3900 -9903.454 0.247 0.130
Chain 1: 4000 -8658.355 0.261 0.144
Chain 1: 4100 -9332.743 0.266 0.144
Chain 1: 4200 -9100.080 0.230 0.130
Chain 1: 4300 -9667.659 0.196 0.120
Chain 1: 4400 -8764.503 0.193 0.103
Chain 1: 4500 -9097.026 0.195 0.103
Chain 1: 4600 -8528.549 0.154 0.072
Chain 1: 4700 -8743.576 0.072 0.069
Chain 1: 4800 -8283.025 0.071 0.067
Chain 1: 4900 -9407.099 0.071 0.067
Chain 1: 5000 -11718.200 0.076 0.067
Chain 1: 5100 -8806.318 0.102 0.067
Chain 1: 5200 -12757.822 0.130 0.103
Chain 1: 5300 -8422.301 0.176 0.119
Chain 1: 5400 -8392.205 0.166 0.119
Chain 1: 5500 -13097.254 0.198 0.197
Chain 1: 5600 -11655.179 0.204 0.197
Chain 1: 5700 -9383.039 0.226 0.242
Chain 1: 5800 -12716.033 0.246 0.262
Chain 1: 5900 -12085.429 0.240 0.262
Chain 1: 6000 -11482.267 0.225 0.262
Chain 1: 6100 -10797.564 0.198 0.242
Chain 1: 6200 -8480.023 0.195 0.242
Chain 1: 6300 -8335.435 0.145 0.124
Chain 1: 6400 -8404.657 0.145 0.124
Chain 1: 6500 -12481.272 0.142 0.124
Chain 1: 6600 -10774.953 0.146 0.158
Chain 1: 6700 -8107.620 0.154 0.158
Chain 1: 6800 -8871.423 0.137 0.086
Chain 1: 6900 -11252.497 0.153 0.158
Chain 1: 7000 -8908.517 0.174 0.212
Chain 1: 7100 -8053.645 0.178 0.212
Chain 1: 7200 -10848.643 0.176 0.212
Chain 1: 7300 -11060.125 0.177 0.212
Chain 1: 7400 -8065.213 0.213 0.258
Chain 1: 7500 -8085.989 0.180 0.212
Chain 1: 7600 -8176.708 0.166 0.212
Chain 1: 7700 -8557.732 0.137 0.106
Chain 1: 7800 -11136.337 0.152 0.212
Chain 1: 7900 -8211.787 0.166 0.232
Chain 1: 8000 -8724.101 0.146 0.106
Chain 1: 8100 -9048.571 0.139 0.059
Chain 1: 8200 -10752.367 0.129 0.059
Chain 1: 8300 -8153.978 0.159 0.158
Chain 1: 8400 -7867.714 0.125 0.059
Chain 1: 8500 -8450.399 0.132 0.069
Chain 1: 8600 -11406.569 0.157 0.158
Chain 1: 8700 -9591.796 0.171 0.189
Chain 1: 8800 -8232.499 0.165 0.165
Chain 1: 8900 -10698.788 0.152 0.165
Chain 1: 9000 -9362.365 0.161 0.165
Chain 1: 9100 -8354.548 0.169 0.165
Chain 1: 9200 -8502.928 0.155 0.165
Chain 1: 9300 -8225.902 0.126 0.143
Chain 1: 9400 -8295.043 0.124 0.143
Chain 1: 9500 -8237.608 0.117 0.143
Chain 1: 9600 -8275.004 0.092 0.121
Chain 1: 9700 -8116.784 0.075 0.034
Chain 1: 9800 -8023.022 0.060 0.019
Chain 1: 9900 -8038.151 0.037 0.017
Chain 1: 10000 -8140.467 0.024 0.013
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001412 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.12 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56875.781 1.000 1.000
Chain 1: 200 -17260.643 1.648 2.295
Chain 1: 300 -8615.962 1.433 1.003
Chain 1: 400 -8227.700 1.086 1.003
Chain 1: 500 -7794.907 0.880 1.000
Chain 1: 600 -8300.165 0.744 1.000
Chain 1: 700 -8492.627 0.641 0.061
Chain 1: 800 -8032.069 0.568 0.061
Chain 1: 900 -7893.084 0.507 0.057
Chain 1: 1000 -7780.642 0.457 0.057
Chain 1: 1100 -7675.578 0.359 0.056
Chain 1: 1200 -7718.889 0.130 0.047
Chain 1: 1300 -7704.891 0.030 0.023
Chain 1: 1400 -7748.453 0.026 0.018
Chain 1: 1500 -7557.521 0.022 0.018
Chain 1: 1600 -7627.232 0.017 0.014
Chain 1: 1700 -7479.081 0.017 0.014
Chain 1: 1800 -7513.921 0.012 0.014
Chain 1: 1900 -7527.086 0.010 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.006057 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 60.57 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86581.362 1.000 1.000
Chain 1: 200 -13278.028 3.260 5.521
Chain 1: 300 -9638.981 2.299 1.000
Chain 1: 400 -10495.977 1.745 1.000
Chain 1: 500 -8617.247 1.440 0.378
Chain 1: 600 -8381.684 1.204 0.378
Chain 1: 700 -8259.444 1.034 0.218
Chain 1: 800 -8512.874 0.909 0.218
Chain 1: 900 -8456.721 0.809 0.082
Chain 1: 1000 -8260.529 0.730 0.082
Chain 1: 1100 -8408.703 0.632 0.030 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -7970.625 0.085 0.030
Chain 1: 1300 -8256.987 0.051 0.030
Chain 1: 1400 -8316.249 0.044 0.028
Chain 1: 1500 -8209.458 0.023 0.024
Chain 1: 1600 -8318.698 0.022 0.018
Chain 1: 1700 -8395.683 0.021 0.018
Chain 1: 1800 -7982.515 0.023 0.018
Chain 1: 1900 -8078.736 0.024 0.018
Chain 1: 2000 -8052.039 0.022 0.013
Chain 1: 2100 -8175.017 0.021 0.013
Chain 1: 2200 -7994.792 0.018 0.013
Chain 1: 2300 -8073.515 0.016 0.013
Chain 1: 2400 -8143.202 0.016 0.013
Chain 1: 2500 -8088.714 0.015 0.012
Chain 1: 2600 -8088.323 0.014 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003548 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.48 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8406279.751 1.000 1.000
Chain 1: 200 -1587307.112 2.648 4.296
Chain 1: 300 -890952.907 2.026 1.000
Chain 1: 400 -457347.742 1.756 1.000
Chain 1: 500 -357286.284 1.461 0.948
Chain 1: 600 -232331.414 1.307 0.948
Chain 1: 700 -118786.066 1.257 0.948
Chain 1: 800 -86023.678 1.148 0.948
Chain 1: 900 -66422.012 1.053 0.782
Chain 1: 1000 -51258.141 0.977 0.782
Chain 1: 1100 -38767.647 0.909 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37949.212 0.482 0.381
Chain 1: 1300 -25943.445 0.450 0.381
Chain 1: 1400 -25666.264 0.356 0.322
Chain 1: 1500 -22262.236 0.344 0.322
Chain 1: 1600 -21481.167 0.293 0.296
Chain 1: 1700 -20359.695 0.203 0.295
Chain 1: 1800 -20304.935 0.166 0.153
Chain 1: 1900 -20631.043 0.138 0.055
Chain 1: 2000 -19144.319 0.116 0.055
Chain 1: 2100 -19382.856 0.085 0.036
Chain 1: 2200 -19608.752 0.084 0.036
Chain 1: 2300 -19226.418 0.040 0.020
Chain 1: 2400 -18998.533 0.040 0.020
Chain 1: 2500 -18800.264 0.025 0.016
Chain 1: 2600 -18430.873 0.024 0.016
Chain 1: 2700 -18387.902 0.018 0.012
Chain 1: 2800 -18104.638 0.020 0.016
Chain 1: 2900 -18385.807 0.020 0.015
Chain 1: 3000 -18372.132 0.012 0.012
Chain 1: 3100 -18457.102 0.011 0.012
Chain 1: 3200 -18147.895 0.012 0.015
Chain 1: 3300 -18352.503 0.011 0.012
Chain 1: 3400 -17827.504 0.013 0.015
Chain 1: 3500 -18439.209 0.015 0.016
Chain 1: 3600 -17746.071 0.017 0.016
Chain 1: 3700 -18132.702 0.019 0.017
Chain 1: 3800 -17092.673 0.023 0.021
Chain 1: 3900 -17088.758 0.022 0.021
Chain 1: 4000 -17206.125 0.022 0.021
Chain 1: 4100 -17119.878 0.022 0.021
Chain 1: 4200 -16936.159 0.022 0.021
Chain 1: 4300 -17074.581 0.021 0.021
Chain 1: 4400 -17031.457 0.019 0.011
Chain 1: 4500 -16933.934 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001256 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.56 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12300.409 1.000 1.000
Chain 1: 200 -9143.789 0.673 1.000
Chain 1: 300 -7888.003 0.501 0.345
Chain 1: 400 -8053.997 0.381 0.345
Chain 1: 500 -8005.615 0.306 0.159
Chain 1: 600 -8041.659 0.256 0.159
Chain 1: 700 -7817.417 0.223 0.029
Chain 1: 800 -7809.263 0.196 0.029
Chain 1: 900 -7761.660 0.175 0.021
Chain 1: 1000 -7843.208 0.158 0.021
Chain 1: 1100 -7836.633 0.058 0.010
Chain 1: 1200 -7800.887 0.024 0.006 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001425 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.25 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56744.865 1.000 1.000
Chain 1: 200 -17321.678 1.638 2.276
Chain 1: 300 -8708.057 1.422 1.000
Chain 1: 400 -8210.039 1.081 1.000
Chain 1: 500 -8097.005 0.868 0.989
Chain 1: 600 -8744.629 0.736 0.989
Chain 1: 700 -7910.673 0.646 0.105
Chain 1: 800 -8126.814 0.568 0.105
Chain 1: 900 -7772.196 0.510 0.074
Chain 1: 1000 -7810.712 0.460 0.074
Chain 1: 1100 -7926.432 0.361 0.061
Chain 1: 1200 -7921.539 0.134 0.046
Chain 1: 1300 -7640.817 0.038 0.037
Chain 1: 1400 -8040.485 0.037 0.037
Chain 1: 1500 -7661.060 0.041 0.046
Chain 1: 1600 -7552.997 0.035 0.037
Chain 1: 1700 -7545.671 0.024 0.027
Chain 1: 1800 -7598.976 0.022 0.015
Chain 1: 1900 -7595.367 0.018 0.014
Chain 1: 2000 -7621.631 0.018 0.014
Chain 1: 2100 -7626.987 0.016 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003873 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 38.73 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86755.233 1.000 1.000
Chain 1: 200 -13370.986 3.244 5.488
Chain 1: 300 -9757.866 2.286 1.000
Chain 1: 400 -10748.340 1.738 1.000
Chain 1: 500 -8583.402 1.441 0.370
Chain 1: 600 -8269.898 1.207 0.370
Chain 1: 700 -8283.249 1.035 0.252
Chain 1: 800 -8461.670 0.908 0.252
Chain 1: 900 -8575.631 0.809 0.092
Chain 1: 1000 -8365.182 0.730 0.092
Chain 1: 1100 -8538.777 0.632 0.038 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8153.050 0.088 0.038
Chain 1: 1300 -8360.549 0.054 0.025
Chain 1: 1400 -8479.719 0.046 0.025
Chain 1: 1500 -8316.679 0.023 0.021
Chain 1: 1600 -8434.268 0.020 0.020
Chain 1: 1700 -8514.649 0.021 0.020
Chain 1: 1800 -8104.763 0.024 0.020
Chain 1: 1900 -8201.151 0.024 0.020
Chain 1: 2000 -8173.605 0.022 0.020
Chain 1: 2100 -8295.230 0.021 0.015
Chain 1: 2200 -8137.200 0.018 0.015
Chain 1: 2300 -8199.389 0.016 0.014
Chain 1: 2400 -8265.930 0.016 0.014
Chain 1: 2500 -8211.381 0.015 0.012
Chain 1: 2600 -8209.707 0.013 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003714 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 37.14 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8405002.404 1.000 1.000
Chain 1: 200 -1583496.240 2.654 4.308
Chain 1: 300 -889917.267 2.029 1.000
Chain 1: 400 -456774.250 1.759 1.000
Chain 1: 500 -357447.982 1.463 0.948
Chain 1: 600 -232392.113 1.309 0.948
Chain 1: 700 -118876.184 1.258 0.948
Chain 1: 800 -86145.925 1.148 0.948
Chain 1: 900 -66528.782 1.053 0.779
Chain 1: 1000 -51359.294 0.978 0.779
Chain 1: 1100 -38867.475 0.910 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38046.017 0.481 0.380
Chain 1: 1300 -26034.241 0.449 0.380
Chain 1: 1400 -25755.322 0.356 0.321
Chain 1: 1500 -22350.801 0.343 0.321
Chain 1: 1600 -21569.663 0.293 0.295
Chain 1: 1700 -20447.213 0.203 0.295
Chain 1: 1800 -20392.258 0.165 0.152
Chain 1: 1900 -20718.301 0.137 0.055
Chain 1: 2000 -19231.599 0.115 0.055
Chain 1: 2100 -19469.907 0.085 0.036
Chain 1: 2200 -19695.960 0.084 0.036
Chain 1: 2300 -19313.499 0.039 0.020
Chain 1: 2400 -19085.668 0.039 0.020
Chain 1: 2500 -18887.508 0.025 0.016
Chain 1: 2600 -18518.034 0.024 0.016
Chain 1: 2700 -18475.086 0.018 0.012
Chain 1: 2800 -18191.977 0.020 0.016
Chain 1: 2900 -18473.040 0.020 0.015
Chain 1: 3000 -18459.287 0.012 0.012
Chain 1: 3100 -18544.290 0.011 0.012
Chain 1: 3200 -18235.091 0.012 0.015
Chain 1: 3300 -18439.694 0.011 0.012
Chain 1: 3400 -17914.822 0.013 0.015
Chain 1: 3500 -18526.421 0.015 0.016
Chain 1: 3600 -17833.372 0.017 0.016
Chain 1: 3700 -18219.956 0.019 0.017
Chain 1: 3800 -17180.164 0.023 0.021
Chain 1: 3900 -17176.290 0.022 0.021
Chain 1: 4000 -17293.589 0.022 0.021
Chain 1: 4100 -17207.419 0.022 0.021
Chain 1: 4200 -17023.755 0.022 0.021
Chain 1: 4300 -17162.116 0.021 0.021
Chain 1: 4400 -17119.024 0.019 0.011
Chain 1: 4500 -17021.543 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001431 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.31 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48616.589 1.000 1.000
Chain 1: 200 -22476.101 1.082 1.163
Chain 1: 300 -29090.876 0.797 1.000
Chain 1: 400 -15929.425 0.804 1.000
Chain 1: 500 -12371.323 0.701 0.826
Chain 1: 600 -14411.321 0.608 0.826
Chain 1: 700 -11212.398 0.562 0.288
Chain 1: 800 -13334.159 0.511 0.288
Chain 1: 900 -10615.157 0.483 0.285
Chain 1: 1000 -10623.933 0.435 0.285
Chain 1: 1100 -19299.810 0.380 0.285
Chain 1: 1200 -18289.820 0.269 0.256
Chain 1: 1300 -12087.837 0.297 0.285
Chain 1: 1400 -11642.060 0.219 0.256
Chain 1: 1500 -9964.244 0.207 0.168
Chain 1: 1600 -12133.182 0.210 0.179
Chain 1: 1700 -20019.440 0.221 0.179
Chain 1: 1800 -12885.199 0.261 0.256
Chain 1: 1900 -10524.580 0.258 0.224
Chain 1: 2000 -10149.952 0.261 0.224
Chain 1: 2100 -19948.000 0.265 0.224
Chain 1: 2200 -9898.332 0.361 0.394
Chain 1: 2300 -9477.264 0.315 0.224
Chain 1: 2400 -11631.062 0.329 0.224
Chain 1: 2500 -9074.326 0.341 0.282
Chain 1: 2600 -9404.544 0.326 0.282
Chain 1: 2700 -15374.423 0.326 0.282
Chain 1: 2800 -12987.399 0.289 0.224
Chain 1: 2900 -9213.374 0.307 0.282
Chain 1: 3000 -8767.691 0.309 0.282
Chain 1: 3100 -9753.009 0.270 0.185
Chain 1: 3200 -13162.251 0.194 0.185
Chain 1: 3300 -8985.966 0.236 0.259
Chain 1: 3400 -14280.564 0.254 0.282
Chain 1: 3500 -8722.156 0.290 0.371
Chain 1: 3600 -9543.338 0.295 0.371
Chain 1: 3700 -14427.488 0.290 0.339
Chain 1: 3800 -14934.758 0.275 0.339
Chain 1: 3900 -13307.066 0.246 0.259
Chain 1: 4000 -9007.440 0.289 0.339
Chain 1: 4100 -8635.237 0.283 0.339
Chain 1: 4200 -9453.188 0.266 0.339
Chain 1: 4300 -9792.781 0.223 0.122
Chain 1: 4400 -9500.624 0.189 0.087
Chain 1: 4500 -8605.855 0.136 0.087
Chain 1: 4600 -8257.812 0.131 0.087
Chain 1: 4700 -8274.379 0.098 0.043
Chain 1: 4800 -12642.217 0.129 0.087
Chain 1: 4900 -9812.371 0.145 0.087
Chain 1: 5000 -9295.100 0.103 0.056
Chain 1: 5100 -8520.374 0.108 0.087
Chain 1: 5200 -13849.563 0.138 0.091
Chain 1: 5300 -8538.684 0.197 0.104
Chain 1: 5400 -14305.396 0.234 0.288
Chain 1: 5500 -14161.068 0.224 0.288
Chain 1: 5600 -8634.847 0.284 0.345
Chain 1: 5700 -11468.530 0.309 0.345
Chain 1: 5800 -8743.911 0.305 0.312
Chain 1: 5900 -8227.276 0.283 0.312
Chain 1: 6000 -9729.484 0.293 0.312
Chain 1: 6100 -12211.565 0.304 0.312
Chain 1: 6200 -8180.393 0.315 0.312
Chain 1: 6300 -12075.181 0.285 0.312
Chain 1: 6400 -12142.526 0.245 0.247
Chain 1: 6500 -11509.755 0.249 0.247
Chain 1: 6600 -8827.383 0.216 0.247
Chain 1: 6700 -8585.734 0.194 0.203
Chain 1: 6800 -8188.381 0.168 0.154
Chain 1: 6900 -10927.635 0.186 0.203
Chain 1: 7000 -8313.204 0.202 0.251
Chain 1: 7100 -8626.930 0.186 0.251
Chain 1: 7200 -8133.625 0.143 0.061
Chain 1: 7300 -10425.231 0.132 0.061
Chain 1: 7400 -12306.189 0.147 0.153
Chain 1: 7500 -9652.135 0.169 0.220
Chain 1: 7600 -9011.980 0.146 0.153
Chain 1: 7700 -8196.761 0.153 0.153
Chain 1: 7800 -9059.368 0.158 0.153
Chain 1: 7900 -12306.063 0.159 0.153
Chain 1: 8000 -8709.606 0.169 0.153
Chain 1: 8100 -8529.421 0.167 0.153
Chain 1: 8200 -11322.928 0.186 0.220
Chain 1: 8300 -9641.401 0.181 0.174
Chain 1: 8400 -8148.936 0.184 0.183
Chain 1: 8500 -8206.930 0.157 0.174
Chain 1: 8600 -7943.940 0.154 0.174
Chain 1: 8700 -11632.451 0.175 0.183
Chain 1: 8800 -8079.570 0.210 0.247
Chain 1: 8900 -8641.933 0.190 0.183
Chain 1: 9000 -8323.798 0.153 0.174
Chain 1: 9100 -8358.042 0.151 0.174
Chain 1: 9200 -9663.367 0.140 0.135
Chain 1: 9300 -11149.846 0.136 0.133
Chain 1: 9400 -13344.182 0.134 0.133
Chain 1: 9500 -8084.777 0.198 0.135
Chain 1: 9600 -8694.188 0.202 0.135
Chain 1: 9700 -10117.587 0.184 0.135
Chain 1: 9800 -8846.524 0.155 0.135
Chain 1: 9900 -8876.840 0.148 0.135
Chain 1: 10000 -9277.780 0.149 0.135
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001435 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.35 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57288.830 1.000 1.000
Chain 1: 200 -17365.340 1.650 2.299
Chain 1: 300 -8629.892 1.437 1.012
Chain 1: 400 -8222.044 1.090 1.012
Chain 1: 500 -8045.306 0.877 1.000
Chain 1: 600 -8357.975 0.737 1.000
Chain 1: 700 -7741.942 0.643 0.080
Chain 1: 800 -8002.306 0.567 0.080
Chain 1: 900 -7866.261 0.506 0.050
Chain 1: 1000 -8127.062 0.458 0.050
Chain 1: 1100 -7524.498 0.366 0.050
Chain 1: 1200 -7559.390 0.137 0.037
Chain 1: 1300 -7540.342 0.036 0.033
Chain 1: 1400 -7801.754 0.034 0.033
Chain 1: 1500 -7516.371 0.036 0.034
Chain 1: 1600 -7672.170 0.034 0.033
Chain 1: 1700 -7471.150 0.029 0.032
Chain 1: 1800 -7528.522 0.026 0.027
Chain 1: 1900 -7538.240 0.025 0.027
Chain 1: 2000 -7478.187 0.022 0.020
Chain 1: 2100 -7431.024 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003257 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.57 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85864.314 1.000 1.000
Chain 1: 200 -13362.351 3.213 5.426
Chain 1: 300 -9739.671 2.266 1.000
Chain 1: 400 -10707.742 1.722 1.000
Chain 1: 500 -8718.495 1.423 0.372
Chain 1: 600 -8222.850 1.196 0.372
Chain 1: 700 -8283.074 1.026 0.228
Chain 1: 800 -9064.913 0.909 0.228
Chain 1: 900 -8571.242 0.814 0.090
Chain 1: 1000 -8384.191 0.735 0.090
Chain 1: 1100 -8526.381 0.637 0.086 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8259.280 0.097 0.060
Chain 1: 1300 -8388.488 0.062 0.058
Chain 1: 1400 -8428.334 0.053 0.032
Chain 1: 1500 -8321.210 0.032 0.022
Chain 1: 1600 -8425.689 0.027 0.017
Chain 1: 1700 -8517.952 0.027 0.017
Chain 1: 1800 -8107.279 0.024 0.017
Chain 1: 1900 -8203.236 0.019 0.015
Chain 1: 2000 -8175.911 0.017 0.013
Chain 1: 2100 -8297.587 0.017 0.013
Chain 1: 2200 -8134.839 0.016 0.013
Chain 1: 2300 -8200.357 0.015 0.012
Chain 1: 2400 -8267.744 0.015 0.012
Chain 1: 2500 -8213.525 0.015 0.012
Chain 1: 2600 -8211.899 0.013 0.011
Chain 1: 2700 -8129.133 0.013 0.010
Chain 1: 2800 -8094.214 0.009 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003599 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.99 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8391224.325 1.000 1.000
Chain 1: 200 -1585082.056 2.647 4.294
Chain 1: 300 -892029.343 2.024 1.000
Chain 1: 400 -458429.287 1.754 1.000
Chain 1: 500 -358579.689 1.459 0.946
Chain 1: 600 -233386.945 1.305 0.946
Chain 1: 700 -119343.215 1.255 0.946
Chain 1: 800 -86488.726 1.146 0.946
Chain 1: 900 -66786.090 1.051 0.777
Chain 1: 1000 -51543.784 0.976 0.777
Chain 1: 1100 -38985.867 0.908 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38159.433 0.481 0.380
Chain 1: 1300 -26082.623 0.449 0.380
Chain 1: 1400 -25798.344 0.356 0.322
Chain 1: 1500 -22377.232 0.343 0.322
Chain 1: 1600 -21591.436 0.293 0.296
Chain 1: 1700 -20461.236 0.203 0.295
Chain 1: 1800 -20404.564 0.166 0.153
Chain 1: 1900 -20730.590 0.138 0.055
Chain 1: 2000 -19240.110 0.116 0.055
Chain 1: 2100 -19478.451 0.085 0.036
Chain 1: 2200 -19705.196 0.084 0.036
Chain 1: 2300 -19322.224 0.040 0.020
Chain 1: 2400 -19094.331 0.040 0.020
Chain 1: 2500 -18896.546 0.025 0.016
Chain 1: 2600 -18526.578 0.024 0.016
Chain 1: 2700 -18483.568 0.018 0.012
Chain 1: 2800 -18200.479 0.020 0.016
Chain 1: 2900 -18481.844 0.020 0.015
Chain 1: 3000 -18467.928 0.012 0.012
Chain 1: 3100 -18552.892 0.011 0.012
Chain 1: 3200 -18243.600 0.012 0.015
Chain 1: 3300 -18448.374 0.011 0.012
Chain 1: 3400 -17923.347 0.013 0.015
Chain 1: 3500 -18535.127 0.015 0.016
Chain 1: 3600 -17842.057 0.017 0.016
Chain 1: 3700 -18228.667 0.019 0.017
Chain 1: 3800 -17188.720 0.023 0.021
Chain 1: 3900 -17184.939 0.022 0.021
Chain 1: 4000 -17302.208 0.022 0.021
Chain 1: 4100 -17215.928 0.022 0.021
Chain 1: 4200 -17032.326 0.022 0.021
Chain 1: 4300 -17170.592 0.021 0.021
Chain 1: 4400 -17127.480 0.019 0.011
Chain 1: 4500 -17030.084 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002929 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.29 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12527.956 1.000 1.000
Chain 1: 200 -9420.400 0.665 1.000
Chain 1: 300 -8184.594 0.494 0.330
Chain 1: 400 -8290.689 0.373 0.330
Chain 1: 500 -8325.279 0.300 0.151
Chain 1: 600 -8079.306 0.255 0.151
Chain 1: 700 -8001.769 0.220 0.030
Chain 1: 800 -8004.032 0.192 0.030
Chain 1: 900 -8033.239 0.171 0.013
Chain 1: 1000 -8029.864 0.154 0.013
Chain 1: 1100 -8075.587 0.055 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003005 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.05 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56762.019 1.000 1.000
Chain 1: 200 -17531.264 1.619 2.238
Chain 1: 300 -8804.005 1.410 1.000
Chain 1: 400 -8223.265 1.075 1.000
Chain 1: 500 -8626.008 0.869 0.991
Chain 1: 600 -8490.093 0.727 0.991
Chain 1: 700 -8135.994 0.629 0.071
Chain 1: 800 -8238.722 0.552 0.071
Chain 1: 900 -7953.313 0.495 0.047
Chain 1: 1000 -7776.148 0.448 0.047
Chain 1: 1100 -7705.334 0.349 0.044
Chain 1: 1200 -7643.195 0.126 0.036
Chain 1: 1300 -7825.003 0.029 0.023
Chain 1: 1400 -7933.688 0.023 0.023
Chain 1: 1500 -7625.484 0.023 0.023
Chain 1: 1600 -7807.218 0.023 0.023
Chain 1: 1700 -7542.275 0.022 0.023
Chain 1: 1800 -7644.664 0.023 0.023
Chain 1: 1900 -7542.278 0.020 0.023
Chain 1: 2000 -7653.833 0.019 0.015
Chain 1: 2100 -7628.323 0.019 0.015
Chain 1: 2200 -7761.833 0.020 0.017
Chain 1: 2300 -7629.945 0.019 0.017
Chain 1: 2400 -7694.719 0.019 0.017
Chain 1: 2500 -7608.120 0.016 0.015
Chain 1: 2600 -7587.865 0.014 0.014
Chain 1: 2700 -7583.831 0.010 0.013
Chain 1: 2800 -7526.363 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.006863 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 68.63 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86931.392 1.000 1.000
Chain 1: 200 -13634.289 3.188 5.376
Chain 1: 300 -9981.858 2.247 1.000
Chain 1: 400 -10839.494 1.705 1.000
Chain 1: 500 -8959.513 1.406 0.366
Chain 1: 600 -8710.431 1.177 0.366
Chain 1: 700 -8651.434 1.009 0.210
Chain 1: 800 -9309.984 0.892 0.210
Chain 1: 900 -8745.783 0.800 0.079
Chain 1: 1000 -8602.432 0.722 0.079
Chain 1: 1100 -8833.945 0.624 0.071 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8340.480 0.093 0.065
Chain 1: 1300 -8671.033 0.060 0.059
Chain 1: 1400 -8673.930 0.052 0.038
Chain 1: 1500 -8546.370 0.033 0.029
Chain 1: 1600 -8654.120 0.031 0.026
Chain 1: 1700 -8737.013 0.031 0.026
Chain 1: 1800 -8321.968 0.029 0.026
Chain 1: 1900 -8418.614 0.024 0.017
Chain 1: 2000 -8392.069 0.023 0.015
Chain 1: 2100 -8515.109 0.021 0.014
Chain 1: 2200 -8334.486 0.018 0.014
Chain 1: 2300 -8413.282 0.015 0.012
Chain 1: 2400 -8483.019 0.016 0.012
Chain 1: 2500 -8428.651 0.015 0.011
Chain 1: 2600 -8428.363 0.013 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003845 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 38.45 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8398150.617 1.000 1.000
Chain 1: 200 -1586271.051 2.647 4.294
Chain 1: 300 -892184.259 2.024 1.000
Chain 1: 400 -458616.303 1.754 1.000
Chain 1: 500 -358728.411 1.459 0.945
Chain 1: 600 -233537.983 1.305 0.945
Chain 1: 700 -119520.856 1.255 0.945
Chain 1: 800 -86666.870 1.146 0.945
Chain 1: 900 -66978.288 1.051 0.778
Chain 1: 1000 -51752.842 0.975 0.778
Chain 1: 1100 -39208.844 0.907 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38382.612 0.480 0.379
Chain 1: 1300 -26322.455 0.448 0.379
Chain 1: 1400 -26040.186 0.355 0.320
Chain 1: 1500 -22622.777 0.342 0.320
Chain 1: 1600 -21837.428 0.292 0.294
Chain 1: 1700 -20709.489 0.202 0.294
Chain 1: 1800 -20653.130 0.164 0.151
Chain 1: 1900 -20979.204 0.136 0.054
Chain 1: 2000 -19489.266 0.115 0.054
Chain 1: 2100 -19727.891 0.084 0.036
Chain 1: 2200 -19954.382 0.083 0.036
Chain 1: 2300 -19571.516 0.039 0.020
Chain 1: 2400 -19343.573 0.039 0.020
Chain 1: 2500 -19145.574 0.025 0.016
Chain 1: 2600 -18775.886 0.023 0.016
Chain 1: 2700 -18732.822 0.018 0.012
Chain 1: 2800 -18449.699 0.019 0.015
Chain 1: 2900 -18730.933 0.019 0.015
Chain 1: 3000 -18717.181 0.012 0.012
Chain 1: 3100 -18802.167 0.011 0.012
Chain 1: 3200 -18492.871 0.012 0.015
Chain 1: 3300 -18697.528 0.011 0.012
Chain 1: 3400 -18172.527 0.012 0.015
Chain 1: 3500 -18784.337 0.015 0.015
Chain 1: 3600 -18091.075 0.017 0.015
Chain 1: 3700 -18477.852 0.018 0.017
Chain 1: 3800 -17437.667 0.023 0.021
Chain 1: 3900 -17433.776 0.021 0.021
Chain 1: 4000 -17551.100 0.022 0.021
Chain 1: 4100 -17464.887 0.022 0.021
Chain 1: 4200 -17281.103 0.021 0.021
Chain 1: 4300 -17419.515 0.021 0.021
Chain 1: 4400 -17376.365 0.018 0.011
Chain 1: 4500 -17278.875 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002775 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 27.75 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49249.937 1.000 1.000
Chain 1: 200 -18447.912 1.335 1.670
Chain 1: 300 -18637.940 0.893 1.000
Chain 1: 400 -14290.803 0.746 1.000
Chain 1: 500 -16999.292 0.629 0.304
Chain 1: 600 -14274.506 0.556 0.304
Chain 1: 700 -12169.554 0.501 0.191
Chain 1: 800 -13850.578 0.454 0.191
Chain 1: 900 -14968.340 0.411 0.173
Chain 1: 1000 -10516.603 0.413 0.191
Chain 1: 1100 -17521.037 0.353 0.191
Chain 1: 1200 -10489.952 0.253 0.191
Chain 1: 1300 -13183.909 0.272 0.204
Chain 1: 1400 -21583.663 0.281 0.204
Chain 1: 1500 -22319.245 0.268 0.204
Chain 1: 1600 -11342.892 0.346 0.389
Chain 1: 1700 -11836.954 0.333 0.389
Chain 1: 1800 -10392.053 0.334 0.389
Chain 1: 1900 -18599.336 0.371 0.400
Chain 1: 2000 -16197.245 0.343 0.389
Chain 1: 2100 -19366.123 0.320 0.204
Chain 1: 2200 -10978.166 0.329 0.204
Chain 1: 2300 -9426.367 0.325 0.165
Chain 1: 2400 -9955.554 0.292 0.164
Chain 1: 2500 -13331.998 0.314 0.165
Chain 1: 2600 -9641.446 0.255 0.165
Chain 1: 2700 -9434.458 0.253 0.165
Chain 1: 2800 -9620.116 0.241 0.165
Chain 1: 2900 -9953.594 0.200 0.164
Chain 1: 3000 -11457.942 0.199 0.164
Chain 1: 3100 -8766.034 0.213 0.165
Chain 1: 3200 -9334.555 0.143 0.131
Chain 1: 3300 -10844.859 0.140 0.131
Chain 1: 3400 -17870.596 0.174 0.139
Chain 1: 3500 -9669.828 0.234 0.139
Chain 1: 3600 -8779.067 0.206 0.131
Chain 1: 3700 -9352.516 0.210 0.131
Chain 1: 3800 -9339.881 0.208 0.131
Chain 1: 3900 -8903.667 0.209 0.131
Chain 1: 4000 -8990.229 0.197 0.101
Chain 1: 4100 -9434.877 0.171 0.061
Chain 1: 4200 -13544.919 0.195 0.101
Chain 1: 4300 -9112.710 0.230 0.101
Chain 1: 4400 -9640.329 0.196 0.061
Chain 1: 4500 -9626.985 0.112 0.055
Chain 1: 4600 -8990.669 0.109 0.055
Chain 1: 4700 -9912.139 0.112 0.055
Chain 1: 4800 -8816.969 0.124 0.071
Chain 1: 4900 -8947.626 0.121 0.071
Chain 1: 5000 -16640.265 0.166 0.093
Chain 1: 5100 -9489.228 0.236 0.124
Chain 1: 5200 -13515.927 0.236 0.124
Chain 1: 5300 -12270.940 0.197 0.101
Chain 1: 5400 -8993.834 0.228 0.124
Chain 1: 5500 -11320.051 0.249 0.205
Chain 1: 5600 -14716.924 0.265 0.231
Chain 1: 5700 -9138.430 0.317 0.298
Chain 1: 5800 -8509.346 0.311 0.298
Chain 1: 5900 -8663.082 0.312 0.298
Chain 1: 6000 -9353.463 0.273 0.231
Chain 1: 6100 -8921.837 0.202 0.205
Chain 1: 6200 -9345.739 0.177 0.101
Chain 1: 6300 -8861.968 0.172 0.074
Chain 1: 6400 -8787.761 0.137 0.074
Chain 1: 6500 -8951.606 0.118 0.055
Chain 1: 6600 -9938.760 0.105 0.055
Chain 1: 6700 -9214.944 0.052 0.055
Chain 1: 6800 -9580.907 0.048 0.048
Chain 1: 6900 -8814.746 0.055 0.055
Chain 1: 7000 -9738.564 0.057 0.055
Chain 1: 7100 -8312.171 0.070 0.079
Chain 1: 7200 -8875.926 0.071 0.079
Chain 1: 7300 -10438.626 0.081 0.087
Chain 1: 7400 -8708.947 0.100 0.095
Chain 1: 7500 -10572.343 0.116 0.099
Chain 1: 7600 -8501.617 0.130 0.150
Chain 1: 7700 -8656.522 0.124 0.150
Chain 1: 7800 -9900.892 0.133 0.150
Chain 1: 7900 -8547.950 0.140 0.158
Chain 1: 8000 -9027.074 0.136 0.158
Chain 1: 8100 -8614.993 0.123 0.150
Chain 1: 8200 -12426.233 0.148 0.158
Chain 1: 8300 -8365.906 0.181 0.176
Chain 1: 8400 -9802.602 0.176 0.158
Chain 1: 8500 -12266.114 0.179 0.158
Chain 1: 8600 -8463.764 0.199 0.158
Chain 1: 8700 -8382.436 0.198 0.158
Chain 1: 8800 -8689.421 0.189 0.158
Chain 1: 8900 -9325.827 0.180 0.147
Chain 1: 9000 -10933.847 0.190 0.147
Chain 1: 9100 -8449.977 0.214 0.201
Chain 1: 9200 -14585.725 0.226 0.201
Chain 1: 9300 -8725.162 0.244 0.201
Chain 1: 9400 -8548.045 0.232 0.201
Chain 1: 9500 -8339.254 0.214 0.147
Chain 1: 9600 -8484.146 0.171 0.068
Chain 1: 9700 -8851.422 0.174 0.068
Chain 1: 9800 -8552.459 0.174 0.068
Chain 1: 9900 -8631.690 0.168 0.041
Chain 1: 10000 -10961.749 0.175 0.041
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003884 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 38.84 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -46905.446 1.000 1.000
Chain 1: 200 -15880.501 1.477 1.954
Chain 1: 300 -8824.602 1.251 1.000
Chain 1: 400 -8676.994 0.943 1.000
Chain 1: 500 -8560.202 0.757 0.800
Chain 1: 600 -8715.460 0.634 0.800
Chain 1: 700 -7833.307 0.559 0.113
Chain 1: 800 -8223.472 0.495 0.113
Chain 1: 900 -7876.285 0.445 0.047
Chain 1: 1000 -7882.630 0.401 0.047
Chain 1: 1100 -7824.611 0.301 0.044
Chain 1: 1200 -7815.929 0.106 0.018
Chain 1: 1300 -7784.507 0.027 0.017
Chain 1: 1400 -8018.960 0.028 0.018
Chain 1: 1500 -7568.795 0.032 0.029
Chain 1: 1600 -7747.734 0.033 0.029
Chain 1: 1700 -7582.525 0.024 0.023
Chain 1: 1800 -7684.173 0.020 0.022
Chain 1: 1900 -7566.175 0.018 0.016
Chain 1: 2000 -7663.280 0.019 0.016
Chain 1: 2100 -7574.422 0.019 0.016
Chain 1: 2200 -7726.141 0.021 0.020
Chain 1: 2300 -7589.681 0.022 0.020
Chain 1: 2400 -7589.338 0.020 0.018
Chain 1: 2500 -7606.586 0.014 0.016
Chain 1: 2600 -7533.612 0.012 0.013
Chain 1: 2700 -7637.268 0.012 0.013
Chain 1: 2800 -7623.997 0.010 0.013
Chain 1: 2900 -7392.318 0.012 0.013
Chain 1: 3000 -7530.380 0.013 0.014
Chain 1: 3100 -7528.261 0.011 0.014
Chain 1: 3200 -7733.353 0.012 0.014
Chain 1: 3300 -7456.717 0.014 0.014
Chain 1: 3400 -7677.013 0.017 0.018
Chain 1: 3500 -7440.640 0.020 0.027
Chain 1: 3600 -7507.200 0.020 0.027
Chain 1: 3700 -7456.365 0.019 0.027
Chain 1: 3800 -7453.944 0.019 0.027
Chain 1: 3900 -7422.081 0.016 0.018
Chain 1: 4000 -7415.358 0.015 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.006565 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 65.65 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86137.351 1.000 1.000
Chain 1: 200 -13748.686 3.133 5.265
Chain 1: 300 -10022.880 2.212 1.000
Chain 1: 400 -11311.093 1.688 1.000
Chain 1: 500 -9045.974 1.400 0.372
Chain 1: 600 -8507.248 1.177 0.372
Chain 1: 700 -9221.765 1.020 0.250
Chain 1: 800 -8336.772 0.906 0.250
Chain 1: 900 -8351.241 0.806 0.114
Chain 1: 1000 -8716.551 0.729 0.114
Chain 1: 1100 -8825.615 0.630 0.106 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8378.416 0.109 0.077
Chain 1: 1300 -8659.014 0.075 0.063
Chain 1: 1400 -8651.594 0.064 0.053
Chain 1: 1500 -8538.664 0.040 0.042
Chain 1: 1600 -8641.435 0.035 0.032
Chain 1: 1700 -8702.241 0.028 0.013
Chain 1: 1800 -8262.827 0.023 0.013
Chain 1: 1900 -8368.516 0.024 0.013
Chain 1: 2000 -8348.756 0.020 0.013
Chain 1: 2100 -8480.213 0.020 0.013
Chain 1: 2200 -8269.744 0.017 0.013
Chain 1: 2300 -8365.425 0.015 0.013
Chain 1: 2400 -8429.573 0.016 0.013
Chain 1: 2500 -8376.911 0.015 0.012
Chain 1: 2600 -8387.333 0.014 0.011
Chain 1: 2700 -8297.464 0.015 0.011
Chain 1: 2800 -8247.272 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004032 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 40.32 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8378497.040 1.000 1.000
Chain 1: 200 -1582261.202 2.648 4.295
Chain 1: 300 -891686.715 2.023 1.000
Chain 1: 400 -458473.349 1.754 1.000
Chain 1: 500 -359210.050 1.458 0.945
Chain 1: 600 -234049.275 1.304 0.945
Chain 1: 700 -119918.990 1.254 0.945
Chain 1: 800 -87007.184 1.144 0.945
Chain 1: 900 -67285.272 1.050 0.774
Chain 1: 1000 -52028.407 0.974 0.774
Chain 1: 1100 -39447.051 0.906 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38620.518 0.479 0.378
Chain 1: 1300 -26515.152 0.447 0.378
Chain 1: 1400 -26229.752 0.354 0.319
Chain 1: 1500 -22800.092 0.341 0.319
Chain 1: 1600 -22011.715 0.291 0.293
Chain 1: 1700 -20878.117 0.201 0.293
Chain 1: 1800 -20820.786 0.164 0.150
Chain 1: 1900 -21147.341 0.136 0.054
Chain 1: 2000 -19653.477 0.114 0.054
Chain 1: 2100 -19892.275 0.084 0.036
Chain 1: 2200 -20119.575 0.083 0.036
Chain 1: 2300 -19735.914 0.039 0.019
Chain 1: 2400 -19507.768 0.039 0.019
Chain 1: 2500 -19309.888 0.025 0.015
Chain 1: 2600 -18939.546 0.023 0.015
Chain 1: 2700 -18896.302 0.018 0.012
Chain 1: 2800 -18612.993 0.019 0.015
Chain 1: 2900 -18894.561 0.019 0.015
Chain 1: 3000 -18880.694 0.012 0.012
Chain 1: 3100 -18965.753 0.011 0.012
Chain 1: 3200 -18656.116 0.012 0.015
Chain 1: 3300 -18861.071 0.011 0.012
Chain 1: 3400 -18335.434 0.012 0.015
Chain 1: 3500 -18948.239 0.015 0.015
Chain 1: 3600 -18253.762 0.016 0.015
Chain 1: 3700 -18641.463 0.018 0.017
Chain 1: 3800 -17599.379 0.023 0.021
Chain 1: 3900 -17595.499 0.021 0.021
Chain 1: 4000 -17712.788 0.022 0.021
Chain 1: 4100 -17626.465 0.022 0.021
Chain 1: 4200 -17442.311 0.021 0.021
Chain 1: 4300 -17580.985 0.021 0.021
Chain 1: 4400 -17537.494 0.018 0.011
Chain 1: 4500 -17439.986 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001298 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.98 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13124.173 1.000 1.000
Chain 1: 200 -9836.675 0.667 1.000
Chain 1: 300 -8490.296 0.498 0.334
Chain 1: 400 -8687.279 0.379 0.334
Chain 1: 500 -8567.203 0.306 0.159
Chain 1: 600 -8424.144 0.258 0.159
Chain 1: 700 -8345.941 0.222 0.023
Chain 1: 800 -8435.991 0.196 0.023
Chain 1: 900 -8385.762 0.175 0.017
Chain 1: 1000 -8369.388 0.157 0.017
Chain 1: 1100 -8400.641 0.058 0.014
Chain 1: 1200 -8326.365 0.025 0.011
Chain 1: 1300 -8275.315 0.010 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00281 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 28.1 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -62766.584 1.000 1.000
Chain 1: 200 -18749.254 1.674 2.348
Chain 1: 300 -9400.399 1.447 1.000
Chain 1: 400 -8770.091 1.104 1.000
Chain 1: 500 -9477.368 0.898 0.995
Chain 1: 600 -9423.225 0.749 0.995
Chain 1: 700 -7907.670 0.669 0.192
Chain 1: 800 -8860.591 0.599 0.192
Chain 1: 900 -8214.112 0.541 0.108
Chain 1: 1000 -7944.868 0.491 0.108
Chain 1: 1100 -8137.325 0.393 0.079
Chain 1: 1200 -7926.673 0.161 0.075
Chain 1: 1300 -7715.850 0.064 0.072
Chain 1: 1400 -7924.505 0.060 0.034
Chain 1: 1500 -7769.793 0.054 0.027
Chain 1: 1600 -7935.760 0.056 0.027
Chain 1: 1700 -7556.722 0.042 0.027
Chain 1: 1800 -7739.566 0.033 0.027
Chain 1: 1900 -7674.129 0.026 0.026
Chain 1: 2000 -7808.786 0.024 0.024
Chain 1: 2100 -7718.634 0.023 0.024
Chain 1: 2200 -7903.090 0.023 0.023
Chain 1: 2300 -7712.621 0.023 0.023
Chain 1: 2400 -7656.999 0.021 0.021
Chain 1: 2500 -7654.801 0.019 0.021
Chain 1: 2600 -7664.437 0.017 0.017
Chain 1: 2700 -7657.392 0.012 0.012
Chain 1: 2800 -7762.945 0.011 0.012
Chain 1: 2900 -7490.446 0.014 0.014
Chain 1: 3000 -7632.268 0.014 0.014
Chain 1: 3100 -7636.435 0.013 0.014
Chain 1: 3200 -7878.296 0.013 0.014
Chain 1: 3300 -7515.959 0.016 0.014
Chain 1: 3400 -7693.235 0.017 0.019
Chain 1: 3500 -7570.479 0.019 0.019
Chain 1: 3600 -7592.736 0.019 0.019
Chain 1: 3700 -7511.776 0.020 0.019
Chain 1: 3800 -7505.865 0.019 0.019
Chain 1: 3900 -7517.865 0.015 0.016
Chain 1: 4000 -7516.895 0.013 0.011
Chain 1: 4100 -7515.692 0.013 0.011
Chain 1: 4200 -7611.942 0.012 0.011
Chain 1: 4300 -7499.769 0.008 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003297 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.97 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86988.622 1.000 1.000
Chain 1: 200 -14354.348 3.030 5.060
Chain 1: 300 -10526.454 2.141 1.000
Chain 1: 400 -12638.159 1.648 1.000
Chain 1: 500 -9068.193 1.397 0.394
Chain 1: 600 -9332.101 1.169 0.394
Chain 1: 700 -9090.669 1.006 0.364
Chain 1: 800 -8682.209 0.886 0.364
Chain 1: 900 -8838.017 0.789 0.167
Chain 1: 1000 -9352.203 0.716 0.167
Chain 1: 1100 -9246.067 0.617 0.055 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8745.444 0.117 0.055
Chain 1: 1300 -9164.454 0.085 0.047
Chain 1: 1400 -8988.791 0.070 0.046
Chain 1: 1500 -8941.367 0.031 0.028
Chain 1: 1600 -9065.020 0.030 0.027
Chain 1: 1700 -9107.568 0.028 0.020
Chain 1: 1800 -8640.509 0.028 0.020
Chain 1: 1900 -8756.774 0.028 0.020
Chain 1: 2000 -8775.963 0.023 0.014
Chain 1: 2100 -8863.697 0.023 0.014
Chain 1: 2200 -8636.170 0.019 0.014
Chain 1: 2300 -8821.074 0.017 0.014
Chain 1: 2400 -8661.692 0.017 0.014
Chain 1: 2500 -8725.665 0.017 0.014
Chain 1: 2600 -8633.276 0.017 0.013
Chain 1: 2700 -8668.167 0.017 0.013
Chain 1: 2800 -8623.779 0.012 0.011
Chain 1: 2900 -8734.554 0.012 0.011
Chain 1: 3000 -8642.767 0.013 0.011
Chain 1: 3100 -8610.481 0.012 0.011
Chain 1: 3200 -8579.960 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003788 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 37.88 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8433402.467 1.000 1.000
Chain 1: 200 -1586128.229 2.658 4.317
Chain 1: 300 -891909.386 2.032 1.000
Chain 1: 400 -459156.318 1.759 1.000
Chain 1: 500 -359311.104 1.463 0.942
Chain 1: 600 -234067.777 1.308 0.942
Chain 1: 700 -120152.609 1.257 0.942
Chain 1: 800 -87392.292 1.147 0.942
Chain 1: 900 -67716.423 1.052 0.778
Chain 1: 1000 -52519.662 0.975 0.778
Chain 1: 1100 -39999.990 0.907 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39183.326 0.477 0.375
Chain 1: 1300 -27118.022 0.444 0.375
Chain 1: 1400 -26840.609 0.350 0.313
Chain 1: 1500 -23422.697 0.337 0.313
Chain 1: 1600 -22639.900 0.287 0.291
Chain 1: 1700 -21509.365 0.198 0.289
Chain 1: 1800 -21453.384 0.160 0.146
Chain 1: 1900 -21780.516 0.133 0.053
Chain 1: 2000 -20288.200 0.111 0.053
Chain 1: 2100 -20526.515 0.081 0.035
Chain 1: 2200 -20754.202 0.080 0.035
Chain 1: 2300 -20370.123 0.038 0.019
Chain 1: 2400 -20141.845 0.038 0.019
Chain 1: 2500 -19944.146 0.024 0.015
Chain 1: 2600 -19572.982 0.023 0.015
Chain 1: 2700 -19529.589 0.018 0.012
Chain 1: 2800 -19246.154 0.019 0.015
Chain 1: 2900 -19527.869 0.019 0.014
Chain 1: 3000 -19513.901 0.011 0.012
Chain 1: 3100 -19599.084 0.011 0.011
Chain 1: 3200 -19289.004 0.011 0.014
Chain 1: 3300 -19494.342 0.010 0.011
Chain 1: 3400 -18968.054 0.012 0.014
Chain 1: 3500 -19581.782 0.014 0.015
Chain 1: 3600 -18886.023 0.016 0.015
Chain 1: 3700 -19274.662 0.018 0.016
Chain 1: 3800 -18230.657 0.022 0.020
Chain 1: 3900 -18226.743 0.021 0.020
Chain 1: 4000 -18344.013 0.021 0.020
Chain 1: 4100 -18257.635 0.021 0.020
Chain 1: 4200 -18073.056 0.021 0.020
Chain 1: 4300 -18211.984 0.020 0.020
Chain 1: 4400 -18168.115 0.018 0.010
Chain 1: 4500 -18070.588 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001513 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.13 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48694.544 1.000 1.000
Chain 1: 200 -14471.590 1.682 2.365
Chain 1: 300 -16516.120 1.163 1.000
Chain 1: 400 -13074.528 0.938 1.000
Chain 1: 500 -14692.130 0.772 0.263
Chain 1: 600 -17642.722 0.672 0.263
Chain 1: 700 -12427.876 0.636 0.263
Chain 1: 800 -14644.614 0.575 0.263
Chain 1: 900 -14366.734 0.513 0.167
Chain 1: 1000 -44993.016 0.530 0.263
Chain 1: 1100 -11012.101 0.739 0.263 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -25101.293 0.558 0.263 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1300 -16229.953 0.601 0.420 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1400 -10897.691 0.623 0.489 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1500 -9384.955 0.628 0.489 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1600 -10039.436 0.618 0.489 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1700 -9879.237 0.578 0.489 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1800 -12232.706 0.582 0.489 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1900 -14004.408 0.593 0.489 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2000 -10020.247 0.564 0.398 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2100 -10291.333 0.258 0.192
Chain 1: 2200 -11328.282 0.211 0.161
Chain 1: 2300 -10194.156 0.168 0.127
Chain 1: 2400 -9951.486 0.121 0.111
Chain 1: 2500 -10120.159 0.107 0.092
Chain 1: 2600 -9477.744 0.107 0.092
Chain 1: 2700 -9051.863 0.110 0.092
Chain 1: 2800 -15505.259 0.133 0.092
Chain 1: 2900 -9464.293 0.184 0.092
Chain 1: 3000 -10775.917 0.156 0.092
Chain 1: 3100 -8737.219 0.177 0.111
Chain 1: 3200 -10251.577 0.182 0.122
Chain 1: 3300 -9726.155 0.177 0.122
Chain 1: 3400 -10429.736 0.181 0.122
Chain 1: 3500 -11715.000 0.190 0.122
Chain 1: 3600 -9589.484 0.206 0.148
Chain 1: 3700 -11193.878 0.215 0.148
Chain 1: 3800 -9065.590 0.197 0.148
Chain 1: 3900 -9506.764 0.138 0.143
Chain 1: 4000 -9421.828 0.127 0.143
Chain 1: 4100 -8659.465 0.112 0.110
Chain 1: 4200 -9406.466 0.105 0.088
Chain 1: 4300 -14152.402 0.134 0.110
Chain 1: 4400 -8719.723 0.189 0.143
Chain 1: 4500 -8661.901 0.179 0.143
Chain 1: 4600 -13466.568 0.192 0.143
Chain 1: 4700 -8308.644 0.240 0.235
Chain 1: 4800 -8295.257 0.217 0.088
Chain 1: 4900 -9100.381 0.221 0.088
Chain 1: 5000 -11660.096 0.242 0.220
Chain 1: 5100 -9017.091 0.262 0.293
Chain 1: 5200 -8767.243 0.257 0.293
Chain 1: 5300 -10169.921 0.238 0.220
Chain 1: 5400 -16934.676 0.215 0.220
Chain 1: 5500 -8303.074 0.319 0.293
Chain 1: 5600 -9560.120 0.296 0.220
Chain 1: 5700 -13469.085 0.263 0.220
Chain 1: 5800 -8343.396 0.324 0.290
Chain 1: 5900 -12063.673 0.346 0.293
Chain 1: 6000 -9980.555 0.345 0.293
Chain 1: 6100 -9105.178 0.325 0.290
Chain 1: 6200 -8585.652 0.329 0.290
Chain 1: 6300 -8746.128 0.317 0.290
Chain 1: 6400 -9676.827 0.286 0.209
Chain 1: 6500 -11552.463 0.199 0.162
Chain 1: 6600 -13235.466 0.198 0.162
Chain 1: 6700 -10682.955 0.193 0.162
Chain 1: 6800 -10108.043 0.137 0.127
Chain 1: 6900 -8474.620 0.126 0.127
Chain 1: 7000 -9560.140 0.116 0.114
Chain 1: 7100 -15664.495 0.146 0.127
Chain 1: 7200 -8493.437 0.224 0.162
Chain 1: 7300 -12239.634 0.253 0.193
Chain 1: 7400 -8612.347 0.285 0.239
Chain 1: 7500 -8655.660 0.270 0.239
Chain 1: 7600 -12597.968 0.288 0.306
Chain 1: 7700 -11099.241 0.278 0.306
Chain 1: 7800 -9266.014 0.292 0.306
Chain 1: 7900 -11968.168 0.295 0.306
Chain 1: 8000 -8354.026 0.327 0.313
Chain 1: 8100 -8663.886 0.292 0.306
Chain 1: 8200 -8709.828 0.208 0.226
Chain 1: 8300 -11961.093 0.204 0.226
Chain 1: 8400 -8016.240 0.211 0.226
Chain 1: 8500 -11513.654 0.241 0.272
Chain 1: 8600 -8031.671 0.253 0.272
Chain 1: 8700 -8648.564 0.247 0.272
Chain 1: 8800 -8092.760 0.234 0.272
Chain 1: 8900 -8658.511 0.218 0.272
Chain 1: 9000 -9635.067 0.185 0.101
Chain 1: 9100 -9291.484 0.185 0.101
Chain 1: 9200 -8111.060 0.199 0.146
Chain 1: 9300 -8994.103 0.182 0.101
Chain 1: 9400 -8267.353 0.141 0.098
Chain 1: 9500 -7817.535 0.117 0.088
Chain 1: 9600 -7978.385 0.075 0.071
Chain 1: 9700 -8670.633 0.076 0.080
Chain 1: 9800 -10134.370 0.084 0.088
Chain 1: 9900 -11255.195 0.087 0.098
Chain 1: 10000 -7954.694 0.119 0.098
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00168 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 16.8 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57760.397 1.000 1.000
Chain 1: 200 -17622.486 1.639 2.278
Chain 1: 300 -8626.810 1.440 1.043
Chain 1: 400 -8177.822 1.094 1.043
Chain 1: 500 -7896.357 0.882 1.000
Chain 1: 600 -8040.779 0.738 1.000
Chain 1: 700 -7709.240 0.639 0.055
Chain 1: 800 -7990.791 0.563 0.055
Chain 1: 900 -7796.216 0.504 0.043
Chain 1: 1000 -7618.279 0.456 0.043
Chain 1: 1100 -7780.735 0.358 0.036
Chain 1: 1200 -7536.607 0.133 0.035
Chain 1: 1300 -7726.501 0.031 0.032
Chain 1: 1400 -7823.435 0.027 0.025
Chain 1: 1500 -7570.589 0.027 0.025
Chain 1: 1600 -7482.632 0.026 0.025
Chain 1: 1700 -7443.938 0.022 0.025
Chain 1: 1800 -7613.162 0.021 0.023
Chain 1: 1900 -7580.264 0.019 0.022
Chain 1: 2000 -7558.629 0.017 0.021
Chain 1: 2100 -7487.476 0.016 0.012
Chain 1: 2200 -7658.940 0.015 0.012
Chain 1: 2300 -7519.737 0.014 0.012
Chain 1: 2400 -7607.496 0.014 0.012
Chain 1: 2500 -7445.666 0.013 0.012
Chain 1: 2600 -7482.888 0.012 0.012
Chain 1: 2700 -7501.448 0.012 0.012
Chain 1: 2800 -7502.340 0.010 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.006234 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 62.34 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86187.992 1.000 1.000
Chain 1: 200 -13410.563 3.213 5.427
Chain 1: 300 -9741.192 2.268 1.000
Chain 1: 400 -10594.205 1.721 1.000
Chain 1: 500 -8553.124 1.425 0.377
Chain 1: 600 -8170.903 1.195 0.377
Chain 1: 700 -8190.816 1.025 0.239
Chain 1: 800 -8593.902 0.902 0.239
Chain 1: 900 -8497.199 0.803 0.081
Chain 1: 1000 -8296.236 0.725 0.081
Chain 1: 1100 -8510.237 0.628 0.047 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8033.035 0.091 0.047
Chain 1: 1300 -8312.783 0.057 0.047
Chain 1: 1400 -8354.500 0.049 0.034
Chain 1: 1500 -8283.617 0.026 0.025
Chain 1: 1600 -8385.981 0.023 0.024
Chain 1: 1700 -8454.994 0.023 0.024
Chain 1: 1800 -8025.728 0.024 0.024
Chain 1: 1900 -8129.427 0.024 0.024
Chain 1: 2000 -8104.414 0.022 0.013
Chain 1: 2100 -8233.582 0.021 0.013
Chain 1: 2200 -8030.845 0.018 0.013
Chain 1: 2300 -8126.025 0.016 0.012
Chain 1: 2400 -8192.530 0.016 0.012
Chain 1: 2500 -8138.410 0.016 0.012
Chain 1: 2600 -8141.708 0.015 0.012
Chain 1: 2700 -8057.507 0.015 0.012
Chain 1: 2800 -8015.142 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00381 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 38.1 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8392256.273 1.000 1.000
Chain 1: 200 -1582607.105 2.651 4.303
Chain 1: 300 -889163.790 2.028 1.000
Chain 1: 400 -456645.548 1.757 1.000
Chain 1: 500 -357263.261 1.462 0.947
Chain 1: 600 -232608.478 1.307 0.947
Chain 1: 700 -119059.293 1.257 0.947
Chain 1: 800 -86274.414 1.147 0.947
Chain 1: 900 -66655.599 1.052 0.780
Chain 1: 1000 -51471.751 0.977 0.780
Chain 1: 1100 -38959.514 0.909 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38141.279 0.481 0.380
Chain 1: 1300 -26109.735 0.449 0.380
Chain 1: 1400 -25830.530 0.355 0.321
Chain 1: 1500 -22419.940 0.343 0.321
Chain 1: 1600 -21637.067 0.293 0.295
Chain 1: 1700 -20512.311 0.203 0.294
Chain 1: 1800 -20456.937 0.165 0.152
Chain 1: 1900 -20783.338 0.137 0.055
Chain 1: 2000 -19294.565 0.115 0.055
Chain 1: 2100 -19533.136 0.084 0.036
Chain 1: 2200 -19759.485 0.083 0.036
Chain 1: 2300 -19376.703 0.039 0.020
Chain 1: 2400 -19148.727 0.039 0.020
Chain 1: 2500 -18950.537 0.025 0.016
Chain 1: 2600 -18580.762 0.024 0.016
Chain 1: 2700 -18537.715 0.018 0.012
Chain 1: 2800 -18254.361 0.020 0.016
Chain 1: 2900 -18535.736 0.020 0.015
Chain 1: 3000 -18521.971 0.012 0.012
Chain 1: 3100 -18606.957 0.011 0.012
Chain 1: 3200 -18297.570 0.012 0.015
Chain 1: 3300 -18502.350 0.011 0.012
Chain 1: 3400 -17977.021 0.013 0.015
Chain 1: 3500 -18589.215 0.015 0.016
Chain 1: 3600 -17895.508 0.017 0.016
Chain 1: 3700 -18282.589 0.019 0.017
Chain 1: 3800 -17241.602 0.023 0.021
Chain 1: 3900 -17237.697 0.022 0.021
Chain 1: 4000 -17355.046 0.022 0.021
Chain 1: 4100 -17268.728 0.022 0.021
Chain 1: 4200 -17084.833 0.022 0.021
Chain 1: 4300 -17223.375 0.021 0.021
Chain 1: 4400 -17180.096 0.019 0.011
Chain 1: 4500 -17082.564 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002008 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 20.08 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12578.896 1.000 1.000
Chain 1: 200 -9415.986 0.668 1.000
Chain 1: 300 -8048.325 0.502 0.336
Chain 1: 400 -8177.182 0.380 0.336
Chain 1: 500 -8140.012 0.305 0.170
Chain 1: 600 -7952.056 0.258 0.170
Chain 1: 700 -7905.293 0.222 0.024
Chain 1: 800 -7914.649 0.195 0.024
Chain 1: 900 -7890.272 0.173 0.016
Chain 1: 1000 -7971.578 0.157 0.016
Chain 1: 1100 -8047.486 0.058 0.010
Chain 1: 1200 -7923.621 0.026 0.010
Chain 1: 1300 -7868.908 0.010 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001646 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 16.46 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -55871.262 1.000 1.000
Chain 1: 200 -17265.864 1.618 2.236
Chain 1: 300 -8727.459 1.405 1.000
Chain 1: 400 -8321.787 1.066 1.000
Chain 1: 500 -8436.883 0.855 0.978
Chain 1: 600 -8348.790 0.715 0.978
Chain 1: 700 -7773.416 0.623 0.074
Chain 1: 800 -8141.341 0.551 0.074
Chain 1: 900 -8125.594 0.490 0.049
Chain 1: 1000 -7815.938 0.445 0.049
Chain 1: 1100 -7630.569 0.347 0.045
Chain 1: 1200 -7826.392 0.126 0.040
Chain 1: 1300 -7585.445 0.031 0.032
Chain 1: 1400 -7679.327 0.028 0.025
Chain 1: 1500 -7560.372 0.028 0.025
Chain 1: 1600 -7854.753 0.031 0.032
Chain 1: 1700 -7565.006 0.027 0.032
Chain 1: 1800 -7655.283 0.024 0.025
Chain 1: 1900 -7640.697 0.024 0.025
Chain 1: 2000 -7632.650 0.020 0.024
Chain 1: 2100 -7640.293 0.018 0.016
Chain 1: 2200 -7741.258 0.016 0.013
Chain 1: 2300 -7603.775 0.015 0.013
Chain 1: 2400 -7665.425 0.015 0.013
Chain 1: 2500 -7605.994 0.014 0.012
Chain 1: 2600 -7529.226 0.011 0.010
Chain 1: 2700 -7568.995 0.008 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.005603 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 56.03 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87006.390 1.000 1.000
Chain 1: 200 -13570.089 3.206 5.412
Chain 1: 300 -9907.832 2.260 1.000
Chain 1: 400 -10744.529 1.715 1.000
Chain 1: 500 -8892.567 1.413 0.370
Chain 1: 600 -8559.137 1.184 0.370
Chain 1: 700 -8433.949 1.017 0.208
Chain 1: 800 -8842.085 0.896 0.208
Chain 1: 900 -8657.843 0.799 0.078
Chain 1: 1000 -8612.584 0.719 0.078
Chain 1: 1100 -8743.124 0.621 0.046 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8256.289 0.086 0.046
Chain 1: 1300 -8590.520 0.053 0.039
Chain 1: 1400 -8597.857 0.045 0.039
Chain 1: 1500 -8467.684 0.026 0.021
Chain 1: 1600 -8576.099 0.023 0.015
Chain 1: 1700 -8654.839 0.022 0.015
Chain 1: 1800 -8234.404 0.023 0.015
Chain 1: 1900 -8333.424 0.022 0.015
Chain 1: 2000 -8307.494 0.022 0.015
Chain 1: 2100 -8432.020 0.022 0.015
Chain 1: 2200 -8241.283 0.018 0.015
Chain 1: 2300 -8328.045 0.015 0.013
Chain 1: 2400 -8397.345 0.016 0.013
Chain 1: 2500 -8343.405 0.015 0.012
Chain 1: 2600 -8344.061 0.014 0.010
Chain 1: 2700 -8261.102 0.014 0.010
Chain 1: 2800 -8222.071 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.005508 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 55.08 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8422845.923 1.000 1.000
Chain 1: 200 -1587356.343 2.653 4.306
Chain 1: 300 -891660.811 2.029 1.000
Chain 1: 400 -457721.908 1.759 1.000
Chain 1: 500 -357854.979 1.463 0.948
Chain 1: 600 -232648.375 1.309 0.948
Chain 1: 700 -119097.298 1.258 0.948
Chain 1: 800 -86348.646 1.148 0.948
Chain 1: 900 -66723.468 1.053 0.780
Chain 1: 1000 -51552.775 0.977 0.780
Chain 1: 1100 -39063.366 0.909 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38245.632 0.481 0.379
Chain 1: 1300 -26237.217 0.449 0.379
Chain 1: 1400 -25959.883 0.355 0.320
Chain 1: 1500 -22555.744 0.342 0.320
Chain 1: 1600 -21775.216 0.292 0.294
Chain 1: 1700 -20653.187 0.202 0.294
Chain 1: 1800 -20598.495 0.164 0.151
Chain 1: 1900 -20924.672 0.136 0.054
Chain 1: 2000 -19437.914 0.115 0.054
Chain 1: 2100 -19676.320 0.084 0.036
Chain 1: 2200 -19902.326 0.083 0.036
Chain 1: 2300 -19519.892 0.039 0.020
Chain 1: 2400 -19291.981 0.039 0.020
Chain 1: 2500 -19093.781 0.025 0.016
Chain 1: 2600 -18724.115 0.023 0.016
Chain 1: 2700 -18681.192 0.018 0.012
Chain 1: 2800 -18397.869 0.019 0.015
Chain 1: 2900 -18679.149 0.019 0.015
Chain 1: 3000 -18665.325 0.012 0.012
Chain 1: 3100 -18750.320 0.011 0.012
Chain 1: 3200 -18441.015 0.012 0.015
Chain 1: 3300 -18645.773 0.011 0.012
Chain 1: 3400 -18120.606 0.012 0.015
Chain 1: 3500 -18732.491 0.015 0.015
Chain 1: 3600 -18039.157 0.017 0.015
Chain 1: 3700 -18425.927 0.018 0.017
Chain 1: 3800 -17385.538 0.023 0.021
Chain 1: 3900 -17381.647 0.021 0.021
Chain 1: 4000 -17498.985 0.022 0.021
Chain 1: 4100 -17412.666 0.022 0.021
Chain 1: 4200 -17228.946 0.021 0.021
Chain 1: 4300 -17367.358 0.021 0.021
Chain 1: 4400 -17324.186 0.018 0.011
Chain 1: 4500 -17226.654 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001322 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.22 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48506.426 1.000 1.000
Chain 1: 200 -16491.963 1.471 1.941
Chain 1: 300 -14292.018 1.032 1.000
Chain 1: 400 -19269.436 0.838 1.000
Chain 1: 500 -12411.772 0.781 0.553
Chain 1: 600 -17203.105 0.697 0.553
Chain 1: 700 -16474.365 0.604 0.279
Chain 1: 800 -10737.363 0.595 0.534
Chain 1: 900 -15379.276 0.563 0.302
Chain 1: 1000 -10883.068 0.548 0.413
Chain 1: 1100 -10008.524 0.457 0.302
Chain 1: 1200 -20671.610 0.314 0.302
Chain 1: 1300 -16781.676 0.322 0.302
Chain 1: 1400 -11312.774 0.344 0.413
Chain 1: 1500 -9804.496 0.304 0.302
Chain 1: 1600 -10755.336 0.285 0.302
Chain 1: 1700 -11835.220 0.290 0.302
Chain 1: 1800 -13569.615 0.249 0.232
Chain 1: 1900 -9886.340 0.257 0.232
Chain 1: 2000 -12344.836 0.235 0.199
Chain 1: 2100 -9120.070 0.262 0.232
Chain 1: 2200 -9874.702 0.218 0.199
Chain 1: 2300 -11201.466 0.206 0.154
Chain 1: 2400 -15948.437 0.188 0.154
Chain 1: 2500 -14670.625 0.181 0.128
Chain 1: 2600 -9346.912 0.229 0.199
Chain 1: 2700 -13152.771 0.249 0.289
Chain 1: 2800 -8775.062 0.286 0.298
Chain 1: 2900 -12061.458 0.276 0.289
Chain 1: 3000 -9347.910 0.285 0.290
Chain 1: 3100 -9364.466 0.250 0.289
Chain 1: 3200 -9085.992 0.246 0.289
Chain 1: 3300 -10380.066 0.246 0.289
Chain 1: 3400 -14774.304 0.246 0.289
Chain 1: 3500 -9980.882 0.286 0.290
Chain 1: 3600 -14024.667 0.257 0.289
Chain 1: 3700 -9302.357 0.279 0.290
Chain 1: 3800 -9983.960 0.236 0.288
Chain 1: 3900 -9340.531 0.216 0.288
Chain 1: 4000 -9385.133 0.187 0.125
Chain 1: 4100 -8843.744 0.193 0.125
Chain 1: 4200 -11504.218 0.213 0.231
Chain 1: 4300 -12899.225 0.212 0.231
Chain 1: 4400 -10450.293 0.205 0.231
Chain 1: 4500 -9482.212 0.167 0.108
Chain 1: 4600 -8354.975 0.152 0.108
Chain 1: 4700 -8356.741 0.101 0.102
Chain 1: 4800 -8436.389 0.096 0.102
Chain 1: 4900 -9886.819 0.103 0.108
Chain 1: 5000 -8867.431 0.114 0.115
Chain 1: 5100 -8639.662 0.111 0.115
Chain 1: 5200 -8688.484 0.088 0.108
Chain 1: 5300 -9419.292 0.085 0.102
Chain 1: 5400 -8663.604 0.071 0.087
Chain 1: 5500 -11844.045 0.087 0.087
Chain 1: 5600 -9952.958 0.093 0.087
Chain 1: 5700 -8521.828 0.109 0.115
Chain 1: 5800 -8425.442 0.110 0.115
Chain 1: 5900 -8838.240 0.100 0.087
Chain 1: 6000 -9142.335 0.091 0.078
Chain 1: 6100 -8329.171 0.099 0.087
Chain 1: 6200 -8845.870 0.104 0.087
Chain 1: 6300 -8476.930 0.100 0.087
Chain 1: 6400 -15081.606 0.136 0.098
Chain 1: 6500 -11064.660 0.145 0.098
Chain 1: 6600 -8319.986 0.159 0.098
Chain 1: 6700 -9374.491 0.153 0.098
Chain 1: 6800 -12808.237 0.179 0.112
Chain 1: 6900 -8172.980 0.231 0.268
Chain 1: 7000 -12183.491 0.261 0.329
Chain 1: 7100 -10974.940 0.262 0.329
Chain 1: 7200 -10132.600 0.264 0.329
Chain 1: 7300 -8925.488 0.274 0.329
Chain 1: 7400 -11220.463 0.250 0.268
Chain 1: 7500 -8219.386 0.250 0.268
Chain 1: 7600 -9102.131 0.227 0.205
Chain 1: 7700 -8465.018 0.223 0.205
Chain 1: 7800 -9006.972 0.203 0.135
Chain 1: 7900 -8276.627 0.155 0.110
Chain 1: 8000 -10909.681 0.146 0.110
Chain 1: 8100 -12933.102 0.151 0.135
Chain 1: 8200 -11698.793 0.153 0.135
Chain 1: 8300 -10860.065 0.147 0.106
Chain 1: 8400 -11746.877 0.134 0.097
Chain 1: 8500 -8213.198 0.141 0.097
Chain 1: 8600 -8526.190 0.135 0.088
Chain 1: 8700 -8650.886 0.129 0.088
Chain 1: 8800 -8896.730 0.125 0.088
Chain 1: 8900 -9155.960 0.119 0.077
Chain 1: 9000 -8293.400 0.106 0.077
Chain 1: 9100 -11062.700 0.115 0.077
Chain 1: 9200 -8678.008 0.132 0.077
Chain 1: 9300 -8119.948 0.131 0.075
Chain 1: 9400 -8439.292 0.127 0.069
Chain 1: 9500 -10509.060 0.104 0.069
Chain 1: 9600 -8921.115 0.118 0.104
Chain 1: 9700 -10723.624 0.133 0.168
Chain 1: 9800 -8472.735 0.157 0.178
Chain 1: 9900 -9601.530 0.166 0.178
Chain 1: 10000 -8713.064 0.166 0.178
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001579 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.79 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56765.542 1.000 1.000
Chain 1: 200 -17242.511 1.646 2.292
Chain 1: 300 -8642.422 1.429 1.000
Chain 1: 400 -8071.439 1.090 1.000
Chain 1: 500 -8286.320 0.877 0.995
Chain 1: 600 -7942.584 0.738 0.995
Chain 1: 700 -7949.532 0.633 0.071
Chain 1: 800 -8017.538 0.555 0.071
Chain 1: 900 -8047.895 0.493 0.043
Chain 1: 1000 -7844.306 0.447 0.043
Chain 1: 1100 -7662.718 0.349 0.026
Chain 1: 1200 -7738.644 0.121 0.026
Chain 1: 1300 -7757.206 0.021 0.024
Chain 1: 1400 -7660.537 0.016 0.013
Chain 1: 1500 -7612.089 0.014 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003778 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 37.78 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85913.938 1.000 1.000
Chain 1: 200 -13298.303 3.230 5.461
Chain 1: 300 -9771.560 2.274 1.000
Chain 1: 400 -10641.155 1.726 1.000
Chain 1: 500 -8701.308 1.425 0.361
Chain 1: 600 -8535.652 1.191 0.361
Chain 1: 700 -8560.229 1.021 0.223
Chain 1: 800 -9032.412 0.900 0.223
Chain 1: 900 -8585.904 0.806 0.082
Chain 1: 1000 -8356.245 0.728 0.082
Chain 1: 1100 -8531.091 0.630 0.052 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8292.867 0.087 0.052
Chain 1: 1300 -8350.425 0.051 0.029
Chain 1: 1400 -8441.875 0.044 0.027
Chain 1: 1500 -8375.590 0.023 0.020
Chain 1: 1600 -8383.097 0.021 0.020
Chain 1: 1700 -8319.062 0.022 0.020
Chain 1: 1800 -8199.117 0.018 0.015
Chain 1: 1900 -8315.616 0.014 0.014
Chain 1: 2000 -8275.579 0.012 0.011
Chain 1: 2100 -8410.863 0.011 0.011
Chain 1: 2200 -8198.892 0.011 0.011
Chain 1: 2300 -8339.651 0.012 0.014
Chain 1: 2400 -8350.424 0.011 0.014
Chain 1: 2500 -8318.992 0.011 0.014
Chain 1: 2600 -8315.005 0.011 0.014
Chain 1: 2700 -8224.887 0.011 0.014
Chain 1: 2800 -8204.134 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.005684 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 56.84 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8388904.246 1.000 1.000
Chain 1: 200 -1582089.557 2.651 4.302
Chain 1: 300 -890124.580 2.027 1.000
Chain 1: 400 -456754.393 1.757 1.000
Chain 1: 500 -357416.527 1.461 0.949
Chain 1: 600 -232626.028 1.307 0.949
Chain 1: 700 -118964.755 1.257 0.949
Chain 1: 800 -86181.584 1.147 0.949
Chain 1: 900 -66539.233 1.053 0.777
Chain 1: 1000 -51336.540 0.977 0.777
Chain 1: 1100 -38813.662 0.909 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37986.768 0.481 0.380
Chain 1: 1300 -25953.304 0.450 0.380
Chain 1: 1400 -25670.334 0.356 0.323
Chain 1: 1500 -22260.402 0.344 0.323
Chain 1: 1600 -21476.694 0.294 0.296
Chain 1: 1700 -20352.375 0.204 0.295
Chain 1: 1800 -20296.618 0.166 0.153
Chain 1: 1900 -20622.174 0.138 0.055
Chain 1: 2000 -19135.393 0.116 0.055
Chain 1: 2100 -19373.653 0.085 0.036
Chain 1: 2200 -19599.501 0.084 0.036
Chain 1: 2300 -19217.401 0.040 0.020
Chain 1: 2400 -18989.752 0.040 0.020
Chain 1: 2500 -18791.682 0.025 0.016
Chain 1: 2600 -18422.586 0.024 0.016
Chain 1: 2700 -18379.786 0.018 0.012
Chain 1: 2800 -18096.876 0.020 0.016
Chain 1: 2900 -18377.845 0.020 0.015
Chain 1: 3000 -18364.105 0.012 0.012
Chain 1: 3100 -18448.969 0.011 0.012
Chain 1: 3200 -18140.084 0.012 0.015
Chain 1: 3300 -18344.487 0.011 0.012
Chain 1: 3400 -17820.145 0.013 0.015
Chain 1: 3500 -18430.873 0.015 0.016
Chain 1: 3600 -17739.104 0.017 0.016
Chain 1: 3700 -18124.743 0.019 0.017
Chain 1: 3800 -17086.791 0.023 0.021
Chain 1: 3900 -17083.004 0.022 0.021
Chain 1: 4000 -17200.303 0.022 0.021
Chain 1: 4100 -17114.148 0.022 0.021
Chain 1: 4200 -16930.952 0.022 0.021
Chain 1: 4300 -17068.974 0.021 0.021
Chain 1: 4400 -17026.225 0.019 0.011
Chain 1: 4500 -16928.838 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001373 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.73 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48952.678 1.000 1.000
Chain 1: 200 -20094.494 1.218 1.436
Chain 1: 300 -15008.878 0.925 1.000
Chain 1: 400 -13611.974 0.719 1.000
Chain 1: 500 -15964.642 0.605 0.339
Chain 1: 600 -16757.434 0.512 0.339
Chain 1: 700 -10606.008 0.522 0.339
Chain 1: 800 -13142.049 0.481 0.339
Chain 1: 900 -11117.107 0.447 0.193
Chain 1: 1000 -34724.983 0.471 0.339
Chain 1: 1100 -13805.541 0.522 0.339 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -14006.485 0.380 0.193
Chain 1: 1300 -17514.753 0.366 0.193
Chain 1: 1400 -11343.425 0.410 0.200
Chain 1: 1500 -9612.394 0.414 0.200
Chain 1: 1600 -10070.553 0.413 0.200
Chain 1: 1700 -15050.358 0.389 0.200
Chain 1: 1800 -20399.541 0.395 0.262
Chain 1: 1900 -10454.533 0.472 0.331
Chain 1: 2000 -14828.326 0.434 0.295
Chain 1: 2100 -9561.277 0.337 0.295
Chain 1: 2200 -10518.421 0.345 0.295
Chain 1: 2300 -9281.261 0.338 0.295
Chain 1: 2400 -10084.570 0.292 0.262
Chain 1: 2500 -10252.369 0.276 0.262
Chain 1: 2600 -9177.840 0.283 0.262
Chain 1: 2700 -9003.616 0.252 0.133
Chain 1: 2800 -13499.832 0.259 0.133
Chain 1: 2900 -9404.112 0.207 0.133
Chain 1: 3000 -10797.453 0.191 0.129
Chain 1: 3100 -10112.135 0.142 0.117
Chain 1: 3200 -9146.194 0.144 0.117
Chain 1: 3300 -13333.318 0.162 0.117
Chain 1: 3400 -18015.625 0.180 0.129
Chain 1: 3500 -15069.792 0.198 0.195
Chain 1: 3600 -13953.602 0.194 0.195
Chain 1: 3700 -14772.499 0.198 0.195
Chain 1: 3800 -9875.993 0.214 0.195
Chain 1: 3900 -10415.426 0.175 0.129
Chain 1: 4000 -8607.965 0.184 0.195
Chain 1: 4100 -9907.777 0.190 0.195
Chain 1: 4200 -9354.797 0.185 0.195
Chain 1: 4300 -9009.894 0.158 0.131
Chain 1: 4400 -8691.031 0.135 0.080
Chain 1: 4500 -9655.574 0.126 0.080
Chain 1: 4600 -9212.562 0.123 0.059
Chain 1: 4700 -9066.228 0.119 0.059
Chain 1: 4800 -8883.778 0.071 0.052
Chain 1: 4900 -8498.487 0.071 0.048
Chain 1: 5000 -9301.486 0.058 0.048
Chain 1: 5100 -8688.056 0.052 0.048
Chain 1: 5200 -8691.531 0.046 0.045
Chain 1: 5300 -9505.327 0.051 0.048
Chain 1: 5400 -9175.160 0.051 0.048
Chain 1: 5500 -9142.410 0.041 0.045
Chain 1: 5600 -8559.577 0.043 0.045
Chain 1: 5700 -9234.784 0.049 0.068
Chain 1: 5800 -9799.978 0.053 0.068
Chain 1: 5900 -8684.646 0.061 0.071
Chain 1: 6000 -12078.891 0.080 0.071
Chain 1: 6100 -11632.183 0.077 0.068
Chain 1: 6200 -8762.599 0.110 0.073
Chain 1: 6300 -8531.391 0.104 0.068
Chain 1: 6400 -9855.888 0.114 0.073
Chain 1: 6500 -8666.024 0.127 0.128
Chain 1: 6600 -8366.248 0.124 0.128
Chain 1: 6700 -10603.095 0.138 0.134
Chain 1: 6800 -8964.326 0.150 0.137
Chain 1: 6900 -8856.826 0.139 0.137
Chain 1: 7000 -8847.415 0.111 0.134
Chain 1: 7100 -8958.862 0.108 0.134
Chain 1: 7200 -12228.321 0.102 0.134
Chain 1: 7300 -13867.071 0.111 0.134
Chain 1: 7400 -8374.165 0.163 0.137
Chain 1: 7500 -9663.390 0.163 0.133
Chain 1: 7600 -8607.632 0.172 0.133
Chain 1: 7700 -8727.410 0.152 0.123
Chain 1: 7800 -11429.598 0.157 0.123
Chain 1: 7900 -8105.893 0.197 0.133
Chain 1: 8000 -8183.652 0.198 0.133
Chain 1: 8100 -8365.695 0.199 0.133
Chain 1: 8200 -8773.969 0.177 0.123
Chain 1: 8300 -8253.627 0.171 0.123
Chain 1: 8400 -8876.474 0.113 0.070
Chain 1: 8500 -8310.600 0.106 0.068
Chain 1: 8600 -9788.927 0.109 0.068
Chain 1: 8700 -9358.563 0.112 0.068
Chain 1: 8800 -8679.681 0.096 0.068
Chain 1: 8900 -8572.045 0.057 0.063
Chain 1: 9000 -9809.782 0.068 0.068
Chain 1: 9100 -8711.475 0.079 0.070
Chain 1: 9200 -9069.776 0.078 0.070
Chain 1: 9300 -8689.100 0.076 0.070
Chain 1: 9400 -10956.270 0.090 0.078
Chain 1: 9500 -8165.715 0.117 0.126
Chain 1: 9600 -9550.176 0.117 0.126
Chain 1: 9700 -8864.386 0.120 0.126
Chain 1: 9800 -9006.633 0.113 0.126
Chain 1: 9900 -9222.654 0.115 0.126
Chain 1: 10000 -8068.659 0.116 0.126
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004171 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 41.71 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58119.805 1.000 1.000
Chain 1: 200 -17767.279 1.636 2.271
Chain 1: 300 -8715.376 1.437 1.039
Chain 1: 400 -8137.646 1.095 1.039
Chain 1: 500 -8401.469 0.882 1.000
Chain 1: 600 -8393.060 0.736 1.000
Chain 1: 700 -8022.691 0.637 0.071
Chain 1: 800 -8257.214 0.561 0.071
Chain 1: 900 -8051.683 0.501 0.046
Chain 1: 1000 -7678.388 0.456 0.049
Chain 1: 1100 -7710.598 0.357 0.046
Chain 1: 1200 -7610.881 0.131 0.031
Chain 1: 1300 -7640.207 0.027 0.028
Chain 1: 1400 -7677.884 0.021 0.026
Chain 1: 1500 -7604.139 0.019 0.013
Chain 1: 1600 -7648.551 0.019 0.013
Chain 1: 1700 -7518.741 0.016 0.013
Chain 1: 1800 -7541.845 0.014 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00446 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 44.6 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85695.961 1.000 1.000
Chain 1: 200 -13585.179 3.154 5.308
Chain 1: 300 -9893.604 2.227 1.000
Chain 1: 400 -11026.699 1.696 1.000
Chain 1: 500 -8880.519 1.405 0.373
Chain 1: 600 -8724.566 1.174 0.373
Chain 1: 700 -8247.281 1.014 0.242
Chain 1: 800 -8633.167 0.893 0.242
Chain 1: 900 -8633.633 0.794 0.103
Chain 1: 1000 -8483.735 0.716 0.103
Chain 1: 1100 -8679.400 0.619 0.058 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8178.607 0.094 0.058
Chain 1: 1300 -8525.526 0.061 0.045
Chain 1: 1400 -8512.318 0.051 0.041
Chain 1: 1500 -8433.577 0.027 0.023
Chain 1: 1600 -8536.959 0.027 0.023
Chain 1: 1700 -8598.351 0.022 0.018
Chain 1: 1800 -8166.138 0.023 0.018
Chain 1: 1900 -8270.240 0.024 0.018
Chain 1: 2000 -8245.331 0.022 0.013
Chain 1: 2100 -8208.020 0.021 0.012
Chain 1: 2200 -8187.196 0.015 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003223 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.23 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8383424.705 1.000 1.000
Chain 1: 200 -1580362.166 2.652 4.305
Chain 1: 300 -890739.565 2.026 1.000
Chain 1: 400 -458126.831 1.756 1.000
Chain 1: 500 -358977.664 1.460 0.944
Chain 1: 600 -233944.429 1.306 0.944
Chain 1: 700 -119801.252 1.255 0.944
Chain 1: 800 -86891.766 1.146 0.944
Chain 1: 900 -67150.080 1.051 0.774
Chain 1: 1000 -51879.533 0.975 0.774
Chain 1: 1100 -39289.267 0.907 0.534 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38460.880 0.479 0.379
Chain 1: 1300 -26344.916 0.448 0.379
Chain 1: 1400 -26058.171 0.354 0.320
Chain 1: 1500 -22626.357 0.342 0.320
Chain 1: 1600 -21837.763 0.292 0.294
Chain 1: 1700 -20702.590 0.202 0.294
Chain 1: 1800 -20644.849 0.165 0.152
Chain 1: 1900 -20971.332 0.137 0.055
Chain 1: 2000 -19477.303 0.115 0.055
Chain 1: 2100 -19715.922 0.084 0.036
Chain 1: 2200 -19943.338 0.083 0.036
Chain 1: 2300 -19559.620 0.039 0.020
Chain 1: 2400 -19331.535 0.039 0.020
Chain 1: 2500 -19133.841 0.025 0.016
Chain 1: 2600 -18763.425 0.023 0.016
Chain 1: 2700 -18720.213 0.018 0.012
Chain 1: 2800 -18437.043 0.019 0.015
Chain 1: 2900 -18718.569 0.019 0.015
Chain 1: 3000 -18704.636 0.012 0.012
Chain 1: 3100 -18789.700 0.011 0.012
Chain 1: 3200 -18480.102 0.012 0.015
Chain 1: 3300 -18685.056 0.011 0.012
Chain 1: 3400 -18159.585 0.012 0.015
Chain 1: 3500 -18772.133 0.015 0.015
Chain 1: 3600 -18077.993 0.017 0.015
Chain 1: 3700 -18465.495 0.018 0.017
Chain 1: 3800 -17423.933 0.023 0.021
Chain 1: 3900 -17420.105 0.021 0.021
Chain 1: 4000 -17537.363 0.022 0.021
Chain 1: 4100 -17451.066 0.022 0.021
Chain 1: 4200 -17267.057 0.021 0.021
Chain 1: 4300 -17405.604 0.021 0.021
Chain 1: 4400 -17362.225 0.018 0.011
Chain 1: 4500 -17264.757 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.005859 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 58.59 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12102.686 1.000 1.000
Chain 1: 200 -9089.782 0.666 1.000
Chain 1: 300 -7981.854 0.490 0.331
Chain 1: 400 -8119.290 0.372 0.331
Chain 1: 500 -8040.910 0.299 0.139
Chain 1: 600 -7837.846 0.254 0.139
Chain 1: 700 -7776.435 0.219 0.026
Chain 1: 800 -7787.394 0.192 0.026
Chain 1: 900 -7826.740 0.171 0.017
Chain 1: 1000 -7834.734 0.154 0.017
Chain 1: 1100 -7895.624 0.055 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001488 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.88 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -55390.638 1.000 1.000
Chain 1: 200 -16840.099 1.645 2.289
Chain 1: 300 -8562.248 1.419 1.000
Chain 1: 400 -8797.979 1.071 1.000
Chain 1: 500 -8433.310 0.865 0.967
Chain 1: 600 -8855.094 0.729 0.967
Chain 1: 700 -7770.147 0.645 0.140
Chain 1: 800 -8201.980 0.571 0.140
Chain 1: 900 -7918.153 0.511 0.053
Chain 1: 1000 -7594.493 0.464 0.053
Chain 1: 1100 -7622.517 0.365 0.048
Chain 1: 1200 -7648.229 0.136 0.043
Chain 1: 1300 -7637.087 0.040 0.043
Chain 1: 1400 -7829.873 0.039 0.043
Chain 1: 1500 -7624.583 0.038 0.036
Chain 1: 1600 -7586.159 0.034 0.027
Chain 1: 1700 -7507.489 0.021 0.025
Chain 1: 1800 -7558.847 0.016 0.010
Chain 1: 1900 -7569.267 0.013 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003491 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.91 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86649.293 1.000 1.000
Chain 1: 200 -13177.531 3.288 5.576
Chain 1: 300 -9639.410 2.314 1.000
Chain 1: 400 -10372.122 1.753 1.000
Chain 1: 500 -8561.317 1.445 0.367
Chain 1: 600 -8299.005 1.209 0.367
Chain 1: 700 -8607.702 1.042 0.212
Chain 1: 800 -8549.952 0.912 0.212
Chain 1: 900 -8502.142 0.812 0.071
Chain 1: 1000 -8234.960 0.734 0.071
Chain 1: 1100 -8556.874 0.637 0.038 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8109.427 0.085 0.038
Chain 1: 1300 -8351.473 0.052 0.036
Chain 1: 1400 -8376.435 0.045 0.032
Chain 1: 1500 -8285.850 0.025 0.032
Chain 1: 1600 -8381.449 0.023 0.029
Chain 1: 1700 -8478.628 0.020 0.011
Chain 1: 1800 -8085.073 0.025 0.029
Chain 1: 1900 -8186.585 0.025 0.029
Chain 1: 2000 -8156.789 0.022 0.012
Chain 1: 2100 -8294.639 0.020 0.012
Chain 1: 2200 -8076.622 0.017 0.012
Chain 1: 2300 -8218.537 0.016 0.012
Chain 1: 2400 -8225.991 0.016 0.012
Chain 1: 2500 -8195.608 0.015 0.012
Chain 1: 2600 -8191.901 0.014 0.012
Chain 1: 2700 -8102.554 0.014 0.012
Chain 1: 2800 -8082.120 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003788 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 37.88 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8421669.413 1.000 1.000
Chain 1: 200 -1586063.779 2.655 4.310
Chain 1: 300 -890777.643 2.030 1.000
Chain 1: 400 -457302.148 1.760 1.000
Chain 1: 500 -357239.173 1.464 0.948
Chain 1: 600 -232140.796 1.310 0.948
Chain 1: 700 -118614.785 1.259 0.948
Chain 1: 800 -85880.286 1.149 0.948
Chain 1: 900 -66266.528 1.055 0.781
Chain 1: 1000 -51094.288 0.979 0.781
Chain 1: 1100 -38610.620 0.911 0.539 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37789.474 0.482 0.381
Chain 1: 1300 -25796.478 0.451 0.381
Chain 1: 1400 -25518.753 0.357 0.323
Chain 1: 1500 -22118.901 0.344 0.323
Chain 1: 1600 -21338.712 0.294 0.297
Chain 1: 1700 -20218.988 0.204 0.296
Chain 1: 1800 -20164.487 0.166 0.154
Chain 1: 1900 -20490.017 0.138 0.055
Chain 1: 2000 -19005.809 0.116 0.055
Chain 1: 2100 -19243.998 0.085 0.037
Chain 1: 2200 -19469.362 0.084 0.037
Chain 1: 2300 -19087.699 0.040 0.020
Chain 1: 2400 -18860.045 0.040 0.020
Chain 1: 2500 -18661.866 0.026 0.016
Chain 1: 2600 -18292.869 0.024 0.016
Chain 1: 2700 -18250.191 0.019 0.012
Chain 1: 2800 -17967.139 0.020 0.016
Chain 1: 2900 -18248.116 0.020 0.015
Chain 1: 3000 -18234.380 0.012 0.012
Chain 1: 3100 -18319.223 0.011 0.012
Chain 1: 3200 -18010.409 0.012 0.015
Chain 1: 3300 -18214.797 0.011 0.012
Chain 1: 3400 -17690.472 0.013 0.015
Chain 1: 3500 -18301.087 0.015 0.016
Chain 1: 3600 -17609.447 0.017 0.016
Chain 1: 3700 -17994.920 0.019 0.017
Chain 1: 3800 -16957.139 0.023 0.021
Chain 1: 3900 -16953.323 0.022 0.021
Chain 1: 4000 -17070.658 0.022 0.021
Chain 1: 4100 -16984.451 0.023 0.021
Chain 1: 4200 -16801.320 0.022 0.021
Chain 1: 4300 -16939.311 0.022 0.021
Chain 1: 4400 -16896.587 0.019 0.011
Chain 1: 4500 -16799.168 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001458 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.58 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12299.717 1.000 1.000
Chain 1: 200 -9218.587 0.667 1.000
Chain 1: 300 -8046.688 0.493 0.334
Chain 1: 400 -8171.629 0.374 0.334
Chain 1: 500 -8113.136 0.300 0.146
Chain 1: 600 -7983.247 0.253 0.146
Chain 1: 700 -7913.543 0.218 0.016
Chain 1: 800 -7896.759 0.191 0.016
Chain 1: 900 -7843.900 0.171 0.015
Chain 1: 1000 -7948.386 0.155 0.015
Chain 1: 1100 -7916.612 0.055 0.013
Chain 1: 1200 -7956.253 0.022 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001609 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 16.09 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61614.584 1.000 1.000
Chain 1: 200 -17786.783 1.732 2.464
Chain 1: 300 -8790.983 1.496 1.023
Chain 1: 400 -9141.556 1.131 1.023
Chain 1: 500 -7948.462 0.935 1.000
Chain 1: 600 -8277.232 0.786 1.000
Chain 1: 700 -7847.464 0.681 0.150
Chain 1: 800 -8125.731 0.601 0.150
Chain 1: 900 -7971.868 0.536 0.055
Chain 1: 1000 -7744.914 0.485 0.055
Chain 1: 1100 -7729.040 0.386 0.040
Chain 1: 1200 -7534.533 0.142 0.038
Chain 1: 1300 -7817.275 0.043 0.036
Chain 1: 1400 -7785.132 0.040 0.034
Chain 1: 1500 -7595.366 0.027 0.029
Chain 1: 1600 -7495.032 0.024 0.026
Chain 1: 1700 -7495.747 0.019 0.025
Chain 1: 1800 -7572.211 0.017 0.019
Chain 1: 1900 -7626.455 0.015 0.013
Chain 1: 2000 -7555.093 0.013 0.010
Chain 1: 2100 -7568.769 0.013 0.010
Chain 1: 2200 -7653.748 0.012 0.010
Chain 1: 2300 -7558.865 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.005486 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 54.86 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86867.533 1.000 1.000
Chain 1: 200 -13409.172 3.239 5.478
Chain 1: 300 -9829.749 2.281 1.000
Chain 1: 400 -10689.580 1.731 1.000
Chain 1: 500 -8753.856 1.429 0.364
Chain 1: 600 -8601.051 1.194 0.364
Chain 1: 700 -8456.146 1.026 0.221
Chain 1: 800 -8892.674 0.903 0.221
Chain 1: 900 -8674.767 0.806 0.080
Chain 1: 1000 -8478.581 0.728 0.080
Chain 1: 1100 -8709.775 0.630 0.049 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8489.369 0.085 0.027
Chain 1: 1300 -8585.263 0.050 0.026
Chain 1: 1400 -8583.096 0.042 0.025
Chain 1: 1500 -8434.273 0.021 0.023
Chain 1: 1600 -8549.522 0.021 0.023
Chain 1: 1700 -8631.390 0.020 0.023
Chain 1: 1800 -8239.848 0.020 0.023
Chain 1: 1900 -8341.856 0.019 0.018
Chain 1: 2000 -8312.239 0.017 0.013
Chain 1: 2100 -8438.338 0.016 0.013
Chain 1: 2200 -8224.203 0.016 0.013
Chain 1: 2300 -8370.703 0.016 0.015
Chain 1: 2400 -8386.209 0.016 0.015
Chain 1: 2500 -8352.726 0.015 0.013
Chain 1: 2600 -8354.634 0.014 0.012
Chain 1: 2700 -8261.535 0.014 0.012
Chain 1: 2800 -8234.612 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003083 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.83 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8427405.189 1.000 1.000
Chain 1: 200 -1590521.866 2.649 4.299
Chain 1: 300 -892611.329 2.027 1.000
Chain 1: 400 -458307.432 1.757 1.000
Chain 1: 500 -358008.132 1.462 0.948
Chain 1: 600 -232778.234 1.308 0.948
Chain 1: 700 -119009.062 1.257 0.948
Chain 1: 800 -86224.824 1.148 0.948
Chain 1: 900 -66580.342 1.053 0.782
Chain 1: 1000 -51389.859 0.977 0.782
Chain 1: 1100 -38885.803 0.909 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38060.965 0.482 0.380
Chain 1: 1300 -26046.934 0.450 0.380
Chain 1: 1400 -25767.099 0.356 0.322
Chain 1: 1500 -22362.437 0.343 0.322
Chain 1: 1600 -21580.668 0.293 0.296
Chain 1: 1700 -20458.500 0.203 0.295
Chain 1: 1800 -20403.319 0.165 0.152
Chain 1: 1900 -20729.015 0.137 0.055
Chain 1: 2000 -19243.112 0.115 0.055
Chain 1: 2100 -19481.341 0.084 0.036
Chain 1: 2200 -19707.135 0.083 0.036
Chain 1: 2300 -19325.035 0.039 0.020
Chain 1: 2400 -19097.310 0.039 0.020
Chain 1: 2500 -18899.191 0.025 0.016
Chain 1: 2600 -18529.962 0.024 0.016
Chain 1: 2700 -18487.095 0.018 0.012
Chain 1: 2800 -18204.070 0.020 0.016
Chain 1: 2900 -18485.098 0.020 0.015
Chain 1: 3000 -18471.366 0.012 0.012
Chain 1: 3100 -18556.270 0.011 0.012
Chain 1: 3200 -18247.262 0.012 0.015
Chain 1: 3300 -18451.745 0.011 0.012
Chain 1: 3400 -17927.175 0.013 0.015
Chain 1: 3500 -18538.212 0.015 0.016
Chain 1: 3600 -17845.991 0.017 0.016
Chain 1: 3700 -18231.938 0.019 0.017
Chain 1: 3800 -17193.308 0.023 0.021
Chain 1: 3900 -17189.463 0.022 0.021
Chain 1: 4000 -17306.798 0.022 0.021
Chain 1: 4100 -17220.616 0.022 0.021
Chain 1: 4200 -17037.225 0.022 0.021
Chain 1: 4300 -17175.373 0.021 0.021
Chain 1: 4400 -17132.495 0.019 0.011
Chain 1: 4500 -17035.066 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001299 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.99 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48595.680 1.000 1.000
Chain 1: 200 -15887.236 1.529 2.059
Chain 1: 300 -18745.450 1.070 1.000
Chain 1: 400 -14723.208 0.871 1.000
Chain 1: 500 -22670.374 0.767 0.351
Chain 1: 600 -11947.175 0.789 0.898
Chain 1: 700 -16583.187 0.716 0.351
Chain 1: 800 -12531.395 0.667 0.351
Chain 1: 900 -14446.098 0.608 0.323
Chain 1: 1000 -11949.836 0.568 0.323
Chain 1: 1100 -15022.065 0.488 0.280
Chain 1: 1200 -12485.531 0.303 0.273
Chain 1: 1300 -10776.171 0.303 0.273
Chain 1: 1400 -18744.788 0.318 0.280
Chain 1: 1500 -10259.529 0.366 0.280
Chain 1: 1600 -10524.343 0.279 0.209
Chain 1: 1700 -9619.261 0.260 0.205
Chain 1: 1800 -17170.403 0.272 0.205
Chain 1: 1900 -9523.842 0.339 0.209
Chain 1: 2000 -9371.236 0.320 0.205
Chain 1: 2100 -11433.837 0.317 0.203
Chain 1: 2200 -10004.905 0.311 0.180
Chain 1: 2300 -11454.587 0.308 0.180
Chain 1: 2400 -17160.781 0.299 0.180
Chain 1: 2500 -9991.499 0.288 0.180
Chain 1: 2600 -8804.020 0.299 0.180
Chain 1: 2700 -9291.558 0.295 0.180
Chain 1: 2800 -9235.969 0.251 0.143
Chain 1: 2900 -9530.293 0.174 0.135
Chain 1: 3000 -10173.823 0.179 0.135
Chain 1: 3100 -9451.043 0.168 0.127
Chain 1: 3200 -9014.681 0.159 0.076
Chain 1: 3300 -9536.567 0.152 0.063
Chain 1: 3400 -9180.627 0.122 0.055
Chain 1: 3500 -9386.097 0.053 0.052
Chain 1: 3600 -8864.150 0.045 0.052
Chain 1: 3700 -9412.033 0.046 0.055
Chain 1: 3800 -10581.193 0.056 0.058
Chain 1: 3900 -10083.204 0.058 0.058
Chain 1: 4000 -9317.613 0.060 0.058
Chain 1: 4100 -9041.636 0.055 0.055
Chain 1: 4200 -13845.252 0.085 0.058
Chain 1: 4300 -9215.267 0.130 0.059
Chain 1: 4400 -8761.111 0.131 0.059
Chain 1: 4500 -14125.918 0.167 0.082
Chain 1: 4600 -13224.937 0.168 0.082
Chain 1: 4700 -9156.972 0.207 0.110
Chain 1: 4800 -11925.882 0.219 0.232
Chain 1: 4900 -11181.385 0.220 0.232
Chain 1: 5000 -10134.733 0.223 0.232
Chain 1: 5100 -8778.013 0.235 0.232
Chain 1: 5200 -8691.644 0.201 0.155
Chain 1: 5300 -8824.201 0.153 0.103
Chain 1: 5400 -10770.508 0.165 0.155
Chain 1: 5500 -13077.134 0.145 0.155
Chain 1: 5600 -8492.909 0.192 0.176
Chain 1: 5700 -9010.066 0.154 0.155
Chain 1: 5800 -9740.214 0.138 0.103
Chain 1: 5900 -11216.039 0.144 0.132
Chain 1: 6000 -9010.360 0.159 0.155
Chain 1: 6100 -11380.367 0.164 0.176
Chain 1: 6200 -8391.266 0.199 0.181
Chain 1: 6300 -8499.508 0.198 0.181
Chain 1: 6400 -11747.163 0.208 0.208
Chain 1: 6500 -9311.975 0.216 0.245
Chain 1: 6600 -12600.240 0.188 0.245
Chain 1: 6700 -9078.937 0.222 0.261
Chain 1: 6800 -9521.589 0.219 0.261
Chain 1: 6900 -10237.781 0.213 0.261
Chain 1: 7000 -12205.269 0.204 0.261
Chain 1: 7100 -8136.349 0.233 0.262
Chain 1: 7200 -11824.840 0.229 0.262
Chain 1: 7300 -8445.937 0.268 0.276
Chain 1: 7400 -8086.218 0.244 0.262
Chain 1: 7500 -8944.605 0.228 0.261
Chain 1: 7600 -9031.984 0.203 0.161
Chain 1: 7700 -11191.091 0.183 0.161
Chain 1: 7800 -8291.357 0.214 0.193
Chain 1: 7900 -11301.278 0.233 0.266
Chain 1: 8000 -8402.299 0.252 0.312
Chain 1: 8100 -8354.322 0.202 0.266
Chain 1: 8200 -8869.498 0.177 0.193
Chain 1: 8300 -10308.424 0.151 0.140
Chain 1: 8400 -10143.451 0.148 0.140
Chain 1: 8500 -9033.158 0.151 0.140
Chain 1: 8600 -8701.910 0.153 0.140
Chain 1: 8700 -8063.042 0.142 0.123
Chain 1: 8800 -8693.153 0.114 0.079
Chain 1: 8900 -8515.403 0.090 0.072
Chain 1: 9000 -9454.455 0.065 0.072
Chain 1: 9100 -8112.701 0.081 0.079
Chain 1: 9200 -9812.025 0.093 0.099
Chain 1: 9300 -8104.388 0.100 0.099
Chain 1: 9400 -8199.047 0.099 0.099
Chain 1: 9500 -8331.689 0.089 0.079
Chain 1: 9600 -8488.132 0.087 0.079
Chain 1: 9700 -8633.890 0.080 0.072
Chain 1: 9800 -8716.282 0.074 0.021
Chain 1: 9900 -8371.261 0.076 0.041
Chain 1: 10000 -8256.514 0.068 0.018
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001609 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 16.09 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61633.621 1.000 1.000
Chain 1: 200 -17595.580 1.751 2.503
Chain 1: 300 -8723.508 1.507 1.017
Chain 1: 400 -8965.646 1.137 1.017
Chain 1: 500 -7925.265 0.936 1.000
Chain 1: 600 -8667.353 0.794 1.000
Chain 1: 700 -8172.732 0.689 0.131
Chain 1: 800 -8255.104 0.604 0.131
Chain 1: 900 -7867.400 0.543 0.086
Chain 1: 1000 -7852.932 0.489 0.086
Chain 1: 1100 -7669.308 0.391 0.061
Chain 1: 1200 -7573.113 0.142 0.049
Chain 1: 1300 -7745.476 0.042 0.027
Chain 1: 1400 -7659.316 0.041 0.024
Chain 1: 1500 -7593.246 0.029 0.022
Chain 1: 1600 -7503.530 0.021 0.013
Chain 1: 1700 -7498.438 0.015 0.012
Chain 1: 1800 -7577.683 0.015 0.012
Chain 1: 1900 -7462.726 0.012 0.012
Chain 1: 2000 -7558.497 0.013 0.013
Chain 1: 2100 -7595.924 0.011 0.012
Chain 1: 2200 -7683.975 0.011 0.011
Chain 1: 2300 -7567.361 0.010 0.011
Chain 1: 2400 -7582.972 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003597 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.97 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86322.613 1.000 1.000
Chain 1: 200 -13249.300 3.258 5.515
Chain 1: 300 -9706.912 2.293 1.000
Chain 1: 400 -10621.147 1.742 1.000
Chain 1: 500 -8613.265 1.440 0.365
Chain 1: 600 -8290.675 1.206 0.365
Chain 1: 700 -8369.188 1.035 0.233
Chain 1: 800 -8949.586 0.914 0.233
Chain 1: 900 -8600.894 0.817 0.086
Chain 1: 1000 -8285.336 0.739 0.086
Chain 1: 1100 -8409.278 0.641 0.065 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8235.890 0.091 0.041
Chain 1: 1300 -8461.459 0.057 0.039
Chain 1: 1400 -8439.843 0.049 0.038
Chain 1: 1500 -8348.325 0.027 0.027
Chain 1: 1600 -8442.495 0.024 0.021
Chain 1: 1700 -8536.770 0.024 0.021
Chain 1: 1800 -8149.410 0.022 0.021
Chain 1: 1900 -8251.429 0.020 0.015
Chain 1: 2000 -8221.456 0.016 0.012
Chain 1: 2100 -8357.881 0.016 0.012
Chain 1: 2200 -8140.879 0.017 0.012
Chain 1: 2300 -8282.416 0.016 0.012
Chain 1: 2400 -8291.896 0.016 0.012
Chain 1: 2500 -8260.306 0.015 0.012
Chain 1: 2600 -8257.723 0.014 0.012
Chain 1: 2700 -8167.668 0.014 0.012
Chain 1: 2800 -8146.627 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00403 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 40.3 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8416341.150 1.000 1.000
Chain 1: 200 -1589107.381 2.648 4.296
Chain 1: 300 -891520.074 2.026 1.000
Chain 1: 400 -457569.398 1.757 1.000
Chain 1: 500 -357599.799 1.461 0.948
Chain 1: 600 -232360.748 1.308 0.948
Chain 1: 700 -118718.004 1.258 0.948
Chain 1: 800 -85968.241 1.148 0.948
Chain 1: 900 -66354.057 1.053 0.782
Chain 1: 1000 -51186.816 0.978 0.782
Chain 1: 1100 -38698.396 0.910 0.539 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37872.277 0.482 0.381
Chain 1: 1300 -25875.938 0.451 0.381
Chain 1: 1400 -25595.590 0.357 0.323
Chain 1: 1500 -22196.333 0.344 0.323
Chain 1: 1600 -21415.555 0.294 0.296
Chain 1: 1700 -20295.928 0.204 0.296
Chain 1: 1800 -20241.088 0.166 0.153
Chain 1: 1900 -20566.636 0.138 0.055
Chain 1: 2000 -19082.171 0.116 0.055
Chain 1: 2100 -19320.241 0.085 0.036
Chain 1: 2200 -19545.867 0.084 0.036
Chain 1: 2300 -19163.962 0.040 0.020
Chain 1: 2400 -18936.360 0.040 0.020
Chain 1: 2500 -18738.175 0.025 0.016
Chain 1: 2600 -18369.264 0.024 0.016
Chain 1: 2700 -18326.387 0.019 0.012
Chain 1: 2800 -18043.523 0.020 0.016
Chain 1: 2900 -18324.343 0.020 0.015
Chain 1: 3000 -18310.623 0.012 0.012
Chain 1: 3100 -18395.569 0.011 0.012
Chain 1: 3200 -18086.683 0.012 0.015
Chain 1: 3300 -18291.004 0.011 0.012
Chain 1: 3400 -17766.717 0.013 0.015
Chain 1: 3500 -18377.398 0.015 0.016
Chain 1: 3600 -17685.557 0.017 0.016
Chain 1: 3700 -18071.280 0.019 0.017
Chain 1: 3800 -17033.300 0.023 0.021
Chain 1: 3900 -17029.469 0.022 0.021
Chain 1: 4000 -17146.785 0.022 0.021
Chain 1: 4100 -17060.735 0.022 0.021
Chain 1: 4200 -16877.418 0.022 0.021
Chain 1: 4300 -17015.511 0.022 0.021
Chain 1: 4400 -16972.744 0.019 0.011
Chain 1: 4500 -16875.336 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001347 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.47 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48373.338 1.000 1.000
Chain 1: 200 -19909.739 1.215 1.430
Chain 1: 300 -18783.035 0.830 1.000
Chain 1: 400 -112043.581 0.830 1.000
Chain 1: 500 -14031.398 2.061 1.000
Chain 1: 600 -16968.297 1.747 1.000
Chain 1: 700 -11734.501 1.561 0.832
Chain 1: 800 -13395.751 1.381 0.832
Chain 1: 900 -10574.395 1.257 0.446
Chain 1: 1000 -11633.362 1.141 0.446
Chain 1: 1100 -11225.473 1.044 0.267 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -15518.022 0.929 0.267 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1300 -12096.200 0.951 0.277 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1400 -10934.656 0.879 0.267 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1500 -11621.910 0.186 0.173
Chain 1: 1600 -9506.100 0.191 0.223
Chain 1: 1700 -16844.972 0.190 0.223
Chain 1: 1800 -17528.882 0.182 0.223
Chain 1: 1900 -22406.228 0.177 0.218
Chain 1: 2000 -10256.942 0.286 0.223
Chain 1: 2100 -11110.370 0.290 0.223
Chain 1: 2200 -10738.054 0.266 0.218
Chain 1: 2300 -9254.715 0.254 0.160
Chain 1: 2400 -9179.201 0.244 0.160
Chain 1: 2500 -17441.413 0.285 0.218
Chain 1: 2600 -17829.005 0.265 0.160
Chain 1: 2700 -8935.084 0.321 0.160
Chain 1: 2800 -13274.664 0.350 0.218
Chain 1: 2900 -9037.268 0.375 0.327
Chain 1: 3000 -11529.920 0.278 0.216
Chain 1: 3100 -9095.773 0.297 0.268
Chain 1: 3200 -12226.085 0.319 0.268
Chain 1: 3300 -9933.231 0.327 0.268
Chain 1: 3400 -15804.514 0.363 0.327
Chain 1: 3500 -8702.638 0.397 0.327
Chain 1: 3600 -9039.101 0.399 0.327
Chain 1: 3700 -15114.875 0.339 0.327
Chain 1: 3800 -9883.976 0.360 0.371
Chain 1: 3900 -8722.198 0.326 0.268
Chain 1: 4000 -8434.013 0.308 0.268
Chain 1: 4100 -8650.927 0.284 0.256
Chain 1: 4200 -9597.740 0.268 0.231
Chain 1: 4300 -10211.186 0.251 0.133
Chain 1: 4400 -8774.553 0.230 0.133
Chain 1: 4500 -8848.280 0.149 0.099
Chain 1: 4600 -12041.599 0.172 0.133
Chain 1: 4700 -9846.415 0.154 0.133
Chain 1: 4800 -8363.050 0.119 0.133
Chain 1: 4900 -8765.612 0.110 0.099
Chain 1: 5000 -10181.027 0.121 0.139
Chain 1: 5100 -8686.205 0.135 0.164
Chain 1: 5200 -15302.420 0.169 0.172
Chain 1: 5300 -11669.803 0.194 0.177
Chain 1: 5400 -14926.650 0.199 0.218
Chain 1: 5500 -9111.562 0.262 0.223
Chain 1: 5600 -8126.657 0.248 0.218
Chain 1: 5700 -8819.617 0.233 0.177
Chain 1: 5800 -8385.949 0.221 0.172
Chain 1: 5900 -12852.804 0.251 0.218
Chain 1: 6000 -10522.292 0.259 0.221
Chain 1: 6100 -8353.859 0.268 0.260
Chain 1: 6200 -8448.366 0.226 0.221
Chain 1: 6300 -8136.559 0.199 0.218
Chain 1: 6400 -13747.060 0.218 0.221
Chain 1: 6500 -9671.678 0.196 0.221
Chain 1: 6600 -8547.953 0.197 0.221
Chain 1: 6700 -8072.221 0.195 0.221
Chain 1: 6800 -10245.776 0.211 0.221
Chain 1: 6900 -9814.953 0.181 0.212
Chain 1: 7000 -9838.939 0.159 0.131
Chain 1: 7100 -8139.743 0.154 0.131
Chain 1: 7200 -9396.228 0.166 0.134
Chain 1: 7300 -8448.465 0.173 0.134
Chain 1: 7400 -10443.955 0.152 0.134
Chain 1: 7500 -8467.702 0.133 0.134
Chain 1: 7600 -8372.775 0.121 0.134
Chain 1: 7700 -8248.615 0.116 0.134
Chain 1: 7800 -10887.050 0.119 0.134
Chain 1: 7900 -7966.425 0.152 0.191
Chain 1: 8000 -8100.457 0.153 0.191
Chain 1: 8100 -9404.399 0.146 0.139
Chain 1: 8200 -9957.667 0.138 0.139
Chain 1: 8300 -7943.425 0.152 0.191
Chain 1: 8400 -7950.376 0.133 0.139
Chain 1: 8500 -8016.335 0.111 0.056
Chain 1: 8600 -11220.868 0.138 0.139
Chain 1: 8700 -10090.372 0.148 0.139
Chain 1: 8800 -9146.080 0.134 0.112
Chain 1: 8900 -11084.622 0.115 0.112
Chain 1: 9000 -10034.072 0.124 0.112
Chain 1: 9100 -9660.408 0.114 0.105
Chain 1: 9200 -8427.540 0.123 0.112
Chain 1: 9300 -8046.952 0.102 0.105
Chain 1: 9400 -8196.863 0.104 0.105
Chain 1: 9500 -10980.093 0.128 0.112
Chain 1: 9600 -8055.810 0.136 0.112
Chain 1: 9700 -8101.199 0.126 0.105
Chain 1: 9800 -8665.404 0.122 0.105
Chain 1: 9900 -9154.114 0.110 0.065
Chain 1: 10000 -7818.997 0.116 0.065
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.0014 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56662.921 1.000 1.000
Chain 1: 200 -17072.676 1.659 2.319
Chain 1: 300 -8487.897 1.443 1.011
Chain 1: 400 -8911.992 1.094 1.011
Chain 1: 500 -8474.681 0.886 1.000
Chain 1: 600 -8603.328 0.741 1.000
Chain 1: 700 -7658.375 0.653 0.123
Chain 1: 800 -7927.658 0.575 0.123
Chain 1: 900 -7787.679 0.513 0.052
Chain 1: 1000 -7630.018 0.464 0.052
Chain 1: 1100 -7511.110 0.366 0.048
Chain 1: 1200 -7472.696 0.134 0.034
Chain 1: 1300 -7590.871 0.035 0.021
Chain 1: 1400 -7786.733 0.032 0.021
Chain 1: 1500 -7488.838 0.031 0.021
Chain 1: 1600 -7467.633 0.030 0.021
Chain 1: 1700 -7390.739 0.019 0.018
Chain 1: 1800 -7455.698 0.016 0.016
Chain 1: 1900 -7508.497 0.015 0.016
Chain 1: 2000 -7503.066 0.013 0.010
Chain 1: 2100 -7525.261 0.012 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003963 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 39.63 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85195.125 1.000 1.000
Chain 1: 200 -13069.778 3.259 5.518
Chain 1: 300 -9541.381 2.296 1.000
Chain 1: 400 -10384.067 1.742 1.000
Chain 1: 500 -8445.869 1.440 0.370
Chain 1: 600 -8108.951 1.207 0.370
Chain 1: 700 -8332.054 1.038 0.229
Chain 1: 800 -8488.854 0.911 0.229
Chain 1: 900 -8414.814 0.811 0.081
Chain 1: 1000 -8168.628 0.732 0.081
Chain 1: 1100 -8442.188 0.636 0.042 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8183.752 0.087 0.032
Chain 1: 1300 -8292.785 0.051 0.032
Chain 1: 1400 -8309.378 0.043 0.030
Chain 1: 1500 -8191.038 0.022 0.027
Chain 1: 1600 -8282.397 0.019 0.018
Chain 1: 1700 -8377.433 0.017 0.014
Chain 1: 1800 -7993.855 0.020 0.014
Chain 1: 1900 -8095.236 0.021 0.014
Chain 1: 2000 -8064.964 0.018 0.013
Chain 1: 2100 -8204.375 0.016 0.013
Chain 1: 2200 -7985.519 0.016 0.013
Chain 1: 2300 -8127.823 0.016 0.014
Chain 1: 2400 -8013.779 0.018 0.014
Chain 1: 2500 -8071.658 0.017 0.014
Chain 1: 2600 -8085.637 0.016 0.014
Chain 1: 2700 -8007.821 0.016 0.014
Chain 1: 2800 -7989.655 0.011 0.013
Chain 1: 2900 -8001.612 0.010 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00625 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 62.5 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8384213.760 1.000 1.000
Chain 1: 200 -1583857.606 2.647 4.294
Chain 1: 300 -892003.513 2.023 1.000
Chain 1: 400 -458065.318 1.754 1.000
Chain 1: 500 -358569.637 1.459 0.947
Chain 1: 600 -233328.206 1.305 0.947
Chain 1: 700 -119166.022 1.256 0.947
Chain 1: 800 -86243.005 1.146 0.947
Chain 1: 900 -66514.033 1.052 0.776
Chain 1: 1000 -51247.863 0.976 0.776
Chain 1: 1100 -38670.099 0.909 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37831.033 0.482 0.382
Chain 1: 1300 -25750.588 0.451 0.382
Chain 1: 1400 -25461.210 0.358 0.325
Chain 1: 1500 -22039.573 0.345 0.325
Chain 1: 1600 -21251.638 0.295 0.298
Chain 1: 1700 -20122.520 0.205 0.297
Chain 1: 1800 -20065.193 0.167 0.155
Chain 1: 1900 -20390.673 0.139 0.056
Chain 1: 2000 -18901.130 0.117 0.056
Chain 1: 2100 -19139.496 0.086 0.037
Chain 1: 2200 -19365.817 0.085 0.037
Chain 1: 2300 -18983.307 0.040 0.020
Chain 1: 2400 -18755.643 0.040 0.020
Chain 1: 2500 -18557.691 0.026 0.016
Chain 1: 2600 -18188.579 0.024 0.016
Chain 1: 2700 -18145.634 0.019 0.012
Chain 1: 2800 -17862.895 0.020 0.016
Chain 1: 2900 -18143.813 0.020 0.015
Chain 1: 3000 -18130.074 0.012 0.012
Chain 1: 3100 -18214.979 0.011 0.012
Chain 1: 3200 -17906.084 0.012 0.015
Chain 1: 3300 -18110.405 0.011 0.012
Chain 1: 3400 -17586.142 0.013 0.015
Chain 1: 3500 -18196.885 0.015 0.016
Chain 1: 3600 -17505.054 0.017 0.016
Chain 1: 3700 -17890.824 0.019 0.017
Chain 1: 3800 -16852.855 0.024 0.022
Chain 1: 3900 -16849.070 0.022 0.022
Chain 1: 4000 -16966.352 0.023 0.022
Chain 1: 4100 -16880.300 0.023 0.022
Chain 1: 4200 -16696.988 0.022 0.022
Chain 1: 4300 -16835.054 0.022 0.022
Chain 1: 4400 -16792.312 0.019 0.011
Chain 1: 4500 -16694.938 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001493 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.93 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12449.999 1.000 1.000
Chain 1: 200 -9325.446 0.668 1.000
Chain 1: 300 -8203.795 0.491 0.335
Chain 1: 400 -8384.191 0.373 0.335
Chain 1: 500 -8209.204 0.303 0.137
Chain 1: 600 -8137.607 0.254 0.137
Chain 1: 700 -8053.843 0.219 0.022
Chain 1: 800 -8130.006 0.193 0.022
Chain 1: 900 -7961.503 0.174 0.021
Chain 1: 1000 -8081.681 0.158 0.021
Chain 1: 1100 -8088.701 0.058 0.021
Chain 1: 1200 -8067.328 0.025 0.015
Chain 1: 1300 -8149.947 0.012 0.010
Chain 1: 1400 -8048.027 0.011 0.010
Chain 1: 1500 -8135.073 0.010 0.010
Chain 1: 1600 -8087.300 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001399 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.99 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58375.629 1.000 1.000
Chain 1: 200 -17738.828 1.645 2.291
Chain 1: 300 -8714.197 1.442 1.036
Chain 1: 400 -8216.951 1.097 1.036
Chain 1: 500 -8537.291 0.885 1.000
Chain 1: 600 -8403.181 0.740 1.000
Chain 1: 700 -8004.641 0.641 0.061
Chain 1: 800 -8147.697 0.563 0.061
Chain 1: 900 -7930.553 0.504 0.050
Chain 1: 1000 -7761.096 0.456 0.050
Chain 1: 1100 -7830.658 0.357 0.038
Chain 1: 1200 -7632.902 0.130 0.027
Chain 1: 1300 -7824.028 0.029 0.026
Chain 1: 1400 -7889.593 0.024 0.024
Chain 1: 1500 -7607.859 0.024 0.024
Chain 1: 1600 -7731.285 0.024 0.024
Chain 1: 1700 -7533.595 0.021 0.024
Chain 1: 1800 -7652.918 0.021 0.024
Chain 1: 1900 -7651.767 0.018 0.022
Chain 1: 2000 -7668.597 0.016 0.016
Chain 1: 2100 -7629.011 0.016 0.016
Chain 1: 2200 -7716.693 0.015 0.016
Chain 1: 2300 -7611.825 0.014 0.014
Chain 1: 2400 -7664.832 0.013 0.014
Chain 1: 2500 -7569.136 0.011 0.013
Chain 1: 2600 -7554.342 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003384 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.84 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86213.188 1.000 1.000
Chain 1: 200 -13548.308 3.182 5.363
Chain 1: 300 -9973.945 2.241 1.000
Chain 1: 400 -10772.101 1.699 1.000
Chain 1: 500 -8920.539 1.401 0.358
Chain 1: 600 -8463.691 1.176 0.358
Chain 1: 700 -8836.419 1.014 0.208
Chain 1: 800 -9267.535 0.893 0.208
Chain 1: 900 -8839.216 0.799 0.074
Chain 1: 1000 -8584.764 0.722 0.074
Chain 1: 1100 -8793.180 0.625 0.054 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8467.187 0.092 0.048
Chain 1: 1300 -8687.455 0.059 0.047
Chain 1: 1400 -8682.916 0.052 0.042
Chain 1: 1500 -8580.424 0.032 0.039
Chain 1: 1600 -8681.236 0.028 0.030
Chain 1: 1700 -8770.441 0.025 0.025
Chain 1: 1800 -8370.225 0.025 0.025
Chain 1: 1900 -8470.888 0.021 0.024
Chain 1: 2000 -8441.820 0.018 0.012
Chain 1: 2100 -8562.271 0.018 0.012
Chain 1: 2200 -8338.797 0.016 0.012
Chain 1: 2300 -8500.069 0.016 0.012
Chain 1: 2400 -8512.174 0.016 0.012
Chain 1: 2500 -8483.900 0.015 0.012
Chain 1: 2600 -8487.258 0.014 0.012
Chain 1: 2700 -8392.391 0.014 0.012
Chain 1: 2800 -8362.528 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003982 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 39.82 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8431779.964 1.000 1.000
Chain 1: 200 -1588000.385 2.655 4.310
Chain 1: 300 -890658.482 2.031 1.000
Chain 1: 400 -457848.478 1.759 1.000
Chain 1: 500 -357755.423 1.464 0.945
Chain 1: 600 -232639.985 1.309 0.945
Chain 1: 700 -119038.556 1.259 0.945
Chain 1: 800 -86309.334 1.149 0.945
Chain 1: 900 -66690.467 1.054 0.783
Chain 1: 1000 -51518.506 0.978 0.783
Chain 1: 1100 -39034.532 0.910 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38210.228 0.481 0.379
Chain 1: 1300 -26210.563 0.448 0.379
Chain 1: 1400 -25932.074 0.355 0.320
Chain 1: 1500 -22531.764 0.342 0.320
Chain 1: 1600 -21751.723 0.292 0.294
Chain 1: 1700 -20630.874 0.202 0.294
Chain 1: 1800 -20576.154 0.164 0.151
Chain 1: 1900 -20902.017 0.136 0.054
Chain 1: 2000 -19416.734 0.115 0.054
Chain 1: 2100 -19654.747 0.084 0.036
Chain 1: 2200 -19880.697 0.083 0.036
Chain 1: 2300 -19498.439 0.039 0.020
Chain 1: 2400 -19270.688 0.039 0.020
Chain 1: 2500 -19072.649 0.025 0.016
Chain 1: 2600 -18703.246 0.023 0.016
Chain 1: 2700 -18660.306 0.018 0.012
Chain 1: 2800 -18377.308 0.019 0.015
Chain 1: 2900 -18658.318 0.019 0.015
Chain 1: 3000 -18644.537 0.012 0.012
Chain 1: 3100 -18729.505 0.011 0.012
Chain 1: 3200 -18420.427 0.012 0.015
Chain 1: 3300 -18624.942 0.011 0.012
Chain 1: 3400 -18100.274 0.012 0.015
Chain 1: 3500 -18711.523 0.015 0.015
Chain 1: 3600 -18018.954 0.017 0.015
Chain 1: 3700 -18405.185 0.018 0.017
Chain 1: 3800 -17366.100 0.023 0.021
Chain 1: 3900 -17362.263 0.021 0.021
Chain 1: 4000 -17479.575 0.022 0.021
Chain 1: 4100 -17393.428 0.022 0.021
Chain 1: 4200 -17209.903 0.021 0.021
Chain 1: 4300 -17348.139 0.021 0.021
Chain 1: 4400 -17305.167 0.018 0.011
Chain 1: 4500 -17207.728 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001401 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.01 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48691.715 1.000 1.000
Chain 1: 200 -15121.793 1.610 2.220
Chain 1: 300 -17872.282 1.125 1.000
Chain 1: 400 -22045.821 0.891 1.000
Chain 1: 500 -12131.016 0.876 0.817
Chain 1: 600 -23180.301 0.810 0.817
Chain 1: 700 -14828.051 0.774 0.563
Chain 1: 800 -13369.279 0.691 0.563
Chain 1: 900 -18676.936 0.646 0.477
Chain 1: 1000 -13461.163 0.620 0.477
Chain 1: 1100 -11884.379 0.533 0.387 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -11103.735 0.318 0.284
Chain 1: 1300 -12457.911 0.314 0.284
Chain 1: 1400 -10921.346 0.309 0.284
Chain 1: 1500 -16121.600 0.260 0.284
Chain 1: 1600 -12064.587 0.246 0.284
Chain 1: 1700 -9447.373 0.217 0.277
Chain 1: 1800 -10890.024 0.219 0.277
Chain 1: 1900 -10320.471 0.196 0.141
Chain 1: 2000 -11937.772 0.171 0.135
Chain 1: 2100 -12987.016 0.166 0.135
Chain 1: 2200 -9737.248 0.192 0.141
Chain 1: 2300 -14182.610 0.213 0.277
Chain 1: 2400 -10697.318 0.231 0.313
Chain 1: 2500 -9865.735 0.207 0.277
Chain 1: 2600 -9188.709 0.181 0.135
Chain 1: 2700 -8858.270 0.157 0.132
Chain 1: 2800 -10581.407 0.160 0.135
Chain 1: 2900 -8653.497 0.177 0.163
Chain 1: 3000 -10568.969 0.182 0.181
Chain 1: 3100 -8749.533 0.194 0.208
Chain 1: 3200 -9033.054 0.164 0.181
Chain 1: 3300 -8878.290 0.134 0.163
Chain 1: 3400 -10050.217 0.114 0.117
Chain 1: 3500 -12295.687 0.123 0.163
Chain 1: 3600 -9096.235 0.151 0.181
Chain 1: 3700 -9302.111 0.150 0.181
Chain 1: 3800 -12971.784 0.162 0.183
Chain 1: 3900 -9247.485 0.180 0.183
Chain 1: 4000 -9749.073 0.167 0.183
Chain 1: 4100 -9056.227 0.154 0.117
Chain 1: 4200 -12751.977 0.179 0.183
Chain 1: 4300 -12763.091 0.178 0.183
Chain 1: 4400 -12356.696 0.169 0.183
Chain 1: 4500 -8920.397 0.190 0.283
Chain 1: 4600 -8423.591 0.160 0.077
Chain 1: 4700 -10265.208 0.176 0.179
Chain 1: 4800 -8651.078 0.166 0.179
Chain 1: 4900 -12654.915 0.158 0.179
Chain 1: 5000 -15007.453 0.168 0.179
Chain 1: 5100 -13265.154 0.174 0.179
Chain 1: 5200 -8702.257 0.197 0.179
Chain 1: 5300 -12874.237 0.230 0.187
Chain 1: 5400 -14503.212 0.238 0.187
Chain 1: 5500 -9302.603 0.255 0.187
Chain 1: 5600 -8922.340 0.253 0.187
Chain 1: 5700 -10853.309 0.253 0.187
Chain 1: 5800 -11383.627 0.239 0.178
Chain 1: 5900 -9138.099 0.232 0.178
Chain 1: 6000 -8721.594 0.221 0.178
Chain 1: 6100 -9230.817 0.214 0.178
Chain 1: 6200 -10765.524 0.175 0.143
Chain 1: 6300 -8589.587 0.168 0.143
Chain 1: 6400 -8948.127 0.161 0.143
Chain 1: 6500 -8603.122 0.109 0.055
Chain 1: 6600 -10280.578 0.121 0.143
Chain 1: 6700 -9222.629 0.115 0.115
Chain 1: 6800 -11891.442 0.133 0.143
Chain 1: 6900 -12221.886 0.111 0.115
Chain 1: 7000 -10109.614 0.127 0.143
Chain 1: 7100 -8096.841 0.146 0.163
Chain 1: 7200 -11867.610 0.164 0.209
Chain 1: 7300 -7986.766 0.187 0.209
Chain 1: 7400 -8304.663 0.187 0.209
Chain 1: 7500 -8288.697 0.183 0.209
Chain 1: 7600 -8082.089 0.169 0.209
Chain 1: 7700 -8118.975 0.158 0.209
Chain 1: 7800 -8208.578 0.137 0.038
Chain 1: 7900 -8668.619 0.140 0.053
Chain 1: 8000 -8687.483 0.119 0.038
Chain 1: 8100 -8420.767 0.097 0.032
Chain 1: 8200 -10690.250 0.087 0.032
Chain 1: 8300 -8001.087 0.072 0.032
Chain 1: 8400 -10783.088 0.094 0.032
Chain 1: 8500 -9634.721 0.105 0.053
Chain 1: 8600 -8550.202 0.115 0.119
Chain 1: 8700 -7981.789 0.122 0.119
Chain 1: 8800 -8213.728 0.124 0.119
Chain 1: 8900 -9631.031 0.133 0.127
Chain 1: 9000 -8143.337 0.151 0.147
Chain 1: 9100 -8188.053 0.149 0.147
Chain 1: 9200 -8070.904 0.129 0.127
Chain 1: 9300 -8488.434 0.100 0.119
Chain 1: 9400 -8286.242 0.077 0.071
Chain 1: 9500 -8409.655 0.066 0.049
Chain 1: 9600 -8097.197 0.058 0.039
Chain 1: 9700 -9353.768 0.064 0.039
Chain 1: 9800 -8205.136 0.075 0.049
Chain 1: 9900 -8986.497 0.069 0.049
Chain 1: 10000 -8188.765 0.061 0.049
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001891 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 18.91 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58488.585 1.000 1.000
Chain 1: 200 -17269.934 1.693 2.387
Chain 1: 300 -8625.677 1.463 1.002
Chain 1: 400 -8191.002 1.110 1.002
Chain 1: 500 -8036.026 0.892 1.000
Chain 1: 600 -7899.430 0.746 1.000
Chain 1: 700 -7689.154 0.644 0.053
Chain 1: 800 -7949.095 0.567 0.053
Chain 1: 900 -7888.930 0.505 0.033
Chain 1: 1000 -7957.874 0.455 0.033
Chain 1: 1100 -7597.917 0.360 0.033
Chain 1: 1200 -7590.532 0.122 0.027
Chain 1: 1300 -7588.212 0.021 0.019
Chain 1: 1400 -7846.703 0.019 0.019
Chain 1: 1500 -7582.773 0.021 0.027
Chain 1: 1600 -7493.217 0.020 0.027
Chain 1: 1700 -7468.888 0.018 0.012
Chain 1: 1800 -7506.271 0.015 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003288 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.88 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85922.987 1.000 1.000
Chain 1: 200 -13128.167 3.272 5.545
Chain 1: 300 -9613.436 2.304 1.000
Chain 1: 400 -10389.203 1.746 1.000
Chain 1: 500 -8521.268 1.441 0.366
Chain 1: 600 -8195.906 1.207 0.366
Chain 1: 700 -8419.530 1.039 0.219
Chain 1: 800 -8447.558 0.909 0.219
Chain 1: 900 -8500.356 0.809 0.075
Chain 1: 1000 -8349.614 0.730 0.075
Chain 1: 1100 -8519.226 0.632 0.040 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8266.781 0.080 0.031
Chain 1: 1300 -8206.110 0.045 0.027
Chain 1: 1400 -8184.978 0.037 0.020
Chain 1: 1500 -8242.142 0.016 0.018
Chain 1: 1600 -8244.838 0.012 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004185 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 41.85 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8409188.620 1.000 1.000
Chain 1: 200 -1582050.582 2.658 4.315
Chain 1: 300 -888825.112 2.032 1.000
Chain 1: 400 -456375.618 1.761 1.000
Chain 1: 500 -356772.547 1.464 0.948
Chain 1: 600 -231886.298 1.310 0.948
Chain 1: 700 -118466.058 1.260 0.948
Chain 1: 800 -85796.049 1.150 0.948
Chain 1: 900 -66196.593 1.055 0.780
Chain 1: 1000 -51037.068 0.979 0.780
Chain 1: 1100 -38563.798 0.912 0.539 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37739.316 0.482 0.381
Chain 1: 1300 -25749.797 0.451 0.381
Chain 1: 1400 -25471.342 0.357 0.323
Chain 1: 1500 -22073.867 0.345 0.323
Chain 1: 1600 -21294.336 0.294 0.297
Chain 1: 1700 -20174.819 0.204 0.296
Chain 1: 1800 -20120.291 0.166 0.154
Chain 1: 1900 -20445.815 0.138 0.055
Chain 1: 2000 -18961.989 0.116 0.055
Chain 1: 2100 -19199.888 0.085 0.037
Chain 1: 2200 -19425.468 0.084 0.037
Chain 1: 2300 -19043.647 0.040 0.020
Chain 1: 2400 -18816.057 0.040 0.020
Chain 1: 2500 -18618.016 0.026 0.016
Chain 1: 2600 -18248.974 0.024 0.016
Chain 1: 2700 -18206.229 0.019 0.012
Chain 1: 2800 -17923.409 0.020 0.016
Chain 1: 2900 -18204.215 0.020 0.015
Chain 1: 3000 -18190.448 0.012 0.012
Chain 1: 3100 -18275.341 0.011 0.012
Chain 1: 3200 -17966.532 0.012 0.015
Chain 1: 3300 -18170.868 0.011 0.012
Chain 1: 3400 -17646.690 0.013 0.015
Chain 1: 3500 -18257.196 0.015 0.016
Chain 1: 3600 -17565.609 0.017 0.016
Chain 1: 3700 -17951.105 0.019 0.017
Chain 1: 3800 -16913.523 0.023 0.021
Chain 1: 3900 -16909.735 0.022 0.021
Chain 1: 4000 -17027.025 0.023 0.021
Chain 1: 4100 -16940.946 0.023 0.021
Chain 1: 4200 -16757.788 0.022 0.021
Chain 1: 4300 -16895.763 0.022 0.021
Chain 1: 4400 -16853.053 0.019 0.011
Chain 1: 4500 -16755.677 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001383 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.83 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12833.353 1.000 1.000
Chain 1: 200 -9718.545 0.660 1.000
Chain 1: 300 -8274.044 0.498 0.321
Chain 1: 400 -8470.354 0.380 0.321
Chain 1: 500 -8386.537 0.306 0.175
Chain 1: 600 -8219.085 0.258 0.175
Chain 1: 700 -8105.772 0.223 0.023
Chain 1: 800 -8103.848 0.195 0.023
Chain 1: 900 -8168.489 0.175 0.020
Chain 1: 1000 -8126.046 0.158 0.020
Chain 1: 1100 -8217.047 0.059 0.014
Chain 1: 1200 -8141.636 0.028 0.011
Chain 1: 1300 -8072.639 0.011 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001804 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 18.04 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58252.800 1.000 1.000
Chain 1: 200 -18056.845 1.613 2.226
Chain 1: 300 -8882.978 1.420 1.033
Chain 1: 400 -8118.008 1.088 1.033
Chain 1: 500 -8337.247 0.876 1.000
Chain 1: 600 -8710.721 0.737 1.000
Chain 1: 700 -8610.128 0.633 0.094
Chain 1: 800 -8260.933 0.560 0.094
Chain 1: 900 -8047.724 0.500 0.043
Chain 1: 1000 -7908.591 0.452 0.043
Chain 1: 1100 -7740.249 0.354 0.042
Chain 1: 1200 -7936.276 0.134 0.026
Chain 1: 1300 -7842.454 0.032 0.026
Chain 1: 1400 -7829.955 0.023 0.025
Chain 1: 1500 -7617.616 0.023 0.025
Chain 1: 1600 -7730.431 0.020 0.022
Chain 1: 1700 -7571.976 0.021 0.022
Chain 1: 1800 -7671.731 0.018 0.021
Chain 1: 1900 -7630.194 0.016 0.018
Chain 1: 2000 -7705.435 0.015 0.015
Chain 1: 2100 -7604.618 0.014 0.013
Chain 1: 2200 -7817.560 0.015 0.013
Chain 1: 2300 -7579.834 0.017 0.015
Chain 1: 2400 -7653.494 0.017 0.015
Chain 1: 2500 -7597.334 0.015 0.013
Chain 1: 2600 -7566.254 0.014 0.013
Chain 1: 2700 -7496.515 0.013 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003304 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.04 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86702.865 1.000 1.000
Chain 1: 200 -13912.062 3.116 5.232
Chain 1: 300 -10178.711 2.200 1.000
Chain 1: 400 -11580.293 1.680 1.000
Chain 1: 500 -9163.969 1.397 0.367
Chain 1: 600 -9232.929 1.165 0.367
Chain 1: 700 -8621.437 1.009 0.264
Chain 1: 800 -8959.311 0.887 0.264
Chain 1: 900 -8970.942 0.789 0.121
Chain 1: 1000 -8736.342 0.713 0.121
Chain 1: 1100 -9016.155 0.616 0.071 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8554.704 0.098 0.054
Chain 1: 1300 -8763.778 0.064 0.038
Chain 1: 1400 -8804.607 0.052 0.031
Chain 1: 1500 -8703.254 0.027 0.027
Chain 1: 1600 -8811.006 0.027 0.027
Chain 1: 1700 -8878.468 0.021 0.024
Chain 1: 1800 -8442.120 0.022 0.024
Chain 1: 1900 -8547.200 0.024 0.024
Chain 1: 2000 -8523.401 0.021 0.012
Chain 1: 2100 -8665.908 0.020 0.012
Chain 1: 2200 -8453.759 0.017 0.012
Chain 1: 2300 -8614.137 0.016 0.012
Chain 1: 2400 -8449.592 0.018 0.016
Chain 1: 2500 -8521.095 0.017 0.016
Chain 1: 2600 -8433.161 0.017 0.016
Chain 1: 2700 -8467.369 0.017 0.016
Chain 1: 2800 -8427.158 0.012 0.012
Chain 1: 2900 -8520.764 0.012 0.011
Chain 1: 3000 -8354.636 0.014 0.016
Chain 1: 3100 -8509.846 0.014 0.018
Chain 1: 3200 -8381.682 0.013 0.015
Chain 1: 3300 -8389.574 0.011 0.011
Chain 1: 3400 -8550.888 0.011 0.011
Chain 1: 3500 -8561.167 0.010 0.011
Chain 1: 3600 -8338.104 0.012 0.015
Chain 1: 3700 -8484.547 0.013 0.017
Chain 1: 3800 -8344.535 0.015 0.017
Chain 1: 3900 -8278.939 0.014 0.017
Chain 1: 4000 -8355.327 0.013 0.017
Chain 1: 4100 -8350.285 0.011 0.015
Chain 1: 4200 -8334.262 0.010 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003438 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.38 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8389769.237 1.000 1.000
Chain 1: 200 -1582347.058 2.651 4.302
Chain 1: 300 -891977.159 2.025 1.000
Chain 1: 400 -458946.196 1.755 1.000
Chain 1: 500 -359378.284 1.459 0.944
Chain 1: 600 -234105.419 1.305 0.944
Chain 1: 700 -119978.004 1.255 0.944
Chain 1: 800 -87095.657 1.145 0.944
Chain 1: 900 -67381.231 1.050 0.774
Chain 1: 1000 -52134.944 0.975 0.774
Chain 1: 1100 -39569.893 0.906 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38742.867 0.478 0.378
Chain 1: 1300 -26652.724 0.446 0.378
Chain 1: 1400 -26368.872 0.353 0.318
Chain 1: 1500 -22943.326 0.340 0.318
Chain 1: 1600 -22156.469 0.290 0.293
Chain 1: 1700 -21024.451 0.200 0.292
Chain 1: 1800 -20967.345 0.163 0.149
Chain 1: 1900 -21293.888 0.135 0.054
Chain 1: 2000 -19801.027 0.114 0.054
Chain 1: 2100 -20039.753 0.083 0.036
Chain 1: 2200 -20266.953 0.082 0.036
Chain 1: 2300 -19883.337 0.039 0.019
Chain 1: 2400 -19655.221 0.039 0.019
Chain 1: 2500 -19457.343 0.025 0.015
Chain 1: 2600 -19087.090 0.023 0.015
Chain 1: 2700 -19043.820 0.018 0.012
Chain 1: 2800 -18760.613 0.019 0.015
Chain 1: 2900 -19042.048 0.019 0.015
Chain 1: 3000 -19028.225 0.012 0.012
Chain 1: 3100 -19113.306 0.011 0.012
Chain 1: 3200 -18803.676 0.011 0.015
Chain 1: 3300 -19008.583 0.011 0.012
Chain 1: 3400 -18483.061 0.012 0.015
Chain 1: 3500 -19095.707 0.014 0.015
Chain 1: 3600 -18401.325 0.016 0.015
Chain 1: 3700 -18789.005 0.018 0.016
Chain 1: 3800 -17747.117 0.022 0.021
Chain 1: 3900 -17743.196 0.021 0.021
Chain 1: 4000 -17860.508 0.022 0.021
Chain 1: 4100 -17774.264 0.022 0.021
Chain 1: 4200 -17590.065 0.021 0.021
Chain 1: 4300 -17728.759 0.021 0.021
Chain 1: 4400 -17685.316 0.018 0.010
Chain 1: 4500 -17587.770 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001336 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.36 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49627.718 1.000 1.000
Chain 1: 200 -22965.729 1.080 1.161
Chain 1: 300 -21320.036 0.746 1.000
Chain 1: 400 -13340.050 0.709 1.000
Chain 1: 500 -16432.051 0.605 0.598
Chain 1: 600 -12091.197 0.564 0.598
Chain 1: 700 -17627.954 0.528 0.359
Chain 1: 800 -12987.257 0.507 0.359
Chain 1: 900 -14405.807 0.461 0.357
Chain 1: 1000 -15744.814 0.424 0.357
Chain 1: 1100 -10939.414 0.368 0.357
Chain 1: 1200 -14291.506 0.275 0.314
Chain 1: 1300 -12661.920 0.280 0.314
Chain 1: 1400 -11659.619 0.229 0.235
Chain 1: 1500 -10742.766 0.219 0.235
Chain 1: 1600 -12701.371 0.198 0.154
Chain 1: 1700 -10182.171 0.192 0.154
Chain 1: 1800 -12523.052 0.175 0.154
Chain 1: 1900 -10664.690 0.182 0.174
Chain 1: 2000 -12336.656 0.187 0.174
Chain 1: 2100 -18262.973 0.176 0.174
Chain 1: 2200 -10794.597 0.221 0.174
Chain 1: 2300 -10113.988 0.215 0.174
Chain 1: 2400 -10223.212 0.208 0.174
Chain 1: 2500 -11524.695 0.211 0.174
Chain 1: 2600 -13203.836 0.208 0.174
Chain 1: 2700 -9845.628 0.217 0.174
Chain 1: 2800 -10560.300 0.205 0.136
Chain 1: 2900 -10280.167 0.191 0.127
Chain 1: 3000 -15843.221 0.212 0.127
Chain 1: 3100 -9495.993 0.247 0.127
Chain 1: 3200 -9453.442 0.178 0.113
Chain 1: 3300 -10521.089 0.181 0.113
Chain 1: 3400 -10027.628 0.185 0.113
Chain 1: 3500 -9758.494 0.177 0.101
Chain 1: 3600 -11565.938 0.179 0.101
Chain 1: 3700 -10014.970 0.161 0.101
Chain 1: 3800 -13526.689 0.180 0.155
Chain 1: 3900 -9536.867 0.219 0.156
Chain 1: 4000 -10402.468 0.192 0.155
Chain 1: 4100 -10080.335 0.129 0.101
Chain 1: 4200 -12823.051 0.150 0.155
Chain 1: 4300 -10340.952 0.163 0.156
Chain 1: 4400 -14309.261 0.186 0.214
Chain 1: 4500 -10981.298 0.214 0.240
Chain 1: 4600 -14743.993 0.224 0.255
Chain 1: 4700 -9340.575 0.266 0.260
Chain 1: 4800 -12951.016 0.268 0.277
Chain 1: 4900 -9532.144 0.262 0.277
Chain 1: 5000 -12914.200 0.280 0.277
Chain 1: 5100 -9896.074 0.307 0.279
Chain 1: 5200 -15352.607 0.321 0.303
Chain 1: 5300 -14197.539 0.306 0.303
Chain 1: 5400 -9066.647 0.334 0.305
Chain 1: 5500 -9813.522 0.312 0.305
Chain 1: 5600 -10395.782 0.292 0.305
Chain 1: 5700 -10722.913 0.237 0.279
Chain 1: 5800 -9506.678 0.222 0.262
Chain 1: 5900 -12759.073 0.212 0.255
Chain 1: 6000 -9732.580 0.216 0.255
Chain 1: 6100 -9142.183 0.192 0.128
Chain 1: 6200 -8743.864 0.161 0.081
Chain 1: 6300 -11205.815 0.175 0.128
Chain 1: 6400 -11544.084 0.122 0.076
Chain 1: 6500 -9870.842 0.131 0.128
Chain 1: 6600 -9152.660 0.133 0.128
Chain 1: 6700 -8984.362 0.132 0.128
Chain 1: 6800 -9845.986 0.128 0.088
Chain 1: 6900 -9151.418 0.110 0.078
Chain 1: 7000 -9172.905 0.079 0.076
Chain 1: 7100 -8945.525 0.075 0.076
Chain 1: 7200 -9047.918 0.072 0.076
Chain 1: 7300 -8852.033 0.052 0.029
Chain 1: 7400 -8855.498 0.049 0.025
Chain 1: 7500 -12063.816 0.059 0.025
Chain 1: 7600 -9926.429 0.073 0.025
Chain 1: 7700 -8827.135 0.083 0.076
Chain 1: 7800 -9331.479 0.080 0.054
Chain 1: 7900 -9001.108 0.076 0.037
Chain 1: 8000 -9017.879 0.076 0.037
Chain 1: 8100 -8823.267 0.075 0.037
Chain 1: 8200 -9055.724 0.077 0.037
Chain 1: 8300 -9468.502 0.079 0.044
Chain 1: 8400 -10045.175 0.085 0.054
Chain 1: 8500 -9877.507 0.060 0.044
Chain 1: 8600 -9937.348 0.039 0.037
Chain 1: 8700 -9139.352 0.035 0.037
Chain 1: 8800 -8747.243 0.034 0.037
Chain 1: 8900 -9457.906 0.038 0.044
Chain 1: 9000 -8936.971 0.044 0.045
Chain 1: 9100 -8898.061 0.042 0.045
Chain 1: 9200 -11853.852 0.064 0.057
Chain 1: 9300 -10224.095 0.076 0.058
Chain 1: 9400 -9113.695 0.082 0.075
Chain 1: 9500 -9904.191 0.089 0.080
Chain 1: 9600 -10514.698 0.094 0.080
Chain 1: 9700 -11477.659 0.093 0.080
Chain 1: 9800 -9260.351 0.113 0.084
Chain 1: 9900 -11156.898 0.122 0.122
Chain 1: 10000 -8779.398 0.144 0.159
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003144 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.44 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -64012.744 1.000 1.000
Chain 1: 200 -18895.584 1.694 2.388
Chain 1: 300 -9095.655 1.488 1.077
Chain 1: 400 -8192.190 1.144 1.077
Chain 1: 500 -8915.347 0.931 1.000
Chain 1: 600 -9434.416 0.785 1.000
Chain 1: 700 -7866.001 0.702 0.199
Chain 1: 800 -8573.982 0.624 0.199
Chain 1: 900 -8016.197 0.563 0.110
Chain 1: 1000 -7640.339 0.511 0.110
Chain 1: 1100 -7662.923 0.412 0.083
Chain 1: 1200 -7914.794 0.176 0.081
Chain 1: 1300 -7890.124 0.069 0.070
Chain 1: 1400 -7778.164 0.059 0.055
Chain 1: 1500 -7567.879 0.054 0.049
Chain 1: 1600 -7739.538 0.050 0.032
Chain 1: 1700 -7454.284 0.034 0.032
Chain 1: 1800 -7506.944 0.027 0.028
Chain 1: 1900 -7562.291 0.020 0.022
Chain 1: 2000 -7600.615 0.016 0.014
Chain 1: 2100 -7541.454 0.016 0.014
Chain 1: 2200 -7757.840 0.016 0.014
Chain 1: 2300 -7549.510 0.019 0.022
Chain 1: 2400 -7578.780 0.017 0.022
Chain 1: 2500 -7579.341 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002905 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.05 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86237.457 1.000 1.000
Chain 1: 200 -14219.758 3.032 5.065
Chain 1: 300 -10527.668 2.138 1.000
Chain 1: 400 -11713.352 1.629 1.000
Chain 1: 500 -9536.905 1.349 0.351
Chain 1: 600 -9150.956 1.131 0.351
Chain 1: 700 -8914.515 0.973 0.228
Chain 1: 800 -9209.797 0.856 0.228
Chain 1: 900 -9325.057 0.762 0.101
Chain 1: 1000 -9216.279 0.687 0.101
Chain 1: 1100 -9284.068 0.588 0.042 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8832.600 0.086 0.042
Chain 1: 1300 -9182.646 0.055 0.038
Chain 1: 1400 -9177.169 0.045 0.032
Chain 1: 1500 -9040.011 0.024 0.027
Chain 1: 1600 -9159.133 0.021 0.015
Chain 1: 1700 -9225.156 0.019 0.013
Chain 1: 1800 -8791.378 0.021 0.013
Chain 1: 1900 -8894.700 0.021 0.013
Chain 1: 2000 -8870.212 0.020 0.013
Chain 1: 2100 -8835.315 0.019 0.013
Chain 1: 2200 -8812.859 0.014 0.012
Chain 1: 2300 -8948.566 0.012 0.012
Chain 1: 2400 -8795.597 0.014 0.013
Chain 1: 2500 -8864.852 0.013 0.012
Chain 1: 2600 -8783.044 0.013 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003445 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.45 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8372150.760 1.000 1.000
Chain 1: 200 -1577392.285 2.654 4.308
Chain 1: 300 -890074.590 2.027 1.000
Chain 1: 400 -457984.853 1.756 1.000
Chain 1: 500 -358945.862 1.460 0.943
Chain 1: 600 -234076.900 1.305 0.943
Chain 1: 700 -120181.683 1.254 0.943
Chain 1: 800 -87370.633 1.144 0.943
Chain 1: 900 -67673.493 1.050 0.772
Chain 1: 1000 -52444.992 0.974 0.772
Chain 1: 1100 -39890.307 0.905 0.533 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39067.739 0.477 0.376
Chain 1: 1300 -26977.353 0.444 0.376
Chain 1: 1400 -26694.340 0.351 0.315
Chain 1: 1500 -23269.230 0.338 0.315
Chain 1: 1600 -22483.246 0.288 0.291
Chain 1: 1700 -21350.269 0.199 0.290
Chain 1: 1800 -21293.437 0.161 0.147
Chain 1: 1900 -21619.959 0.134 0.053
Chain 1: 2000 -20127.273 0.112 0.053
Chain 1: 2100 -20365.699 0.082 0.035
Chain 1: 2200 -20593.051 0.081 0.035
Chain 1: 2300 -20209.400 0.038 0.019
Chain 1: 2400 -19981.326 0.038 0.019
Chain 1: 2500 -19783.618 0.024 0.015
Chain 1: 2600 -19413.105 0.023 0.015
Chain 1: 2700 -19369.925 0.018 0.012
Chain 1: 2800 -19086.736 0.019 0.015
Chain 1: 2900 -19368.215 0.019 0.015
Chain 1: 3000 -19354.309 0.011 0.012
Chain 1: 3100 -19439.374 0.011 0.011
Chain 1: 3200 -19129.722 0.011 0.015
Chain 1: 3300 -19334.745 0.010 0.011
Chain 1: 3400 -18809.211 0.012 0.015
Chain 1: 3500 -19421.865 0.014 0.015
Chain 1: 3600 -18727.570 0.016 0.015
Chain 1: 3700 -19115.123 0.018 0.016
Chain 1: 3800 -18073.406 0.022 0.020
Chain 1: 3900 -18069.588 0.021 0.020
Chain 1: 4000 -18186.818 0.021 0.020
Chain 1: 4100 -18100.527 0.021 0.020
Chain 1: 4200 -17916.517 0.021 0.020
Chain 1: 4300 -18055.072 0.020 0.020
Chain 1: 4400 -18011.634 0.018 0.010
Chain 1: 4500 -17914.187 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001381 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.81 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12187.608 1.000 1.000
Chain 1: 200 -8941.786 0.681 1.000
Chain 1: 300 -7850.089 0.501 0.363
Chain 1: 400 -7997.765 0.380 0.363
Chain 1: 500 -7950.640 0.305 0.139
Chain 1: 600 -7802.631 0.258 0.139
Chain 1: 700 -7722.917 0.222 0.019
Chain 1: 800 -7731.512 0.195 0.019
Chain 1: 900 -7646.891 0.174 0.018
Chain 1: 1000 -7827.507 0.159 0.019
Chain 1: 1100 -7857.450 0.059 0.018
Chain 1: 1200 -7753.263 0.025 0.013
Chain 1: 1300 -7699.544 0.011 0.011
Chain 1: 1400 -7715.560 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001435 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.35 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56901.533 1.000 1.000
Chain 1: 200 -17201.740 1.654 2.308
Chain 1: 300 -8664.647 1.431 1.000
Chain 1: 400 -8357.532 1.082 1.000
Chain 1: 500 -8105.755 0.872 0.985
Chain 1: 600 -8268.231 0.730 0.985
Chain 1: 700 -8160.924 0.628 0.037
Chain 1: 800 -7959.883 0.552 0.037
Chain 1: 900 -7959.887 0.491 0.031
Chain 1: 1000 -7847.988 0.443 0.031
Chain 1: 1100 -7893.794 0.344 0.025
Chain 1: 1200 -7738.004 0.115 0.020
Chain 1: 1300 -7697.159 0.017 0.020
Chain 1: 1400 -7685.489 0.014 0.014
Chain 1: 1500 -7644.248 0.011 0.013
Chain 1: 1600 -7611.458 0.010 0.006 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004086 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 40.86 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85613.444 1.000 1.000
Chain 1: 200 -13278.327 3.224 5.448
Chain 1: 300 -9680.481 2.273 1.000
Chain 1: 400 -10670.554 1.728 1.000
Chain 1: 500 -8574.069 1.431 0.372
Chain 1: 600 -8161.821 1.201 0.372
Chain 1: 700 -8268.502 1.031 0.245
Chain 1: 800 -8704.054 0.909 0.245
Chain 1: 900 -8557.377 0.810 0.093
Chain 1: 1000 -8246.917 0.732 0.093
Chain 1: 1100 -8527.545 0.636 0.051 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8191.687 0.095 0.050
Chain 1: 1300 -8392.301 0.060 0.041
Chain 1: 1400 -8398.334 0.051 0.038
Chain 1: 1500 -8257.718 0.028 0.033
Chain 1: 1600 -8370.039 0.025 0.024
Chain 1: 1700 -8455.628 0.024 0.024
Chain 1: 1800 -8049.472 0.024 0.024
Chain 1: 1900 -8145.881 0.024 0.024
Chain 1: 2000 -8118.045 0.020 0.017
Chain 1: 2100 -8238.685 0.019 0.015
Chain 1: 2200 -8051.726 0.017 0.015
Chain 1: 2300 -8185.910 0.016 0.015
Chain 1: 2400 -8194.081 0.016 0.015
Chain 1: 2500 -8158.935 0.015 0.013
Chain 1: 2600 -8156.954 0.014 0.012
Chain 1: 2700 -8071.341 0.014 0.012
Chain 1: 2800 -8036.322 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003157 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.57 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8413721.727 1.000 1.000
Chain 1: 200 -1587392.141 2.650 4.300
Chain 1: 300 -890709.731 2.028 1.000
Chain 1: 400 -457070.968 1.758 1.000
Chain 1: 500 -357149.649 1.462 0.949
Chain 1: 600 -232121.012 1.308 0.949
Chain 1: 700 -118665.655 1.258 0.949
Chain 1: 800 -85967.920 1.148 0.949
Chain 1: 900 -66374.483 1.053 0.782
Chain 1: 1000 -51227.728 0.978 0.782
Chain 1: 1100 -38756.060 0.910 0.539 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37937.337 0.482 0.380
Chain 1: 1300 -25944.431 0.450 0.380
Chain 1: 1400 -25668.182 0.356 0.322
Chain 1: 1500 -22268.523 0.344 0.322
Chain 1: 1600 -21489.019 0.293 0.296
Chain 1: 1700 -20368.837 0.203 0.295
Chain 1: 1800 -20314.487 0.165 0.153
Chain 1: 1900 -20640.512 0.137 0.055
Chain 1: 2000 -19155.072 0.116 0.055
Chain 1: 2100 -19393.262 0.085 0.036
Chain 1: 2200 -19619.133 0.084 0.036
Chain 1: 2300 -19236.900 0.039 0.020
Chain 1: 2400 -19009.085 0.040 0.020
Chain 1: 2500 -18810.933 0.025 0.016
Chain 1: 2600 -18441.472 0.024 0.016
Chain 1: 2700 -18398.589 0.018 0.012
Chain 1: 2800 -18115.411 0.020 0.016
Chain 1: 2900 -18396.524 0.020 0.015
Chain 1: 3000 -18382.770 0.012 0.012
Chain 1: 3100 -18467.715 0.011 0.012
Chain 1: 3200 -18158.587 0.012 0.015
Chain 1: 3300 -18363.177 0.011 0.012
Chain 1: 3400 -17838.299 0.013 0.015
Chain 1: 3500 -18449.832 0.015 0.016
Chain 1: 3600 -17756.952 0.017 0.016
Chain 1: 3700 -18143.352 0.019 0.017
Chain 1: 3800 -17103.752 0.023 0.021
Chain 1: 3900 -17099.888 0.022 0.021
Chain 1: 4000 -17217.213 0.022 0.021
Chain 1: 4100 -17130.989 0.022 0.021
Chain 1: 4200 -16947.403 0.022 0.021
Chain 1: 4300 -17085.700 0.021 0.021
Chain 1: 4400 -17042.634 0.019 0.011
Chain 1: 4500 -16945.176 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001375 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.75 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49833.498 1.000 1.000
Chain 1: 200 -24012.357 1.038 1.075
Chain 1: 300 -14649.706 0.905 1.000
Chain 1: 400 -17826.857 0.723 1.000
Chain 1: 500 -16372.288 0.596 0.639
Chain 1: 600 -24798.371 0.554 0.639
Chain 1: 700 -16430.010 0.547 0.509
Chain 1: 800 -15635.876 0.485 0.509
Chain 1: 900 -25030.450 0.473 0.375
Chain 1: 1000 -13813.606 0.507 0.509
Chain 1: 1100 -14570.022 0.412 0.375
Chain 1: 1200 -21051.582 0.335 0.340
Chain 1: 1300 -12591.233 0.339 0.340
Chain 1: 1400 -12080.033 0.325 0.340
Chain 1: 1500 -10991.295 0.326 0.340
Chain 1: 1600 -18261.933 0.332 0.375
Chain 1: 1700 -16966.838 0.289 0.308
Chain 1: 1800 -11687.803 0.329 0.375
Chain 1: 1900 -11020.314 0.297 0.308
Chain 1: 2000 -15191.950 0.243 0.275
Chain 1: 2100 -18990.909 0.258 0.275
Chain 1: 2200 -13381.225 0.269 0.275
Chain 1: 2300 -10204.932 0.233 0.275
Chain 1: 2400 -15481.592 0.263 0.311
Chain 1: 2500 -10859.266 0.296 0.341
Chain 1: 2600 -9955.068 0.265 0.311
Chain 1: 2700 -12929.819 0.280 0.311
Chain 1: 2800 -12046.145 0.243 0.275
Chain 1: 2900 -10414.699 0.252 0.275
Chain 1: 3000 -12971.690 0.245 0.230
Chain 1: 3100 -10225.882 0.251 0.269
Chain 1: 3200 -10192.278 0.210 0.230
Chain 1: 3300 -16905.804 0.218 0.230
Chain 1: 3400 -9549.587 0.261 0.230
Chain 1: 3500 -10177.583 0.225 0.197
Chain 1: 3600 -9365.817 0.224 0.197
Chain 1: 3700 -15762.302 0.242 0.197
Chain 1: 3800 -15931.445 0.236 0.197
Chain 1: 3900 -9752.003 0.283 0.269
Chain 1: 4000 -9623.517 0.265 0.269
Chain 1: 4100 -10413.643 0.246 0.087
Chain 1: 4200 -10100.766 0.249 0.087
Chain 1: 4300 -15013.371 0.242 0.087
Chain 1: 4400 -10393.714 0.209 0.087
Chain 1: 4500 -11562.911 0.213 0.101
Chain 1: 4600 -9387.802 0.227 0.232
Chain 1: 4700 -13822.609 0.219 0.232
Chain 1: 4800 -9455.487 0.264 0.321
Chain 1: 4900 -14280.362 0.235 0.321
Chain 1: 5000 -15224.565 0.239 0.321
Chain 1: 5100 -10978.482 0.270 0.327
Chain 1: 5200 -9638.162 0.281 0.327
Chain 1: 5300 -9344.751 0.252 0.321
Chain 1: 5400 -9145.044 0.209 0.232
Chain 1: 5500 -8993.289 0.201 0.232
Chain 1: 5600 -15546.675 0.220 0.321
Chain 1: 5700 -14813.990 0.193 0.139
Chain 1: 5800 -9732.330 0.199 0.139
Chain 1: 5900 -9601.747 0.166 0.062
Chain 1: 6000 -9371.061 0.163 0.049
Chain 1: 6100 -9344.011 0.124 0.031
Chain 1: 6200 -8940.011 0.115 0.031
Chain 1: 6300 -14156.186 0.149 0.045
Chain 1: 6400 -15331.196 0.154 0.049
Chain 1: 6500 -9344.753 0.217 0.077
Chain 1: 6600 -9110.635 0.177 0.049
Chain 1: 6700 -8774.952 0.176 0.045
Chain 1: 6800 -11393.405 0.147 0.045
Chain 1: 6900 -11397.854 0.145 0.045
Chain 1: 7000 -9805.048 0.159 0.077
Chain 1: 7100 -13173.693 0.184 0.162
Chain 1: 7200 -9150.844 0.224 0.230
Chain 1: 7300 -10485.374 0.200 0.162
Chain 1: 7400 -8668.831 0.213 0.210
Chain 1: 7500 -12203.429 0.178 0.210
Chain 1: 7600 -9446.172 0.204 0.230
Chain 1: 7700 -9318.000 0.202 0.230
Chain 1: 7800 -8870.798 0.184 0.210
Chain 1: 7900 -8737.467 0.186 0.210
Chain 1: 8000 -14146.779 0.208 0.256
Chain 1: 8100 -11826.516 0.202 0.210
Chain 1: 8200 -12286.545 0.161 0.196
Chain 1: 8300 -8961.656 0.186 0.210
Chain 1: 8400 -13200.947 0.197 0.290
Chain 1: 8500 -9013.123 0.214 0.292
Chain 1: 8600 -13792.836 0.220 0.321
Chain 1: 8700 -12525.555 0.229 0.321
Chain 1: 8800 -8920.763 0.264 0.347
Chain 1: 8900 -9220.321 0.266 0.347
Chain 1: 9000 -13368.023 0.258 0.321
Chain 1: 9100 -8741.888 0.292 0.347
Chain 1: 9200 -9219.288 0.293 0.347
Chain 1: 9300 -10214.094 0.266 0.321
Chain 1: 9400 -8806.894 0.250 0.310
Chain 1: 9500 -10279.122 0.218 0.160
Chain 1: 9600 -9498.023 0.191 0.143
Chain 1: 9700 -8575.890 0.192 0.143
Chain 1: 9800 -8664.636 0.152 0.108
Chain 1: 9900 -8983.887 0.153 0.108
Chain 1: 10000 -8748.022 0.124 0.097
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001565 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.65 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -62810.111 1.000 1.000
Chain 1: 200 -18385.743 1.708 2.416
Chain 1: 300 -9128.961 1.477 1.014
Chain 1: 400 -8420.808 1.129 1.014
Chain 1: 500 -8083.620 0.911 1.000
Chain 1: 600 -8206.952 0.762 1.000
Chain 1: 700 -8680.982 0.661 0.084
Chain 1: 800 -8294.564 0.584 0.084
Chain 1: 900 -8123.960 0.521 0.055
Chain 1: 1000 -7632.920 0.476 0.064
Chain 1: 1100 -7991.072 0.380 0.055
Chain 1: 1200 -7771.319 0.141 0.047
Chain 1: 1300 -7650.286 0.042 0.045
Chain 1: 1400 -7531.209 0.035 0.042
Chain 1: 1500 -7472.687 0.031 0.028
Chain 1: 1600 -7705.499 0.033 0.030
Chain 1: 1700 -7670.516 0.028 0.028
Chain 1: 1800 -7728.649 0.024 0.021
Chain 1: 1900 -7571.511 0.024 0.021
Chain 1: 2000 -7580.292 0.018 0.016
Chain 1: 2100 -7537.815 0.014 0.016
Chain 1: 2200 -7736.165 0.013 0.016
Chain 1: 2300 -7575.929 0.014 0.016
Chain 1: 2400 -7485.913 0.014 0.012
Chain 1: 2500 -7561.554 0.014 0.012
Chain 1: 2600 -7470.450 0.012 0.012
Chain 1: 2700 -7362.946 0.013 0.012
Chain 1: 2800 -7653.276 0.016 0.015
Chain 1: 2900 -7304.162 0.019 0.015
Chain 1: 3000 -7483.057 0.021 0.021
Chain 1: 3100 -7452.615 0.021 0.021
Chain 1: 3200 -7691.053 0.021 0.021
Chain 1: 3300 -7346.829 0.024 0.024
Chain 1: 3400 -7590.662 0.026 0.031
Chain 1: 3500 -7374.611 0.028 0.031
Chain 1: 3600 -7438.796 0.028 0.031
Chain 1: 3700 -7396.640 0.027 0.031
Chain 1: 3800 -7366.529 0.023 0.029
Chain 1: 3900 -7339.214 0.019 0.024
Chain 1: 4000 -7335.237 0.017 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.006106 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 61.06 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86540.182 1.000 1.000
Chain 1: 200 -14139.626 3.060 5.120
Chain 1: 300 -10461.185 2.157 1.000
Chain 1: 400 -11506.181 1.641 1.000
Chain 1: 500 -9441.987 1.356 0.352
Chain 1: 600 -9209.417 1.134 0.352
Chain 1: 700 -8844.789 0.978 0.219
Chain 1: 800 -9219.074 0.861 0.219
Chain 1: 900 -9342.867 0.767 0.091
Chain 1: 1000 -8979.131 0.694 0.091
Chain 1: 1100 -9075.658 0.595 0.041 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8876.376 0.085 0.041
Chain 1: 1300 -9089.168 0.053 0.041
Chain 1: 1400 -9094.456 0.044 0.025
Chain 1: 1500 -8990.709 0.023 0.023
Chain 1: 1600 -9096.016 0.022 0.022
Chain 1: 1700 -9171.168 0.018 0.013
Chain 1: 1800 -8741.318 0.019 0.013
Chain 1: 1900 -8845.027 0.019 0.012
Chain 1: 2000 -8820.288 0.015 0.012
Chain 1: 2100 -8952.111 0.016 0.012
Chain 1: 2200 -8748.188 0.016 0.012
Chain 1: 2300 -8843.246 0.014 0.012
Chain 1: 2400 -8908.577 0.015 0.012
Chain 1: 2500 -8853.901 0.015 0.012
Chain 1: 2600 -8857.792 0.013 0.011
Chain 1: 2700 -8773.173 0.014 0.011
Chain 1: 2800 -8730.331 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003965 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 39.65 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8375450.733 1.000 1.000
Chain 1: 200 -1577238.271 2.655 4.310
Chain 1: 300 -889634.039 2.028 1.000
Chain 1: 400 -457787.383 1.757 1.000
Chain 1: 500 -358653.102 1.461 0.943
Chain 1: 600 -233779.418 1.306 0.943
Chain 1: 700 -119979.615 1.255 0.943
Chain 1: 800 -87182.006 1.145 0.943
Chain 1: 900 -67510.080 1.050 0.773
Chain 1: 1000 -52291.787 0.974 0.773
Chain 1: 1100 -39752.399 0.906 0.534 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38930.959 0.477 0.376
Chain 1: 1300 -26859.711 0.445 0.376
Chain 1: 1400 -26579.052 0.351 0.315
Chain 1: 1500 -23158.278 0.339 0.315
Chain 1: 1600 -22373.296 0.289 0.291
Chain 1: 1700 -21242.880 0.199 0.291
Chain 1: 1800 -21186.661 0.162 0.148
Chain 1: 1900 -21513.010 0.134 0.053
Chain 1: 2000 -20021.674 0.112 0.053
Chain 1: 2100 -20260.252 0.082 0.035
Chain 1: 2200 -20487.151 0.081 0.035
Chain 1: 2300 -20103.873 0.038 0.019
Chain 1: 2400 -19875.798 0.038 0.019
Chain 1: 2500 -19677.987 0.024 0.015
Chain 1: 2600 -19307.760 0.023 0.015
Chain 1: 2700 -19264.673 0.018 0.012
Chain 1: 2800 -18981.448 0.019 0.015
Chain 1: 2900 -19262.841 0.019 0.015
Chain 1: 3000 -19249.016 0.012 0.012
Chain 1: 3100 -19334.048 0.011 0.011
Chain 1: 3200 -19024.540 0.011 0.015
Chain 1: 3300 -19229.426 0.010 0.011
Chain 1: 3400 -18704.075 0.012 0.015
Chain 1: 3500 -19316.448 0.014 0.015
Chain 1: 3600 -18622.458 0.016 0.015
Chain 1: 3700 -19009.744 0.018 0.016
Chain 1: 3800 -17968.519 0.022 0.020
Chain 1: 3900 -17964.664 0.021 0.020
Chain 1: 4000 -18081.925 0.021 0.020
Chain 1: 4100 -17995.632 0.021 0.020
Chain 1: 4200 -17811.719 0.021 0.020
Chain 1: 4300 -17950.230 0.020 0.020
Chain 1: 4400 -17906.872 0.018 0.010
Chain 1: 4500 -17809.393 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001381 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.81 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49015.484 1.000 1.000
Chain 1: 200 -20747.407 1.181 1.362
Chain 1: 300 -13350.821 0.972 1.000
Chain 1: 400 -19716.523 0.810 1.000
Chain 1: 500 -13537.284 0.739 0.554
Chain 1: 600 -29610.264 0.706 0.554
Chain 1: 700 -15942.359 0.728 0.554
Chain 1: 800 -13836.083 0.656 0.554
Chain 1: 900 -10652.512 0.616 0.543
Chain 1: 1000 -16851.228 0.591 0.543
Chain 1: 1100 -11043.073 0.544 0.526 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -24493.891 0.463 0.526 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1300 -10960.203 0.531 0.526 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1400 -22500.634 0.550 0.526 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1500 -12649.945 0.582 0.543 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1600 -29090.759 0.584 0.549 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1700 -10169.527 0.685 0.549 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1800 -9777.376 0.673 0.549 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1900 -11479.704 0.658 0.549 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2000 -11524.176 0.622 0.549 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2100 -19677.031 0.611 0.549 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2200 -9496.399 0.663 0.565 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2300 -9872.385 0.543 0.513 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2400 -9994.836 0.493 0.414
Chain 1: 2500 -12796.161 0.437 0.219
Chain 1: 2600 -9851.506 0.411 0.219
Chain 1: 2700 -9088.986 0.233 0.148
Chain 1: 2800 -15522.545 0.271 0.219
Chain 1: 2900 -9596.734 0.317 0.299
Chain 1: 3000 -9674.490 0.318 0.299
Chain 1: 3100 -8843.204 0.286 0.219
Chain 1: 3200 -9707.934 0.188 0.094
Chain 1: 3300 -9130.687 0.190 0.094
Chain 1: 3400 -13894.351 0.223 0.219
Chain 1: 3500 -8972.655 0.256 0.299
Chain 1: 3600 -9231.218 0.229 0.094
Chain 1: 3700 -12150.703 0.245 0.240
Chain 1: 3800 -8773.396 0.242 0.240
Chain 1: 3900 -9092.822 0.183 0.094
Chain 1: 4000 -14655.710 0.221 0.240
Chain 1: 4100 -10171.125 0.255 0.343
Chain 1: 4200 -8987.566 0.260 0.343
Chain 1: 4300 -10646.501 0.269 0.343
Chain 1: 4400 -8752.876 0.256 0.240
Chain 1: 4500 -11468.535 0.225 0.237
Chain 1: 4600 -10297.310 0.234 0.237
Chain 1: 4700 -10433.472 0.211 0.216
Chain 1: 4800 -9016.398 0.188 0.157
Chain 1: 4900 -9403.966 0.189 0.157
Chain 1: 5000 -14572.484 0.186 0.157
Chain 1: 5100 -10349.626 0.183 0.157
Chain 1: 5200 -11654.308 0.181 0.157
Chain 1: 5300 -12447.238 0.172 0.157
Chain 1: 5400 -8560.277 0.195 0.157
Chain 1: 5500 -8823.941 0.175 0.114
Chain 1: 5600 -13763.394 0.199 0.157
Chain 1: 5700 -12655.858 0.207 0.157
Chain 1: 5800 -9327.312 0.227 0.355
Chain 1: 5900 -9160.605 0.224 0.355
Chain 1: 6000 -8481.276 0.197 0.112
Chain 1: 6100 -10015.412 0.171 0.112
Chain 1: 6200 -9035.023 0.171 0.109
Chain 1: 6300 -9017.330 0.165 0.109
Chain 1: 6400 -9751.665 0.127 0.088
Chain 1: 6500 -9493.993 0.127 0.088
Chain 1: 6600 -8901.026 0.098 0.080
Chain 1: 6700 -9464.524 0.095 0.075
Chain 1: 6800 -8729.050 0.067 0.075
Chain 1: 6900 -11419.278 0.089 0.080
Chain 1: 7000 -12967.955 0.093 0.084
Chain 1: 7100 -8450.310 0.131 0.084
Chain 1: 7200 -11178.534 0.145 0.084
Chain 1: 7300 -8711.506 0.173 0.119
Chain 1: 7400 -8347.906 0.170 0.119
Chain 1: 7500 -10437.393 0.187 0.200
Chain 1: 7600 -9715.376 0.188 0.200
Chain 1: 7700 -8327.422 0.199 0.200
Chain 1: 7800 -11948.649 0.220 0.236
Chain 1: 7900 -8325.980 0.240 0.244
Chain 1: 8000 -8430.818 0.230 0.244
Chain 1: 8100 -8763.745 0.180 0.200
Chain 1: 8200 -9492.337 0.163 0.167
Chain 1: 8300 -11203.555 0.150 0.153
Chain 1: 8400 -8300.951 0.181 0.167
Chain 1: 8500 -9167.419 0.170 0.153
Chain 1: 8600 -9360.908 0.165 0.153
Chain 1: 8700 -8286.350 0.161 0.130
Chain 1: 8800 -8385.019 0.132 0.095
Chain 1: 8900 -11015.642 0.113 0.095
Chain 1: 9000 -11576.703 0.116 0.095
Chain 1: 9100 -8255.695 0.153 0.130
Chain 1: 9200 -10862.655 0.169 0.153
Chain 1: 9300 -11215.682 0.157 0.130
Chain 1: 9400 -8747.329 0.150 0.130
Chain 1: 9500 -8788.860 0.141 0.130
Chain 1: 9600 -9852.310 0.150 0.130
Chain 1: 9700 -9936.492 0.138 0.108
Chain 1: 9800 -8447.307 0.154 0.176
Chain 1: 9900 -8979.741 0.136 0.108
Chain 1: 10000 -8341.318 0.139 0.108
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001578 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.78 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58093.006 1.000 1.000
Chain 1: 200 -17750.081 1.636 2.273
Chain 1: 300 -8693.353 1.438 1.042
Chain 1: 400 -8204.065 1.094 1.042
Chain 1: 500 -8396.293 0.879 1.000
Chain 1: 600 -8895.580 0.742 1.000
Chain 1: 700 -8102.206 0.650 0.098
Chain 1: 800 -8171.140 0.570 0.098
Chain 1: 900 -7955.718 0.510 0.060
Chain 1: 1000 -7845.470 0.460 0.060
Chain 1: 1100 -7744.721 0.361 0.056
Chain 1: 1200 -7685.964 0.135 0.027
Chain 1: 1300 -7610.757 0.032 0.023
Chain 1: 1400 -7637.388 0.026 0.014
Chain 1: 1500 -7580.495 0.025 0.013
Chain 1: 1600 -7768.518 0.021 0.013
Chain 1: 1700 -7529.074 0.015 0.013
Chain 1: 1800 -7625.336 0.015 0.013
Chain 1: 1900 -7649.500 0.013 0.013
Chain 1: 2000 -7630.519 0.012 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003265 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.65 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86819.380 1.000 1.000
Chain 1: 200 -13536.193 3.207 5.414
Chain 1: 300 -9965.785 2.257 1.000
Chain 1: 400 -10914.398 1.715 1.000
Chain 1: 500 -8883.323 1.418 0.358
Chain 1: 600 -8513.751 1.189 0.358
Chain 1: 700 -8654.918 1.021 0.229
Chain 1: 800 -9317.401 0.902 0.229
Chain 1: 900 -8825.603 0.808 0.087
Chain 1: 1000 -8546.380 0.731 0.087
Chain 1: 1100 -8716.557 0.633 0.071 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8513.010 0.094 0.056
Chain 1: 1300 -8707.340 0.060 0.043
Chain 1: 1400 -8697.492 0.051 0.033
Chain 1: 1500 -8600.404 0.030 0.024
Chain 1: 1600 -8696.907 0.027 0.022
Chain 1: 1700 -8786.404 0.026 0.022
Chain 1: 1800 -8397.342 0.023 0.022
Chain 1: 1900 -8499.725 0.019 0.020
Chain 1: 2000 -8469.785 0.016 0.012
Chain 1: 2100 -8600.499 0.016 0.012
Chain 1: 2200 -8386.695 0.016 0.012
Chain 1: 2300 -8528.873 0.015 0.012
Chain 1: 2400 -8541.733 0.015 0.012
Chain 1: 2500 -8509.603 0.015 0.012
Chain 1: 2600 -8509.873 0.013 0.012
Chain 1: 2700 -8417.801 0.014 0.012
Chain 1: 2800 -8393.343 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003271 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.71 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8410042.501 1.000 1.000
Chain 1: 200 -1589592.072 2.645 4.291
Chain 1: 300 -892679.857 2.024 1.000
Chain 1: 400 -458388.808 1.755 1.000
Chain 1: 500 -358278.927 1.460 0.947
Chain 1: 600 -232964.217 1.306 0.947
Chain 1: 700 -119191.315 1.256 0.947
Chain 1: 800 -86384.810 1.146 0.947
Chain 1: 900 -66736.939 1.052 0.781
Chain 1: 1000 -51539.061 0.976 0.781
Chain 1: 1100 -39025.840 0.908 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38199.432 0.481 0.380
Chain 1: 1300 -26179.835 0.449 0.380
Chain 1: 1400 -25898.540 0.355 0.321
Chain 1: 1500 -22491.895 0.343 0.321
Chain 1: 1600 -21709.200 0.292 0.295
Chain 1: 1700 -20586.710 0.202 0.294
Chain 1: 1800 -20531.343 0.165 0.151
Chain 1: 1900 -20856.965 0.137 0.055
Chain 1: 2000 -19370.817 0.115 0.055
Chain 1: 2100 -19609.152 0.084 0.036
Chain 1: 2200 -19834.865 0.083 0.036
Chain 1: 2300 -19452.837 0.039 0.020
Chain 1: 2400 -19225.131 0.039 0.020
Chain 1: 2500 -19026.966 0.025 0.016
Chain 1: 2600 -18657.888 0.023 0.016
Chain 1: 2700 -18615.062 0.018 0.012
Chain 1: 2800 -18332.034 0.020 0.015
Chain 1: 2900 -18613.039 0.019 0.015
Chain 1: 3000 -18599.330 0.012 0.012
Chain 1: 3100 -18684.218 0.011 0.012
Chain 1: 3200 -18375.293 0.012 0.015
Chain 1: 3300 -18579.705 0.011 0.012
Chain 1: 3400 -18055.235 0.013 0.015
Chain 1: 3500 -18666.135 0.015 0.015
Chain 1: 3600 -17974.117 0.017 0.015
Chain 1: 3700 -18359.908 0.018 0.017
Chain 1: 3800 -17321.578 0.023 0.021
Chain 1: 3900 -17317.739 0.021 0.021
Chain 1: 4000 -17435.077 0.022 0.021
Chain 1: 4100 -17348.902 0.022 0.021
Chain 1: 4200 -17165.579 0.021 0.021
Chain 1: 4300 -17303.695 0.021 0.021
Chain 1: 4400 -17260.882 0.019 0.011
Chain 1: 4500 -17163.446 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001538 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.38 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49326.460 1.000 1.000
Chain 1: 200 -14828.186 1.663 2.327
Chain 1: 300 -13666.936 1.137 1.000
Chain 1: 400 -18585.097 0.919 1.000
Chain 1: 500 -12895.344 0.823 0.441
Chain 1: 600 -17272.973 0.728 0.441
Chain 1: 700 -16356.962 0.632 0.265
Chain 1: 800 -10741.034 0.619 0.441
Chain 1: 900 -19649.465 0.600 0.441
Chain 1: 1000 -15677.883 0.566 0.441
Chain 1: 1100 -10134.845 0.520 0.441 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -11836.257 0.302 0.265
Chain 1: 1300 -11236.432 0.299 0.265
Chain 1: 1400 -10185.913 0.283 0.253
Chain 1: 1500 -11692.689 0.252 0.253
Chain 1: 1600 -23044.105 0.275 0.253
Chain 1: 1700 -15401.516 0.319 0.453
Chain 1: 1800 -12560.054 0.290 0.253
Chain 1: 1900 -10466.425 0.264 0.226
Chain 1: 2000 -15310.843 0.271 0.226
Chain 1: 2100 -10307.725 0.265 0.226
Chain 1: 2200 -10206.558 0.251 0.226
Chain 1: 2300 -9557.791 0.253 0.226
Chain 1: 2400 -9333.683 0.245 0.226
Chain 1: 2500 -16624.368 0.276 0.316
Chain 1: 2600 -9521.541 0.301 0.316
Chain 1: 2700 -9551.987 0.252 0.226
Chain 1: 2800 -12407.963 0.252 0.230
Chain 1: 2900 -14794.025 0.248 0.230
Chain 1: 3000 -9071.436 0.280 0.230
Chain 1: 3100 -9542.850 0.236 0.161
Chain 1: 3200 -9056.263 0.241 0.161
Chain 1: 3300 -9149.292 0.235 0.161
Chain 1: 3400 -8840.213 0.236 0.161
Chain 1: 3500 -8886.522 0.192 0.054
Chain 1: 3600 -9013.300 0.119 0.049
Chain 1: 3700 -9529.742 0.124 0.054
Chain 1: 3800 -8649.371 0.112 0.054
Chain 1: 3900 -10052.399 0.109 0.054
Chain 1: 4000 -8888.298 0.059 0.054
Chain 1: 4100 -14605.099 0.094 0.054
Chain 1: 4200 -12634.562 0.104 0.102
Chain 1: 4300 -11501.977 0.113 0.102
Chain 1: 4400 -9123.478 0.135 0.131
Chain 1: 4500 -10095.033 0.144 0.131
Chain 1: 4600 -9839.931 0.146 0.131
Chain 1: 4700 -13793.393 0.169 0.140
Chain 1: 4800 -8690.591 0.217 0.156
Chain 1: 4900 -8875.329 0.205 0.156
Chain 1: 5000 -9884.736 0.203 0.156
Chain 1: 5100 -9125.641 0.172 0.102
Chain 1: 5200 -8827.805 0.159 0.098
Chain 1: 5300 -11703.420 0.174 0.102
Chain 1: 5400 -9416.366 0.172 0.102
Chain 1: 5500 -8565.885 0.173 0.102
Chain 1: 5600 -8536.964 0.170 0.102
Chain 1: 5700 -8459.732 0.143 0.099
Chain 1: 5800 -8911.189 0.089 0.083
Chain 1: 5900 -11300.904 0.108 0.099
Chain 1: 6000 -8326.116 0.134 0.099
Chain 1: 6100 -8757.101 0.130 0.099
Chain 1: 6200 -11188.959 0.149 0.211
Chain 1: 6300 -8686.644 0.153 0.211
Chain 1: 6400 -13165.534 0.163 0.211
Chain 1: 6500 -8874.003 0.201 0.217
Chain 1: 6600 -9456.844 0.207 0.217
Chain 1: 6700 -8449.172 0.218 0.217
Chain 1: 6800 -8348.412 0.214 0.217
Chain 1: 6900 -8190.365 0.195 0.217
Chain 1: 7000 -14225.925 0.201 0.217
Chain 1: 7100 -8202.444 0.270 0.288
Chain 1: 7200 -9495.967 0.262 0.288
Chain 1: 7300 -8285.324 0.248 0.146
Chain 1: 7400 -8606.950 0.217 0.136
Chain 1: 7500 -8272.048 0.173 0.119
Chain 1: 7600 -9293.622 0.178 0.119
Chain 1: 7700 -8134.079 0.180 0.136
Chain 1: 7800 -9909.310 0.197 0.143
Chain 1: 7900 -8776.141 0.208 0.143
Chain 1: 8000 -8411.599 0.170 0.136
Chain 1: 8100 -8381.151 0.097 0.129
Chain 1: 8200 -8268.255 0.085 0.110
Chain 1: 8300 -8273.608 0.070 0.043
Chain 1: 8400 -9329.161 0.078 0.110
Chain 1: 8500 -8681.486 0.081 0.110
Chain 1: 8600 -8013.189 0.078 0.083
Chain 1: 8700 -8348.271 0.068 0.075
Chain 1: 8800 -10586.602 0.071 0.075
Chain 1: 8900 -8411.387 0.084 0.075
Chain 1: 9000 -8492.219 0.081 0.075
Chain 1: 9100 -8880.993 0.085 0.075
Chain 1: 9200 -8216.487 0.092 0.081
Chain 1: 9300 -8102.487 0.093 0.081
Chain 1: 9400 -8061.629 0.082 0.075
Chain 1: 9500 -8239.133 0.077 0.044
Chain 1: 9600 -8360.141 0.070 0.040
Chain 1: 9700 -11269.755 0.092 0.044
Chain 1: 9800 -9139.703 0.094 0.044
Chain 1: 9900 -9204.170 0.069 0.022
Chain 1: 10000 -8035.880 0.082 0.044
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001564 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.64 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -63313.030 1.000 1.000
Chain 1: 200 -18378.963 1.722 2.445
Chain 1: 300 -8828.006 1.509 1.082
Chain 1: 400 -7926.140 1.160 1.082
Chain 1: 500 -8656.722 0.945 1.000
Chain 1: 600 -9706.700 0.806 1.000
Chain 1: 700 -7941.142 0.722 0.222
Chain 1: 800 -8307.338 0.637 0.222
Chain 1: 900 -7907.920 0.572 0.114
Chain 1: 1000 -7793.967 0.516 0.114
Chain 1: 1100 -7820.015 0.417 0.108
Chain 1: 1200 -7506.574 0.176 0.084
Chain 1: 1300 -7534.494 0.069 0.051
Chain 1: 1400 -7837.173 0.061 0.044
Chain 1: 1500 -7594.422 0.056 0.042
Chain 1: 1600 -7794.489 0.048 0.039
Chain 1: 1700 -7514.333 0.029 0.037
Chain 1: 1800 -7557.824 0.025 0.032
Chain 1: 1900 -7574.743 0.020 0.026
Chain 1: 2000 -7552.385 0.019 0.026
Chain 1: 2100 -7491.631 0.020 0.026
Chain 1: 2200 -7750.007 0.019 0.026
Chain 1: 2300 -7534.564 0.021 0.029
Chain 1: 2400 -7546.908 0.018 0.026
Chain 1: 2500 -7577.775 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003318 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.18 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86283.790 1.000 1.000
Chain 1: 200 -13620.454 3.167 5.335
Chain 1: 300 -9888.125 2.237 1.000
Chain 1: 400 -11249.960 1.708 1.000
Chain 1: 500 -8850.716 1.421 0.377
Chain 1: 600 -8590.541 1.189 0.377
Chain 1: 700 -8476.669 1.021 0.271
Chain 1: 800 -8139.660 0.899 0.271
Chain 1: 900 -8239.146 0.800 0.121
Chain 1: 1000 -8542.050 0.724 0.121
Chain 1: 1100 -8644.869 0.625 0.041 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8221.845 0.097 0.041
Chain 1: 1300 -8550.374 0.063 0.038
Chain 1: 1400 -8306.357 0.053 0.035
Chain 1: 1500 -8387.186 0.027 0.030
Chain 1: 1600 -8491.792 0.026 0.029
Chain 1: 1700 -8552.344 0.025 0.029
Chain 1: 1800 -8107.646 0.026 0.029
Chain 1: 1900 -8215.149 0.026 0.029
Chain 1: 2000 -8201.778 0.023 0.013
Chain 1: 2100 -8318.418 0.023 0.014
Chain 1: 2200 -8112.865 0.021 0.014
Chain 1: 2300 -8208.383 0.018 0.013
Chain 1: 2400 -8275.274 0.016 0.012
Chain 1: 2500 -8223.650 0.015 0.012
Chain 1: 2600 -8237.737 0.014 0.012
Chain 1: 2700 -8145.185 0.015 0.012
Chain 1: 2800 -8092.550 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003236 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.36 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8405508.110 1.000 1.000
Chain 1: 200 -1584176.697 2.653 4.306
Chain 1: 300 -890278.847 2.028 1.000
Chain 1: 400 -456794.506 1.759 1.000
Chain 1: 500 -357002.721 1.463 0.949
Chain 1: 600 -232174.023 1.309 0.949
Chain 1: 700 -118938.137 1.258 0.949
Chain 1: 800 -86240.209 1.148 0.949
Chain 1: 900 -66687.741 1.053 0.779
Chain 1: 1000 -51569.252 0.977 0.779
Chain 1: 1100 -39111.731 0.909 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38305.352 0.480 0.379
Chain 1: 1300 -26315.247 0.448 0.379
Chain 1: 1400 -26042.857 0.354 0.319
Chain 1: 1500 -22642.127 0.341 0.319
Chain 1: 1600 -21863.428 0.291 0.293
Chain 1: 1700 -20742.806 0.201 0.293
Chain 1: 1800 -20688.961 0.163 0.150
Chain 1: 1900 -21015.645 0.136 0.054
Chain 1: 2000 -19528.661 0.114 0.054
Chain 1: 2100 -19767.076 0.083 0.036
Chain 1: 2200 -19993.285 0.082 0.036
Chain 1: 2300 -19610.586 0.039 0.020
Chain 1: 2400 -19382.556 0.039 0.020
Chain 1: 2500 -19184.201 0.025 0.016
Chain 1: 2600 -18814.162 0.023 0.016
Chain 1: 2700 -18771.180 0.018 0.012
Chain 1: 2800 -18487.587 0.019 0.015
Chain 1: 2900 -18769.018 0.019 0.015
Chain 1: 3000 -18755.304 0.012 0.012
Chain 1: 3100 -18840.313 0.011 0.012
Chain 1: 3200 -18530.730 0.012 0.015
Chain 1: 3300 -18735.711 0.011 0.012
Chain 1: 3400 -18209.964 0.012 0.015
Chain 1: 3500 -18822.704 0.015 0.015
Chain 1: 3600 -18128.299 0.017 0.015
Chain 1: 3700 -18515.799 0.018 0.017
Chain 1: 3800 -17473.735 0.023 0.021
Chain 1: 3900 -17469.811 0.021 0.021
Chain 1: 4000 -17587.157 0.022 0.021
Chain 1: 4100 -17500.743 0.022 0.021
Chain 1: 4200 -17316.719 0.021 0.021
Chain 1: 4300 -17455.382 0.021 0.021
Chain 1: 4400 -17411.883 0.018 0.011
Chain 1: 4500 -17314.347 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001299 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.99 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -11922.649 1.000 1.000
Chain 1: 200 -8876.852 0.672 1.000
Chain 1: 300 -7856.745 0.491 0.343
Chain 1: 400 -7950.762 0.371 0.343
Chain 1: 500 -7783.058 0.301 0.130
Chain 1: 600 -7667.194 0.254 0.130
Chain 1: 700 -7756.094 0.219 0.022
Chain 1: 800 -7627.040 0.194 0.022
Chain 1: 900 -7678.897 0.173 0.017
Chain 1: 1000 -7665.513 0.156 0.017
Chain 1: 1100 -7725.210 0.057 0.015
Chain 1: 1200 -7623.392 0.024 0.013
Chain 1: 1300 -7603.117 0.011 0.012
Chain 1: 1400 -7609.231 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001439 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.39 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56416.657 1.000 1.000
Chain 1: 200 -16895.680 1.670 2.339
Chain 1: 300 -8449.230 1.446 1.000
Chain 1: 400 -8591.599 1.089 1.000
Chain 1: 500 -8198.822 0.881 1.000
Chain 1: 600 -8788.391 0.745 1.000
Chain 1: 700 -7640.451 0.660 0.150
Chain 1: 800 -7884.489 0.581 0.150
Chain 1: 900 -7583.213 0.521 0.067
Chain 1: 1000 -7613.060 0.470 0.067
Chain 1: 1100 -7591.763 0.370 0.048
Chain 1: 1200 -7695.953 0.137 0.040
Chain 1: 1300 -7536.141 0.039 0.031
Chain 1: 1400 -7754.993 0.041 0.031
Chain 1: 1500 -7525.717 0.039 0.030
Chain 1: 1600 -7488.211 0.033 0.028
Chain 1: 1700 -7413.239 0.019 0.021
Chain 1: 1800 -7446.792 0.016 0.014
Chain 1: 1900 -7442.212 0.012 0.010
Chain 1: 2000 -7500.786 0.012 0.010
Chain 1: 2100 -7528.819 0.013 0.010
Chain 1: 2200 -7567.794 0.012 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003428 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.28 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86094.605 1.000 1.000
Chain 1: 200 -12969.049 3.319 5.638
Chain 1: 300 -9445.685 2.337 1.000
Chain 1: 400 -10260.913 1.773 1.000
Chain 1: 500 -8294.977 1.466 0.373
Chain 1: 600 -8048.999 1.226 0.373
Chain 1: 700 -8347.159 1.056 0.237
Chain 1: 800 -8522.770 0.927 0.237
Chain 1: 900 -8366.087 0.826 0.079
Chain 1: 1000 -8059.023 0.747 0.079
Chain 1: 1100 -8346.417 0.651 0.038 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8113.259 0.090 0.036
Chain 1: 1300 -8219.273 0.054 0.034
Chain 1: 1400 -8215.748 0.046 0.031
Chain 1: 1500 -8120.639 0.023 0.029
Chain 1: 1600 -8204.923 0.021 0.021
Chain 1: 1700 -8310.027 0.019 0.019
Chain 1: 1800 -7925.062 0.022 0.019
Chain 1: 1900 -8023.795 0.021 0.013
Chain 1: 2000 -7993.799 0.018 0.013
Chain 1: 2100 -8137.844 0.016 0.013
Chain 1: 2200 -7916.273 0.016 0.013
Chain 1: 2300 -8055.788 0.016 0.013
Chain 1: 2400 -7940.202 0.018 0.015
Chain 1: 2500 -8000.680 0.017 0.015
Chain 1: 2600 -8014.158 0.016 0.015
Chain 1: 2700 -7935.191 0.016 0.015
Chain 1: 2800 -7919.549 0.011 0.012
Chain 1: 2900 -7916.510 0.010 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003513 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.13 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8410200.220 1.000 1.000
Chain 1: 200 -1586599.506 2.650 4.301
Chain 1: 300 -891287.569 2.027 1.000
Chain 1: 400 -457493.297 1.757 1.000
Chain 1: 500 -357426.934 1.462 0.948
Chain 1: 600 -232389.504 1.308 0.948
Chain 1: 700 -118617.266 1.258 0.948
Chain 1: 800 -85807.983 1.149 0.948
Chain 1: 900 -66159.084 1.054 0.780
Chain 1: 1000 -50956.642 0.978 0.780
Chain 1: 1100 -38440.732 0.911 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37612.776 0.483 0.382
Chain 1: 1300 -25593.493 0.452 0.382
Chain 1: 1400 -25311.619 0.358 0.326
Chain 1: 1500 -21905.225 0.346 0.326
Chain 1: 1600 -21122.347 0.296 0.298
Chain 1: 1700 -19999.982 0.205 0.297
Chain 1: 1800 -19944.478 0.168 0.156
Chain 1: 1900 -20269.845 0.139 0.056
Chain 1: 2000 -18784.329 0.117 0.056
Chain 1: 2100 -19022.579 0.086 0.037
Chain 1: 2200 -19248.118 0.085 0.037
Chain 1: 2300 -18866.322 0.040 0.020
Chain 1: 2400 -18638.725 0.040 0.020
Chain 1: 2500 -18440.597 0.026 0.016
Chain 1: 2600 -18071.763 0.024 0.016
Chain 1: 2700 -18029.013 0.019 0.013
Chain 1: 2800 -17746.151 0.020 0.016
Chain 1: 2900 -18027.016 0.020 0.016
Chain 1: 3000 -18013.331 0.012 0.013
Chain 1: 3100 -18098.169 0.011 0.012
Chain 1: 3200 -17789.413 0.012 0.016
Chain 1: 3300 -17993.689 0.011 0.012
Chain 1: 3400 -17469.572 0.013 0.016
Chain 1: 3500 -18079.945 0.015 0.016
Chain 1: 3600 -17388.611 0.017 0.016
Chain 1: 3700 -17773.922 0.019 0.017
Chain 1: 3800 -16736.634 0.024 0.022
Chain 1: 3900 -16732.829 0.022 0.022
Chain 1: 4000 -16850.160 0.023 0.022
Chain 1: 4100 -16764.041 0.023 0.022
Chain 1: 4200 -16580.953 0.022 0.022
Chain 1: 4300 -16718.894 0.022 0.022
Chain 1: 4400 -16676.268 0.019 0.011
Chain 1: 4500 -16578.880 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001349 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.49 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12227.653 1.000 1.000
Chain 1: 200 -9201.240 0.664 1.000
Chain 1: 300 -7923.048 0.497 0.329
Chain 1: 400 -8036.657 0.376 0.329
Chain 1: 500 -8006.523 0.302 0.161
Chain 1: 600 -7828.755 0.255 0.161
Chain 1: 700 -7734.605 0.220 0.023
Chain 1: 800 -7745.963 0.193 0.023
Chain 1: 900 -7708.177 0.172 0.014
Chain 1: 1000 -7774.439 0.156 0.014
Chain 1: 1100 -7812.505 0.056 0.012
Chain 1: 1200 -7772.703 0.024 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001393 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.93 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61547.691 1.000 1.000
Chain 1: 200 -17790.620 1.730 2.460
Chain 1: 300 -8832.748 1.491 1.014
Chain 1: 400 -9362.063 1.133 1.014
Chain 1: 500 -8079.586 0.938 1.000
Chain 1: 600 -8354.398 0.787 1.000
Chain 1: 700 -7751.556 0.686 0.159
Chain 1: 800 -8201.076 0.607 0.159
Chain 1: 900 -7927.377 0.543 0.078
Chain 1: 1000 -7832.578 0.490 0.078
Chain 1: 1100 -7685.211 0.392 0.057
Chain 1: 1200 -7730.804 0.147 0.055
Chain 1: 1300 -7673.509 0.046 0.035
Chain 1: 1400 -7830.979 0.042 0.033
Chain 1: 1500 -7607.169 0.029 0.029
Chain 1: 1600 -7757.381 0.028 0.020
Chain 1: 1700 -7501.711 0.024 0.020
Chain 1: 1800 -7586.835 0.019 0.019
Chain 1: 1900 -7557.400 0.016 0.019
Chain 1: 2000 -7601.512 0.016 0.019
Chain 1: 2100 -7584.383 0.014 0.011
Chain 1: 2200 -7695.159 0.015 0.014
Chain 1: 2300 -7602.582 0.015 0.014
Chain 1: 2400 -7580.707 0.014 0.012
Chain 1: 2500 -7582.971 0.011 0.011
Chain 1: 2600 -7551.885 0.009 0.006 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003226 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.26 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85903.473 1.000 1.000
Chain 1: 200 -13411.972 3.202 5.405
Chain 1: 300 -9754.642 2.260 1.000
Chain 1: 400 -10758.257 1.718 1.000
Chain 1: 500 -8723.115 1.421 0.375
Chain 1: 600 -8210.021 1.195 0.375
Chain 1: 700 -8220.746 1.024 0.233
Chain 1: 800 -8444.474 0.900 0.233
Chain 1: 900 -8511.368 0.801 0.093
Chain 1: 1000 -8497.715 0.721 0.093
Chain 1: 1100 -8636.512 0.622 0.062 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8195.029 0.087 0.054
Chain 1: 1300 -8433.256 0.052 0.028
Chain 1: 1400 -8453.733 0.043 0.026
Chain 1: 1500 -8295.472 0.022 0.019
Chain 1: 1600 -8409.576 0.017 0.016
Chain 1: 1700 -8486.274 0.018 0.016
Chain 1: 1800 -8063.261 0.020 0.016
Chain 1: 1900 -8164.213 0.021 0.016
Chain 1: 2000 -8138.543 0.021 0.016
Chain 1: 2100 -8263.980 0.021 0.015
Chain 1: 2200 -8067.677 0.018 0.015
Chain 1: 2300 -8158.958 0.016 0.014
Chain 1: 2400 -8227.797 0.017 0.014
Chain 1: 2500 -8173.978 0.016 0.012
Chain 1: 2600 -8175.242 0.014 0.011
Chain 1: 2700 -8092.052 0.014 0.011
Chain 1: 2800 -8052.044 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002783 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 27.83 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8396149.700 1.000 1.000
Chain 1: 200 -1579125.737 2.658 4.317
Chain 1: 300 -891148.353 2.030 1.000
Chain 1: 400 -458362.091 1.758 1.000
Chain 1: 500 -359323.636 1.462 0.944
Chain 1: 600 -234087.803 1.307 0.944
Chain 1: 700 -119759.021 1.257 0.944
Chain 1: 800 -86835.401 1.147 0.944
Chain 1: 900 -67050.663 1.053 0.772
Chain 1: 1000 -51749.463 0.977 0.772
Chain 1: 1100 -39139.955 0.909 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38302.712 0.480 0.379
Chain 1: 1300 -26167.775 0.449 0.379
Chain 1: 1400 -25877.520 0.355 0.322
Chain 1: 1500 -22442.076 0.343 0.322
Chain 1: 1600 -21652.019 0.293 0.296
Chain 1: 1700 -20514.794 0.203 0.295
Chain 1: 1800 -20456.283 0.166 0.153
Chain 1: 1900 -20782.529 0.138 0.055
Chain 1: 2000 -19287.932 0.116 0.055
Chain 1: 2100 -19526.417 0.085 0.036
Chain 1: 2200 -19754.034 0.084 0.036
Chain 1: 2300 -19370.196 0.040 0.020
Chain 1: 2400 -19142.156 0.040 0.020
Chain 1: 2500 -18944.633 0.025 0.016
Chain 1: 2600 -18574.267 0.024 0.016
Chain 1: 2700 -18531.002 0.018 0.012
Chain 1: 2800 -18248.098 0.020 0.016
Chain 1: 2900 -18529.461 0.020 0.015
Chain 1: 3000 -18515.482 0.012 0.012
Chain 1: 3100 -18600.569 0.011 0.012
Chain 1: 3200 -18291.034 0.012 0.015
Chain 1: 3300 -18495.896 0.011 0.012
Chain 1: 3400 -17970.694 0.013 0.015
Chain 1: 3500 -18582.942 0.015 0.016
Chain 1: 3600 -17889.121 0.017 0.016
Chain 1: 3700 -18276.429 0.019 0.017
Chain 1: 3800 -17235.484 0.023 0.021
Chain 1: 3900 -17231.681 0.022 0.021
Chain 1: 4000 -17348.894 0.022 0.021
Chain 1: 4100 -17262.718 0.022 0.021
Chain 1: 4200 -17078.773 0.022 0.021
Chain 1: 4300 -17217.236 0.021 0.021
Chain 1: 4400 -17173.955 0.019 0.011
Chain 1: 4500 -17076.515 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003853 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 38.53 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13259.571 1.000 1.000
Chain 1: 200 -10034.271 0.661 1.000
Chain 1: 300 -8560.096 0.498 0.321
Chain 1: 400 -8823.946 0.381 0.321
Chain 1: 500 -8683.704 0.308 0.172
Chain 1: 600 -8544.917 0.259 0.172
Chain 1: 700 -8412.630 0.225 0.030
Chain 1: 800 -8355.055 0.197 0.030
Chain 1: 900 -8526.602 0.178 0.020
Chain 1: 1000 -8496.552 0.160 0.020
Chain 1: 1100 -8502.475 0.060 0.016
Chain 1: 1200 -8428.294 0.029 0.016
Chain 1: 1300 -8369.134 0.013 0.016
Chain 1: 1400 -8385.147 0.010 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001388 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.88 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58925.513 1.000 1.000
Chain 1: 200 -18499.041 1.593 2.185
Chain 1: 300 -9158.269 1.402 1.020
Chain 1: 400 -8294.146 1.077 1.020
Chain 1: 500 -9036.013 0.878 1.000
Chain 1: 600 -8369.444 0.745 1.000
Chain 1: 700 -7869.429 0.648 0.104
Chain 1: 800 -8666.361 0.578 0.104
Chain 1: 900 -8079.276 0.522 0.092
Chain 1: 1000 -7779.070 0.474 0.092
Chain 1: 1100 -7691.788 0.375 0.082
Chain 1: 1200 -8093.970 0.161 0.080
Chain 1: 1300 -8032.104 0.060 0.073
Chain 1: 1400 -8114.095 0.051 0.064
Chain 1: 1500 -7641.578 0.049 0.062
Chain 1: 1600 -7823.160 0.043 0.050
Chain 1: 1700 -8029.262 0.039 0.039
Chain 1: 1800 -7825.528 0.033 0.026
Chain 1: 1900 -7612.424 0.028 0.026
Chain 1: 2000 -7859.481 0.028 0.026
Chain 1: 2100 -7739.930 0.028 0.026
Chain 1: 2200 -7957.513 0.026 0.026
Chain 1: 2300 -7714.836 0.028 0.027
Chain 1: 2400 -7898.900 0.029 0.027
Chain 1: 2500 -7723.046 0.025 0.026
Chain 1: 2600 -7661.495 0.024 0.026
Chain 1: 2700 -7600.948 0.022 0.026
Chain 1: 2800 -7775.090 0.022 0.023
Chain 1: 2900 -7507.387 0.023 0.023
Chain 1: 3000 -7655.973 0.021 0.023
Chain 1: 3100 -7649.806 0.020 0.023
Chain 1: 3200 -7818.836 0.019 0.022
Chain 1: 3300 -7537.010 0.020 0.022
Chain 1: 3400 -7771.793 0.021 0.022
Chain 1: 3500 -7572.435 0.021 0.022
Chain 1: 3600 -7640.228 0.021 0.022
Chain 1: 3700 -7593.783 0.021 0.022
Chain 1: 3800 -7564.894 0.019 0.022
Chain 1: 3900 -7536.273 0.016 0.019
Chain 1: 4000 -7528.814 0.014 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004364 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 43.64 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87484.110 1.000 1.000
Chain 1: 200 -14375.440 3.043 5.086
Chain 1: 300 -10599.256 2.147 1.000
Chain 1: 400 -12114.447 1.642 1.000
Chain 1: 500 -9404.943 1.371 0.356
Chain 1: 600 -9118.883 1.148 0.356
Chain 1: 700 -8985.454 0.986 0.288
Chain 1: 800 -9196.799 0.866 0.288
Chain 1: 900 -9332.915 0.771 0.125
Chain 1: 1000 -8991.511 0.698 0.125
Chain 1: 1100 -9310.761 0.601 0.038 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8913.117 0.097 0.038
Chain 1: 1300 -9211.227 0.065 0.034
Chain 1: 1400 -9045.823 0.054 0.032
Chain 1: 1500 -9063.889 0.025 0.031
Chain 1: 1600 -9150.490 0.023 0.023
Chain 1: 1700 -9199.102 0.022 0.023
Chain 1: 1800 -8742.592 0.025 0.032
Chain 1: 1900 -8854.022 0.025 0.032
Chain 1: 2000 -8872.123 0.021 0.018
Chain 1: 2100 -8980.003 0.019 0.013
Chain 1: 2200 -8747.724 0.017 0.013
Chain 1: 2300 -8935.817 0.016 0.013
Chain 1: 2400 -8750.139 0.016 0.013
Chain 1: 2500 -8826.449 0.017 0.013
Chain 1: 2600 -8735.372 0.017 0.013
Chain 1: 2700 -8769.153 0.017 0.013
Chain 1: 2800 -8720.216 0.012 0.012
Chain 1: 2900 -8834.813 0.012 0.012
Chain 1: 3000 -8749.520 0.013 0.012
Chain 1: 3100 -8712.346 0.012 0.010
Chain 1: 3200 -8684.545 0.010 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003575 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.75 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8418581.311 1.000 1.000
Chain 1: 200 -1586192.213 2.654 4.307
Chain 1: 300 -890249.003 2.030 1.000
Chain 1: 400 -457390.956 1.759 1.000
Chain 1: 500 -357458.432 1.463 0.946
Chain 1: 600 -232676.135 1.309 0.946
Chain 1: 700 -119542.533 1.257 0.946
Chain 1: 800 -86915.570 1.147 0.946
Chain 1: 900 -67385.532 1.051 0.782
Chain 1: 1000 -52289.823 0.975 0.782
Chain 1: 1100 -39855.309 0.906 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39053.146 0.478 0.375
Chain 1: 1300 -27077.380 0.444 0.375
Chain 1: 1400 -26808.034 0.350 0.312
Chain 1: 1500 -23411.458 0.337 0.312
Chain 1: 1600 -22634.782 0.286 0.290
Chain 1: 1700 -21515.215 0.197 0.289
Chain 1: 1800 -21461.921 0.160 0.145
Chain 1: 1900 -21788.900 0.132 0.052
Chain 1: 2000 -20302.241 0.111 0.052
Chain 1: 2100 -20540.568 0.081 0.034
Chain 1: 2200 -20766.914 0.080 0.034
Chain 1: 2300 -20384.009 0.037 0.019
Chain 1: 2400 -20155.883 0.037 0.019
Chain 1: 2500 -19957.573 0.024 0.015
Chain 1: 2600 -19587.154 0.022 0.015
Chain 1: 2700 -19544.107 0.017 0.012
Chain 1: 2800 -19260.394 0.019 0.015
Chain 1: 2900 -19541.960 0.019 0.014
Chain 1: 3000 -19528.177 0.011 0.012
Chain 1: 3100 -19613.237 0.011 0.011
Chain 1: 3200 -19303.442 0.011 0.014
Chain 1: 3300 -19508.619 0.010 0.011
Chain 1: 3400 -18982.510 0.012 0.014
Chain 1: 3500 -19595.748 0.014 0.015
Chain 1: 3600 -18900.672 0.016 0.015
Chain 1: 3700 -19288.656 0.018 0.016
Chain 1: 3800 -18245.546 0.022 0.020
Chain 1: 3900 -18241.592 0.020 0.020
Chain 1: 4000 -18358.942 0.021 0.020
Chain 1: 4100 -18272.459 0.021 0.020
Chain 1: 4200 -18088.210 0.021 0.020
Chain 1: 4300 -18227.026 0.020 0.020
Chain 1: 4400 -18183.335 0.018 0.010
Chain 1: 4500 -18085.752 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001383 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.83 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12444.471 1.000 1.000
Chain 1: 200 -9393.230 0.662 1.000
Chain 1: 300 -7925.358 0.503 0.325
Chain 1: 400 -8150.916 0.384 0.325
Chain 1: 500 -7901.306 0.314 0.185
Chain 1: 600 -7866.724 0.262 0.185
Chain 1: 700 -7805.229 0.226 0.032
Chain 1: 800 -7741.860 0.199 0.032
Chain 1: 900 -7731.057 0.177 0.028
Chain 1: 1000 -7977.258 0.162 0.031
Chain 1: 1100 -7899.593 0.063 0.028
Chain 1: 1200 -7817.162 0.032 0.011
Chain 1: 1300 -7738.887 0.014 0.010
Chain 1: 1400 -7767.670 0.012 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00138 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.8 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58127.855 1.000 1.000
Chain 1: 200 -17736.451 1.639 2.277
Chain 1: 300 -8711.263 1.438 1.036
Chain 1: 400 -8106.204 1.097 1.036
Chain 1: 500 -8777.580 0.893 1.000
Chain 1: 600 -8630.671 0.747 1.000
Chain 1: 700 -7775.951 0.656 0.110
Chain 1: 800 -8236.777 0.581 0.110
Chain 1: 900 -7790.131 0.523 0.076
Chain 1: 1000 -7764.780 0.471 0.076
Chain 1: 1100 -7857.671 0.372 0.075
Chain 1: 1200 -7742.408 0.146 0.057
Chain 1: 1300 -7819.654 0.043 0.056
Chain 1: 1400 -7688.237 0.037 0.017
Chain 1: 1500 -7541.875 0.032 0.017
Chain 1: 1600 -7770.728 0.033 0.019
Chain 1: 1700 -7538.067 0.025 0.019
Chain 1: 1800 -7649.010 0.021 0.017
Chain 1: 1900 -7525.804 0.017 0.016
Chain 1: 2000 -7600.529 0.017 0.016
Chain 1: 2100 -7628.488 0.017 0.016
Chain 1: 2200 -7687.459 0.016 0.016
Chain 1: 2300 -7550.635 0.017 0.017
Chain 1: 2400 -7607.138 0.016 0.016
Chain 1: 2500 -7607.391 0.014 0.015
Chain 1: 2600 -7509.349 0.012 0.013
Chain 1: 2700 -7548.960 0.010 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003253 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.53 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86446.964 1.000 1.000
Chain 1: 200 -13571.952 3.185 5.370
Chain 1: 300 -9862.288 2.249 1.000
Chain 1: 400 -11233.387 1.717 1.000
Chain 1: 500 -8838.037 1.428 0.376
Chain 1: 600 -8388.476 1.199 0.376
Chain 1: 700 -8437.641 1.028 0.271
Chain 1: 800 -8169.900 0.904 0.271
Chain 1: 900 -8215.081 0.804 0.122
Chain 1: 1000 -8631.149 0.728 0.122
Chain 1: 1100 -8648.418 0.629 0.054 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8242.098 0.097 0.049
Chain 1: 1300 -8540.471 0.063 0.048
Chain 1: 1400 -8501.551 0.051 0.035
Chain 1: 1500 -8381.457 0.025 0.033
Chain 1: 1600 -8485.281 0.021 0.014
Chain 1: 1700 -8555.805 0.021 0.014
Chain 1: 1800 -8120.630 0.023 0.014
Chain 1: 1900 -8225.128 0.024 0.014
Chain 1: 2000 -8201.146 0.019 0.013
Chain 1: 2100 -8347.967 0.021 0.014
Chain 1: 2200 -8132.988 0.019 0.014
Chain 1: 2300 -8288.599 0.017 0.014
Chain 1: 2400 -8128.069 0.019 0.018
Chain 1: 2500 -8199.087 0.018 0.018
Chain 1: 2600 -8111.364 0.018 0.018
Chain 1: 2700 -8145.431 0.018 0.018
Chain 1: 2800 -8105.559 0.013 0.013
Chain 1: 2900 -8198.730 0.013 0.011
Chain 1: 3000 -8030.741 0.014 0.018
Chain 1: 3100 -8188.231 0.015 0.019
Chain 1: 3200 -8060.242 0.013 0.016
Chain 1: 3300 -8067.938 0.012 0.011
Chain 1: 3400 -8226.528 0.012 0.011
Chain 1: 3500 -8232.364 0.011 0.011
Chain 1: 3600 -8016.844 0.012 0.016
Chain 1: 3700 -8162.398 0.014 0.018
Chain 1: 3800 -8023.423 0.015 0.018
Chain 1: 3900 -7958.073 0.015 0.018
Chain 1: 4000 -8033.216 0.014 0.017
Chain 1: 4100 -8023.798 0.012 0.016
Chain 1: 4200 -8013.753 0.010 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003058 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.58 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8395358.338 1.000 1.000
Chain 1: 200 -1583887.303 2.650 4.300
Chain 1: 300 -891195.132 2.026 1.000
Chain 1: 400 -457996.043 1.756 1.000
Chain 1: 500 -358375.485 1.460 0.946
Chain 1: 600 -233385.320 1.306 0.946
Chain 1: 700 -119477.827 1.256 0.946
Chain 1: 800 -86644.951 1.146 0.946
Chain 1: 900 -66964.464 1.051 0.777
Chain 1: 1000 -51743.716 0.976 0.777
Chain 1: 1100 -39198.982 0.908 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38377.239 0.480 0.379
Chain 1: 1300 -26306.902 0.448 0.379
Chain 1: 1400 -26026.120 0.354 0.320
Chain 1: 1500 -22605.304 0.342 0.320
Chain 1: 1600 -21820.145 0.292 0.294
Chain 1: 1700 -20690.193 0.202 0.294
Chain 1: 1800 -20633.811 0.164 0.151
Chain 1: 1900 -20960.325 0.137 0.055
Chain 1: 2000 -19468.687 0.115 0.055
Chain 1: 2100 -19707.421 0.084 0.036
Chain 1: 2200 -19934.333 0.083 0.036
Chain 1: 2300 -19550.987 0.039 0.020
Chain 1: 2400 -19322.877 0.039 0.020
Chain 1: 2500 -19124.940 0.025 0.016
Chain 1: 2600 -18754.771 0.023 0.016
Chain 1: 2700 -18711.573 0.018 0.012
Chain 1: 2800 -18428.247 0.019 0.015
Chain 1: 2900 -18709.724 0.019 0.015
Chain 1: 3000 -18695.932 0.012 0.012
Chain 1: 3100 -18780.974 0.011 0.012
Chain 1: 3200 -18471.386 0.012 0.015
Chain 1: 3300 -18676.305 0.011 0.012
Chain 1: 3400 -18150.777 0.012 0.015
Chain 1: 3500 -18763.371 0.015 0.015
Chain 1: 3600 -18069.091 0.017 0.015
Chain 1: 3700 -18456.623 0.018 0.017
Chain 1: 3800 -17414.882 0.023 0.021
Chain 1: 3900 -17410.964 0.021 0.021
Chain 1: 4000 -17528.278 0.022 0.021
Chain 1: 4100 -17441.976 0.022 0.021
Chain 1: 4200 -17257.874 0.021 0.021
Chain 1: 4300 -17396.520 0.021 0.021
Chain 1: 4400 -17353.095 0.018 0.011
Chain 1: 4500 -17255.558 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001279 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.79 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48831.976 1.000 1.000
Chain 1: 200 -18233.351 1.339 1.678
Chain 1: 300 -20320.299 0.927 1.000
Chain 1: 400 -15228.515 0.779 1.000
Chain 1: 500 -22604.704 0.688 0.334
Chain 1: 600 -12796.551 0.701 0.766
Chain 1: 700 -12661.988 0.603 0.334
Chain 1: 800 -15172.347 0.548 0.334
Chain 1: 900 -14940.217 0.489 0.326
Chain 1: 1000 -11839.535 0.466 0.326
Chain 1: 1100 -10829.856 0.375 0.262
Chain 1: 1200 -10563.662 0.210 0.165
Chain 1: 1300 -17370.812 0.239 0.262
Chain 1: 1400 -10931.520 0.265 0.262
Chain 1: 1500 -11567.954 0.237 0.165
Chain 1: 1600 -13092.627 0.172 0.116
Chain 1: 1700 -11898.880 0.181 0.116
Chain 1: 1800 -11040.641 0.173 0.100
Chain 1: 1900 -11122.061 0.172 0.100
Chain 1: 2000 -9299.969 0.165 0.100
Chain 1: 2100 -10379.060 0.166 0.104
Chain 1: 2200 -10584.068 0.166 0.104
Chain 1: 2300 -9344.767 0.140 0.104
Chain 1: 2400 -9287.089 0.081 0.100
Chain 1: 2500 -9869.794 0.082 0.100
Chain 1: 2600 -10025.006 0.072 0.078
Chain 1: 2700 -9104.615 0.072 0.078
Chain 1: 2800 -13644.370 0.097 0.101
Chain 1: 2900 -9607.307 0.139 0.104
Chain 1: 3000 -21064.268 0.173 0.104
Chain 1: 3100 -9488.858 0.285 0.133
Chain 1: 3200 -12619.706 0.308 0.248
Chain 1: 3300 -10465.982 0.315 0.248
Chain 1: 3400 -9709.163 0.322 0.248
Chain 1: 3500 -8988.827 0.325 0.248
Chain 1: 3600 -8841.518 0.325 0.248
Chain 1: 3700 -9243.245 0.319 0.248
Chain 1: 3800 -8702.582 0.292 0.206
Chain 1: 3900 -13525.012 0.285 0.206
Chain 1: 4000 -10007.262 0.266 0.206
Chain 1: 4100 -8712.561 0.159 0.149
Chain 1: 4200 -8879.709 0.136 0.080
Chain 1: 4300 -9760.700 0.125 0.080
Chain 1: 4400 -8957.795 0.126 0.090
Chain 1: 4500 -8532.758 0.123 0.090
Chain 1: 4600 -12207.942 0.151 0.090
Chain 1: 4700 -10167.184 0.167 0.149
Chain 1: 4800 -8941.288 0.174 0.149
Chain 1: 4900 -10130.491 0.150 0.137
Chain 1: 5000 -8767.010 0.131 0.137
Chain 1: 5100 -8357.592 0.121 0.117
Chain 1: 5200 -12109.474 0.150 0.137
Chain 1: 5300 -11835.519 0.143 0.137
Chain 1: 5400 -8956.458 0.167 0.156
Chain 1: 5500 -8379.175 0.168 0.156
Chain 1: 5600 -11600.121 0.166 0.156
Chain 1: 5700 -8339.200 0.185 0.156
Chain 1: 5800 -9222.280 0.181 0.156
Chain 1: 5900 -14164.018 0.204 0.278
Chain 1: 6000 -9481.654 0.238 0.310
Chain 1: 6100 -13920.831 0.265 0.319
Chain 1: 6200 -8328.315 0.301 0.321
Chain 1: 6300 -11554.515 0.327 0.321
Chain 1: 6400 -13095.787 0.306 0.319
Chain 1: 6500 -11740.644 0.311 0.319
Chain 1: 6600 -8433.685 0.322 0.349
Chain 1: 6700 -8292.248 0.285 0.319
Chain 1: 6800 -8426.442 0.277 0.319
Chain 1: 6900 -8265.911 0.244 0.279
Chain 1: 7000 -8827.324 0.201 0.118
Chain 1: 7100 -10150.382 0.182 0.118
Chain 1: 7200 -10271.887 0.116 0.115
Chain 1: 7300 -8538.876 0.109 0.115
Chain 1: 7400 -8892.548 0.101 0.064
Chain 1: 7500 -8122.458 0.099 0.064
Chain 1: 7600 -11198.878 0.087 0.064
Chain 1: 7700 -8634.769 0.115 0.095
Chain 1: 7800 -11111.593 0.136 0.130
Chain 1: 7900 -8231.533 0.169 0.203
Chain 1: 8000 -11944.817 0.194 0.223
Chain 1: 8100 -8400.429 0.223 0.275
Chain 1: 8200 -8643.119 0.224 0.275
Chain 1: 8300 -8254.503 0.209 0.275
Chain 1: 8400 -8675.531 0.210 0.275
Chain 1: 8500 -11542.848 0.225 0.275
Chain 1: 8600 -8709.018 0.230 0.297
Chain 1: 8700 -8909.366 0.203 0.248
Chain 1: 8800 -9954.471 0.191 0.248
Chain 1: 8900 -9969.792 0.156 0.105
Chain 1: 9000 -10214.259 0.127 0.049
Chain 1: 9100 -8424.868 0.106 0.049
Chain 1: 9200 -8217.189 0.106 0.049
Chain 1: 9300 -8982.399 0.110 0.085
Chain 1: 9400 -8377.707 0.112 0.085
Chain 1: 9500 -8224.569 0.089 0.072
Chain 1: 9600 -9190.266 0.067 0.072
Chain 1: 9700 -8146.658 0.078 0.085
Chain 1: 9800 -9068.766 0.077 0.085
Chain 1: 9900 -8168.631 0.088 0.102
Chain 1: 10000 -9074.738 0.096 0.102
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001431 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.31 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61546.934 1.000 1.000
Chain 1: 200 -17773.551 1.731 2.463
Chain 1: 300 -8826.227 1.492 1.014
Chain 1: 400 -9211.806 1.130 1.014
Chain 1: 500 -8079.313 0.932 1.000
Chain 1: 600 -8479.489 0.784 1.000
Chain 1: 700 -8108.949 0.679 0.140
Chain 1: 800 -8128.908 0.594 0.140
Chain 1: 900 -7960.241 0.531 0.047
Chain 1: 1000 -7728.764 0.481 0.047
Chain 1: 1100 -7769.903 0.381 0.046
Chain 1: 1200 -7741.501 0.135 0.042
Chain 1: 1300 -7549.080 0.036 0.030
Chain 1: 1400 -7846.812 0.036 0.030
Chain 1: 1500 -7590.254 0.025 0.030
Chain 1: 1600 -7509.772 0.022 0.025
Chain 1: 1700 -7643.696 0.019 0.021
Chain 1: 1800 -7616.343 0.019 0.021
Chain 1: 1900 -7471.592 0.019 0.019
Chain 1: 2000 -7563.629 0.017 0.018
Chain 1: 2100 -7582.246 0.017 0.018
Chain 1: 2200 -7684.730 0.018 0.018
Chain 1: 2300 -7584.412 0.016 0.013
Chain 1: 2400 -7586.537 0.013 0.013
Chain 1: 2500 -7612.972 0.010 0.012 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004512 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 45.12 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86261.120 1.000 1.000
Chain 1: 200 -13427.950 3.212 5.424
Chain 1: 300 -9787.316 2.265 1.000
Chain 1: 400 -10811.558 1.723 1.000
Chain 1: 500 -8667.085 1.428 0.372
Chain 1: 600 -8641.288 1.190 0.372
Chain 1: 700 -8282.515 1.026 0.247
Chain 1: 800 -8683.486 0.904 0.247
Chain 1: 900 -8589.996 0.805 0.095
Chain 1: 1000 -8450.028 0.726 0.095
Chain 1: 1100 -8610.336 0.628 0.046 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8432.824 0.087 0.043
Chain 1: 1300 -8545.471 0.051 0.021
Chain 1: 1400 -8308.127 0.045 0.021
Chain 1: 1500 -8314.234 0.020 0.019
Chain 1: 1600 -8300.185 0.020 0.019
Chain 1: 1700 -8200.370 0.017 0.017
Chain 1: 1800 -8101.703 0.014 0.013
Chain 1: 1900 -8225.452 0.014 0.015
Chain 1: 2000 -8189.346 0.013 0.013
Chain 1: 2100 -8321.005 0.012 0.013
Chain 1: 2200 -8134.568 0.013 0.013
Chain 1: 2300 -8214.181 0.012 0.012
Chain 1: 2400 -8283.652 0.010 0.012
Chain 1: 2500 -8229.145 0.011 0.012
Chain 1: 2600 -8228.783 0.011 0.012
Chain 1: 2700 -8146.151 0.011 0.010
Chain 1: 2800 -8107.973 0.010 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003249 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.49 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8426830.443 1.000 1.000
Chain 1: 200 -1587903.179 2.653 4.307
Chain 1: 300 -891485.160 2.029 1.000
Chain 1: 400 -457857.754 1.759 1.000
Chain 1: 500 -358069.471 1.463 0.947
Chain 1: 600 -232816.165 1.309 0.947
Chain 1: 700 -119059.454 1.258 0.947
Chain 1: 800 -86317.002 1.148 0.947
Chain 1: 900 -66664.828 1.053 0.781
Chain 1: 1000 -51477.333 0.978 0.781
Chain 1: 1100 -38972.539 0.910 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38147.591 0.481 0.379
Chain 1: 1300 -26119.929 0.449 0.379
Chain 1: 1400 -25839.548 0.356 0.321
Chain 1: 1500 -22432.425 0.343 0.321
Chain 1: 1600 -21650.707 0.293 0.295
Chain 1: 1700 -20526.292 0.203 0.295
Chain 1: 1800 -20470.829 0.165 0.152
Chain 1: 1900 -20796.982 0.137 0.055
Chain 1: 2000 -19309.379 0.115 0.055
Chain 1: 2100 -19547.463 0.084 0.036
Chain 1: 2200 -19773.975 0.083 0.036
Chain 1: 2300 -19391.157 0.039 0.020
Chain 1: 2400 -19163.283 0.039 0.020
Chain 1: 2500 -18965.334 0.025 0.016
Chain 1: 2600 -18595.453 0.024 0.016
Chain 1: 2700 -18552.385 0.018 0.012
Chain 1: 2800 -18269.318 0.020 0.015
Chain 1: 2900 -18550.504 0.020 0.015
Chain 1: 3000 -18536.635 0.012 0.012
Chain 1: 3100 -18621.671 0.011 0.012
Chain 1: 3200 -18312.307 0.012 0.015
Chain 1: 3300 -18517.054 0.011 0.012
Chain 1: 3400 -17991.968 0.013 0.015
Chain 1: 3500 -18603.868 0.015 0.015
Chain 1: 3600 -17910.473 0.017 0.015
Chain 1: 3700 -18297.332 0.019 0.017
Chain 1: 3800 -17256.973 0.023 0.021
Chain 1: 3900 -17253.119 0.022 0.021
Chain 1: 4000 -17370.404 0.022 0.021
Chain 1: 4100 -17284.217 0.022 0.021
Chain 1: 4200 -17100.409 0.022 0.021
Chain 1: 4300 -17238.821 0.021 0.021
Chain 1: 4400 -17195.616 0.019 0.011
Chain 1: 4500 -17098.162 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001434 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.34 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12695.621 1.000 1.000
Chain 1: 200 -9584.486 0.662 1.000
Chain 1: 300 -8280.518 0.494 0.325
Chain 1: 400 -8558.158 0.379 0.325
Chain 1: 500 -8388.997 0.307 0.157
Chain 1: 600 -8267.901 0.258 0.157
Chain 1: 700 -8164.310 0.223 0.032
Chain 1: 800 -8171.887 0.195 0.032
Chain 1: 900 -8110.670 0.174 0.020
Chain 1: 1000 -8288.645 0.159 0.021
Chain 1: 1100 -8304.025 0.059 0.020
Chain 1: 1200 -8186.442 0.028 0.015
Chain 1: 1300 -8162.628 0.013 0.014
Chain 1: 1400 -8168.730 0.010 0.013 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001537 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.37 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -46706.470 1.000 1.000
Chain 1: 200 -15916.035 1.467 1.935
Chain 1: 300 -8866.661 1.243 1.000
Chain 1: 400 -8271.073 0.950 1.000
Chain 1: 500 -8811.044 0.773 0.795
Chain 1: 600 -9491.662 0.656 0.795
Chain 1: 700 -8554.450 0.578 0.110
Chain 1: 800 -8080.387 0.513 0.110
Chain 1: 900 -7853.199 0.459 0.072
Chain 1: 1000 -7972.309 0.415 0.072
Chain 1: 1100 -7689.346 0.318 0.072
Chain 1: 1200 -7677.683 0.125 0.061
Chain 1: 1300 -7847.235 0.048 0.059
Chain 1: 1400 -8019.968 0.043 0.037
Chain 1: 1500 -7595.915 0.042 0.037
Chain 1: 1600 -7771.613 0.037 0.029
Chain 1: 1700 -7626.128 0.028 0.023
Chain 1: 1800 -7713.915 0.023 0.022
Chain 1: 1900 -7636.618 0.022 0.022
Chain 1: 2000 -7680.857 0.021 0.022
Chain 1: 2100 -7615.038 0.018 0.019
Chain 1: 2200 -7767.014 0.020 0.020
Chain 1: 2300 -7600.421 0.020 0.020
Chain 1: 2400 -7644.758 0.018 0.019
Chain 1: 2500 -7659.443 0.013 0.011
Chain 1: 2600 -7555.501 0.012 0.011
Chain 1: 2700 -7556.329 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003698 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 36.98 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86944.443 1.000 1.000
Chain 1: 200 -13843.906 3.140 5.280
Chain 1: 300 -10190.570 2.213 1.000
Chain 1: 400 -11048.955 1.679 1.000
Chain 1: 500 -9129.981 1.385 0.359
Chain 1: 600 -8684.293 1.163 0.359
Chain 1: 700 -8681.261 0.997 0.210
Chain 1: 800 -9244.240 0.880 0.210
Chain 1: 900 -8969.644 0.786 0.078
Chain 1: 1000 -8640.256 0.711 0.078
Chain 1: 1100 -8800.907 0.613 0.061 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8613.700 0.087 0.051
Chain 1: 1300 -8872.022 0.054 0.038
Chain 1: 1400 -8886.580 0.046 0.031
Chain 1: 1500 -8733.104 0.027 0.029
Chain 1: 1600 -8846.865 0.023 0.022
Chain 1: 1700 -8924.467 0.024 0.022
Chain 1: 1800 -8502.342 0.023 0.022
Chain 1: 1900 -8602.692 0.021 0.018
Chain 1: 2000 -8577.067 0.017 0.018
Chain 1: 2100 -8702.114 0.017 0.014
Chain 1: 2200 -8507.370 0.017 0.014
Chain 1: 2300 -8597.418 0.015 0.013
Chain 1: 2400 -8666.454 0.016 0.013
Chain 1: 2500 -8612.662 0.015 0.012
Chain 1: 2600 -8613.713 0.014 0.010
Chain 1: 2700 -8530.559 0.014 0.010
Chain 1: 2800 -8490.908 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002625 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 26.25 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8394990.166 1.000 1.000
Chain 1: 200 -1580408.983 2.656 4.312
Chain 1: 300 -890109.037 2.029 1.000
Chain 1: 400 -457546.746 1.758 1.000
Chain 1: 500 -358439.212 1.462 0.945
Chain 1: 600 -233372.099 1.308 0.945
Chain 1: 700 -119603.434 1.257 0.945
Chain 1: 800 -86836.157 1.147 0.945
Chain 1: 900 -67164.899 1.052 0.776
Chain 1: 1000 -51956.979 0.976 0.776
Chain 1: 1100 -39430.152 0.908 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38605.301 0.479 0.377
Chain 1: 1300 -26549.400 0.447 0.377
Chain 1: 1400 -26267.407 0.353 0.318
Chain 1: 1500 -22852.135 0.340 0.318
Chain 1: 1600 -22068.304 0.290 0.293
Chain 1: 1700 -20940.030 0.201 0.293
Chain 1: 1800 -20883.890 0.163 0.149
Chain 1: 1900 -21210.198 0.135 0.054
Chain 1: 2000 -19720.273 0.114 0.054
Chain 1: 2100 -19958.558 0.083 0.036
Chain 1: 2200 -20185.426 0.082 0.036
Chain 1: 2300 -19802.233 0.039 0.019
Chain 1: 2400 -19574.262 0.039 0.019
Chain 1: 2500 -19376.413 0.025 0.015
Chain 1: 2600 -19006.356 0.023 0.015
Chain 1: 2700 -18963.238 0.018 0.012
Chain 1: 2800 -18680.183 0.019 0.015
Chain 1: 2900 -18961.436 0.019 0.015
Chain 1: 3000 -18947.526 0.012 0.012
Chain 1: 3100 -19032.595 0.011 0.012
Chain 1: 3200 -18723.167 0.011 0.015
Chain 1: 3300 -18927.960 0.011 0.012
Chain 1: 3400 -18402.795 0.012 0.015
Chain 1: 3500 -19014.934 0.015 0.015
Chain 1: 3600 -18321.206 0.016 0.015
Chain 1: 3700 -18708.350 0.018 0.017
Chain 1: 3800 -17667.564 0.023 0.021
Chain 1: 3900 -17663.715 0.021 0.021
Chain 1: 4000 -17780.965 0.022 0.021
Chain 1: 4100 -17694.786 0.022 0.021
Chain 1: 4200 -17510.879 0.021 0.021
Chain 1: 4300 -17649.361 0.021 0.021
Chain 1: 4400 -17606.087 0.018 0.011
Chain 1: 4500 -17508.616 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001364 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.64 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12428.839 1.000 1.000
Chain 1: 200 -9372.486 0.663 1.000
Chain 1: 300 -8157.642 0.492 0.326
Chain 1: 400 -8271.470 0.372 0.326
Chain 1: 500 -8130.792 0.301 0.149
Chain 1: 600 -8045.811 0.253 0.149
Chain 1: 700 -7949.983 0.218 0.017
Chain 1: 800 -7995.077 0.192 0.017
Chain 1: 900 -8119.865 0.172 0.015
Chain 1: 1000 -8053.841 0.156 0.015
Chain 1: 1100 -8050.277 0.056 0.014
Chain 1: 1200 -7971.676 0.024 0.012
Chain 1: 1300 -7923.239 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001436 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.36 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56743.784 1.000 1.000
Chain 1: 200 -17439.359 1.627 2.254
Chain 1: 300 -8773.194 1.414 1.000
Chain 1: 400 -8425.546 1.071 1.000
Chain 1: 500 -8794.337 0.865 0.988
Chain 1: 600 -8990.011 0.724 0.988
Chain 1: 700 -8109.301 0.636 0.109
Chain 1: 800 -8227.293 0.559 0.109
Chain 1: 900 -7953.750 0.500 0.042
Chain 1: 1000 -7828.659 0.452 0.042
Chain 1: 1100 -7617.235 0.355 0.041
Chain 1: 1200 -7900.878 0.133 0.036
Chain 1: 1300 -7746.400 0.036 0.034
Chain 1: 1400 -7728.038 0.032 0.028
Chain 1: 1500 -7609.229 0.030 0.022
Chain 1: 1600 -7762.328 0.029 0.020
Chain 1: 1700 -7547.529 0.021 0.020
Chain 1: 1800 -7745.834 0.023 0.026
Chain 1: 1900 -7639.066 0.021 0.020
Chain 1: 2000 -7628.125 0.019 0.020
Chain 1: 2100 -7610.369 0.017 0.020
Chain 1: 2200 -7735.149 0.015 0.016
Chain 1: 2300 -7633.751 0.014 0.016
Chain 1: 2400 -7677.065 0.014 0.016
Chain 1: 2500 -7593.003 0.014 0.014
Chain 1: 2600 -7574.543 0.012 0.013
Chain 1: 2700 -7575.312 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003849 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 38.49 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87001.298 1.000 1.000
Chain 1: 200 -13549.677 3.210 5.421
Chain 1: 300 -9915.299 2.262 1.000
Chain 1: 400 -10787.581 1.717 1.000
Chain 1: 500 -8859.371 1.417 0.367
Chain 1: 600 -8411.705 1.190 0.367
Chain 1: 700 -8490.815 1.021 0.218
Chain 1: 800 -8727.387 0.897 0.218
Chain 1: 900 -8704.408 0.798 0.081
Chain 1: 1000 -8543.144 0.720 0.081
Chain 1: 1100 -8774.701 0.622 0.053 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8292.547 0.086 0.053
Chain 1: 1300 -8582.109 0.053 0.034
Chain 1: 1400 -8608.518 0.045 0.027
Chain 1: 1500 -8500.377 0.025 0.026
Chain 1: 1600 -8609.762 0.020 0.019
Chain 1: 1700 -8688.914 0.020 0.019
Chain 1: 1800 -8278.188 0.023 0.019
Chain 1: 1900 -8374.178 0.024 0.019
Chain 1: 2000 -8347.021 0.022 0.013
Chain 1: 2100 -8469.075 0.021 0.013
Chain 1: 2200 -8288.702 0.017 0.013
Chain 1: 2300 -8370.047 0.015 0.013
Chain 1: 2400 -8438.902 0.015 0.013
Chain 1: 2500 -8384.308 0.015 0.011
Chain 1: 2600 -8383.138 0.013 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003493 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.93 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8409694.373 1.000 1.000
Chain 1: 200 -1590375.768 2.644 4.288
Chain 1: 300 -892862.480 2.023 1.000
Chain 1: 400 -458055.988 1.755 1.000
Chain 1: 500 -357636.216 1.460 0.949
Chain 1: 600 -232354.902 1.306 0.949
Chain 1: 700 -118908.452 1.256 0.949
Chain 1: 800 -86169.897 1.147 0.949
Chain 1: 900 -66590.262 1.052 0.781
Chain 1: 1000 -51456.703 0.976 0.781
Chain 1: 1100 -38993.163 0.908 0.539 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38181.835 0.481 0.380
Chain 1: 1300 -26199.646 0.449 0.380
Chain 1: 1400 -25925.585 0.355 0.320
Chain 1: 1500 -22527.534 0.342 0.320
Chain 1: 1600 -21748.807 0.292 0.294
Chain 1: 1700 -20629.779 0.202 0.294
Chain 1: 1800 -20575.780 0.164 0.151
Chain 1: 1900 -20901.861 0.136 0.054
Chain 1: 2000 -19416.804 0.114 0.054
Chain 1: 2100 -19655.170 0.084 0.036
Chain 1: 2200 -19880.831 0.083 0.036
Chain 1: 2300 -19498.737 0.039 0.020
Chain 1: 2400 -19270.890 0.039 0.020
Chain 1: 2500 -19072.597 0.025 0.016
Chain 1: 2600 -18703.177 0.023 0.016
Chain 1: 2700 -18660.345 0.018 0.012
Chain 1: 2800 -18377.002 0.019 0.015
Chain 1: 2900 -18658.193 0.019 0.015
Chain 1: 3000 -18644.494 0.012 0.012
Chain 1: 3100 -18729.431 0.011 0.012
Chain 1: 3200 -18420.240 0.012 0.015
Chain 1: 3300 -18624.918 0.011 0.012
Chain 1: 3400 -18099.895 0.012 0.015
Chain 1: 3500 -18711.524 0.015 0.015
Chain 1: 3600 -18018.569 0.017 0.015
Chain 1: 3700 -18404.991 0.018 0.017
Chain 1: 3800 -17365.177 0.023 0.021
Chain 1: 3900 -17361.299 0.021 0.021
Chain 1: 4000 -17478.657 0.022 0.021
Chain 1: 4100 -17392.337 0.022 0.021
Chain 1: 4200 -17208.778 0.021 0.021
Chain 1: 4300 -17347.094 0.021 0.021
Chain 1: 4400 -17304.018 0.019 0.011
Chain 1: 4500 -17206.514 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001353 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.53 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13070.517 1.000 1.000
Chain 1: 200 -9877.112 0.662 1.000
Chain 1: 300 -8914.300 0.477 0.323
Chain 1: 400 -8389.113 0.373 0.323
Chain 1: 500 -8321.795 0.300 0.108
Chain 1: 600 -8240.780 0.252 0.108
Chain 1: 700 -8108.857 0.218 0.063
Chain 1: 800 -8137.803 0.191 0.063
Chain 1: 900 -8014.732 0.172 0.016
Chain 1: 1000 -8152.078 0.156 0.017
Chain 1: 1100 -8290.288 0.058 0.017
Chain 1: 1200 -8161.457 0.027 0.016
Chain 1: 1300 -8088.378 0.017 0.016
Chain 1: 1400 -8125.061 0.012 0.015
Chain 1: 1500 -8215.753 0.012 0.015
Chain 1: 1600 -8126.514 0.012 0.015
Chain 1: 1700 -8092.993 0.011 0.011
Chain 1: 1800 -8064.393 0.011 0.011
Chain 1: 1900 -8092.311 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001394 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.94 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58559.532 1.000 1.000
Chain 1: 200 -18321.126 1.598 2.196
Chain 1: 300 -8982.538 1.412 1.040
Chain 1: 400 -8164.612 1.084 1.040
Chain 1: 500 -8358.202 0.872 1.000
Chain 1: 600 -9291.118 0.743 1.000
Chain 1: 700 -8366.854 0.653 0.110
Chain 1: 800 -8471.074 0.573 0.110
Chain 1: 900 -8073.060 0.515 0.100
Chain 1: 1000 -7833.823 0.466 0.100
Chain 1: 1100 -7991.376 0.368 0.100
Chain 1: 1200 -8061.559 0.149 0.049
Chain 1: 1300 -7839.240 0.048 0.031
Chain 1: 1400 -7783.472 0.039 0.028
Chain 1: 1500 -7575.023 0.039 0.028
Chain 1: 1600 -7842.336 0.033 0.028
Chain 1: 1700 -7627.959 0.025 0.028
Chain 1: 1800 -7765.669 0.025 0.028
Chain 1: 1900 -7658.593 0.022 0.028
Chain 1: 2000 -7725.191 0.019 0.020
Chain 1: 2100 -7643.357 0.018 0.018
Chain 1: 2200 -7878.284 0.021 0.028
Chain 1: 2300 -7597.030 0.021 0.028
Chain 1: 2400 -7605.155 0.021 0.028
Chain 1: 2500 -7616.253 0.018 0.018
Chain 1: 2600 -7592.022 0.015 0.014
Chain 1: 2700 -7501.964 0.014 0.012
Chain 1: 2800 -7578.432 0.013 0.011
Chain 1: 2900 -7431.779 0.013 0.011
Chain 1: 3000 -7590.130 0.015 0.012
Chain 1: 3100 -7585.387 0.014 0.012
Chain 1: 3200 -7807.239 0.013 0.012
Chain 1: 3300 -7525.527 0.013 0.012
Chain 1: 3400 -7772.460 0.017 0.020
Chain 1: 3500 -7498.221 0.020 0.021
Chain 1: 3600 -7561.205 0.021 0.021
Chain 1: 3700 -7513.899 0.020 0.021
Chain 1: 3800 -7520.763 0.019 0.021
Chain 1: 3900 -7474.560 0.018 0.021
Chain 1: 4000 -7461.091 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002532 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.32 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86450.043 1.000 1.000
Chain 1: 200 -14066.970 3.073 5.146
Chain 1: 300 -10291.814 2.171 1.000
Chain 1: 400 -12035.327 1.664 1.000
Chain 1: 500 -8774.832 1.406 0.372
Chain 1: 600 -8590.592 1.175 0.372
Chain 1: 700 -8842.709 1.011 0.367
Chain 1: 800 -9096.148 0.888 0.367
Chain 1: 900 -9021.388 0.791 0.145
Chain 1: 1000 -9299.173 0.714 0.145
Chain 1: 1100 -8967.082 0.618 0.037 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8600.661 0.108 0.037
Chain 1: 1300 -8912.809 0.075 0.035
Chain 1: 1400 -8820.987 0.061 0.030
Chain 1: 1500 -8772.227 0.025 0.029
Chain 1: 1600 -8869.184 0.024 0.029
Chain 1: 1700 -8924.283 0.021 0.028
Chain 1: 1800 -8466.459 0.024 0.030
Chain 1: 1900 -8578.318 0.024 0.030
Chain 1: 2000 -8583.949 0.022 0.013
Chain 1: 2100 -8528.011 0.019 0.011
Chain 1: 2200 -8500.233 0.015 0.010
Chain 1: 2300 -8684.266 0.013 0.010
Chain 1: 2400 -8474.736 0.015 0.011
Chain 1: 2500 -8546.687 0.015 0.011
Chain 1: 2600 -8459.638 0.015 0.010
Chain 1: 2700 -8495.223 0.015 0.010
Chain 1: 2800 -8446.487 0.010 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003389 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.89 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8378269.387 1.000 1.000
Chain 1: 200 -1582019.341 2.648 4.296
Chain 1: 300 -890810.439 2.024 1.000
Chain 1: 400 -457713.648 1.755 1.000
Chain 1: 500 -358479.751 1.459 0.946
Chain 1: 600 -233594.656 1.305 0.946
Chain 1: 700 -119897.419 1.254 0.946
Chain 1: 800 -87081.589 1.144 0.946
Chain 1: 900 -67435.596 1.050 0.776
Chain 1: 1000 -52239.221 0.974 0.776
Chain 1: 1100 -39704.187 0.905 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38889.933 0.478 0.377
Chain 1: 1300 -26822.975 0.445 0.377
Chain 1: 1400 -26543.933 0.352 0.316
Chain 1: 1500 -23123.329 0.339 0.316
Chain 1: 1600 -22338.521 0.289 0.291
Chain 1: 1700 -21208.840 0.199 0.291
Chain 1: 1800 -21152.763 0.162 0.148
Chain 1: 1900 -21479.698 0.134 0.053
Chain 1: 2000 -19987.368 0.113 0.053
Chain 1: 2100 -20226.179 0.082 0.035
Chain 1: 2200 -20453.286 0.081 0.035
Chain 1: 2300 -20069.706 0.038 0.019
Chain 1: 2400 -19841.478 0.038 0.019
Chain 1: 2500 -19643.436 0.024 0.015
Chain 1: 2600 -19272.989 0.023 0.015
Chain 1: 2700 -19229.744 0.018 0.012
Chain 1: 2800 -18946.179 0.019 0.015
Chain 1: 2900 -19227.886 0.019 0.015
Chain 1: 3000 -19214.048 0.012 0.012
Chain 1: 3100 -19299.112 0.011 0.012
Chain 1: 3200 -18989.332 0.011 0.015
Chain 1: 3300 -19194.438 0.010 0.012
Chain 1: 3400 -18668.443 0.012 0.015
Chain 1: 3500 -19281.678 0.014 0.015
Chain 1: 3600 -18586.649 0.016 0.015
Chain 1: 3700 -18974.732 0.018 0.016
Chain 1: 3800 -17931.710 0.022 0.020
Chain 1: 3900 -17927.774 0.021 0.020
Chain 1: 4000 -18045.110 0.021 0.020
Chain 1: 4100 -17958.690 0.021 0.020
Chain 1: 4200 -17774.361 0.021 0.020
Chain 1: 4300 -17913.192 0.021 0.020
Chain 1: 4400 -17869.548 0.018 0.010
Chain 1: 4500 -17771.972 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001341 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.41 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48729.057 1.000 1.000
Chain 1: 200 -18921.600 1.288 1.575
Chain 1: 300 -20862.794 0.889 1.000
Chain 1: 400 -19142.779 0.690 1.000
Chain 1: 500 -23699.189 0.590 0.192
Chain 1: 600 -18724.850 0.536 0.266
Chain 1: 700 -14361.051 0.503 0.266
Chain 1: 800 -14260.169 0.441 0.266
Chain 1: 900 -13260.765 0.400 0.192
Chain 1: 1000 -12425.674 0.367 0.192
Chain 1: 1100 -12651.096 0.269 0.093
Chain 1: 1200 -10634.504 0.130 0.093
Chain 1: 1300 -11219.377 0.126 0.090
Chain 1: 1400 -13817.320 0.136 0.188
Chain 1: 1500 -10115.686 0.153 0.188
Chain 1: 1600 -13344.559 0.151 0.188
Chain 1: 1700 -12172.638 0.130 0.096
Chain 1: 1800 -10583.595 0.144 0.150
Chain 1: 1900 -19978.481 0.184 0.188
Chain 1: 2000 -16058.810 0.202 0.190
Chain 1: 2100 -9440.308 0.270 0.242
Chain 1: 2200 -9370.839 0.252 0.242
Chain 1: 2300 -9509.711 0.248 0.242
Chain 1: 2400 -18279.763 0.277 0.244
Chain 1: 2500 -10831.715 0.309 0.244
Chain 1: 2600 -9060.998 0.305 0.244
Chain 1: 2700 -9174.309 0.296 0.244
Chain 1: 2800 -9368.444 0.283 0.244
Chain 1: 2900 -9570.695 0.238 0.195
Chain 1: 3000 -9095.542 0.219 0.052
Chain 1: 3100 -9951.312 0.158 0.052
Chain 1: 3200 -9177.878 0.165 0.084
Chain 1: 3300 -9170.028 0.164 0.084
Chain 1: 3400 -9775.251 0.122 0.062
Chain 1: 3500 -10324.462 0.059 0.053
Chain 1: 3600 -9325.235 0.050 0.053
Chain 1: 3700 -9319.122 0.049 0.053
Chain 1: 3800 -13480.454 0.078 0.062
Chain 1: 3900 -9839.124 0.113 0.084
Chain 1: 4000 -10408.985 0.113 0.084
Chain 1: 4100 -8865.701 0.122 0.084
Chain 1: 4200 -12574.822 0.143 0.107
Chain 1: 4300 -13721.495 0.151 0.107
Chain 1: 4400 -9122.039 0.195 0.174
Chain 1: 4500 -12572.028 0.217 0.274
Chain 1: 4600 -12562.504 0.207 0.274
Chain 1: 4700 -11952.576 0.212 0.274
Chain 1: 4800 -8547.911 0.221 0.274
Chain 1: 4900 -10130.214 0.199 0.174
Chain 1: 5000 -9264.219 0.203 0.174
Chain 1: 5100 -13135.512 0.215 0.274
Chain 1: 5200 -9568.600 0.223 0.274
Chain 1: 5300 -14321.003 0.248 0.295
Chain 1: 5400 -9557.165 0.247 0.295
Chain 1: 5500 -13280.725 0.248 0.295
Chain 1: 5600 -11797.838 0.260 0.295
Chain 1: 5700 -8713.160 0.291 0.332
Chain 1: 5800 -8508.090 0.253 0.295
Chain 1: 5900 -8901.964 0.242 0.295
Chain 1: 6000 -11263.329 0.254 0.295
Chain 1: 6100 -8473.342 0.257 0.329
Chain 1: 6200 -8431.426 0.220 0.280
Chain 1: 6300 -9546.968 0.199 0.210
Chain 1: 6400 -8308.932 0.164 0.149
Chain 1: 6500 -9365.689 0.147 0.126
Chain 1: 6600 -8564.997 0.144 0.117
Chain 1: 6700 -8804.997 0.111 0.113
Chain 1: 6800 -12998.257 0.141 0.117
Chain 1: 6900 -9987.708 0.167 0.149
Chain 1: 7000 -8741.119 0.160 0.143
Chain 1: 7100 -9341.572 0.134 0.117
Chain 1: 7200 -8559.642 0.142 0.117
Chain 1: 7300 -11633.678 0.157 0.143
Chain 1: 7400 -8857.169 0.173 0.143
Chain 1: 7500 -8248.088 0.169 0.143
Chain 1: 7600 -8512.697 0.163 0.143
Chain 1: 7700 -8837.325 0.164 0.143
Chain 1: 7800 -11741.415 0.157 0.143
Chain 1: 7900 -8462.907 0.165 0.143
Chain 1: 8000 -9958.934 0.166 0.150
Chain 1: 8100 -8318.359 0.179 0.197
Chain 1: 8200 -11886.968 0.200 0.247
Chain 1: 8300 -12251.524 0.177 0.197
Chain 1: 8400 -12983.952 0.151 0.150
Chain 1: 8500 -8409.768 0.198 0.197
Chain 1: 8600 -8234.528 0.197 0.197
Chain 1: 8700 -8333.403 0.195 0.197
Chain 1: 8800 -10261.799 0.189 0.188
Chain 1: 8900 -10218.504 0.150 0.150
Chain 1: 9000 -10854.853 0.141 0.059
Chain 1: 9100 -8875.639 0.144 0.059
Chain 1: 9200 -10168.487 0.126 0.059
Chain 1: 9300 -10247.255 0.124 0.059
Chain 1: 9400 -10997.877 0.125 0.068
Chain 1: 9500 -10211.395 0.079 0.068
Chain 1: 9600 -8348.782 0.099 0.077
Chain 1: 9700 -8414.972 0.098 0.077
Chain 1: 9800 -9685.477 0.093 0.077
Chain 1: 9900 -10189.170 0.097 0.077
Chain 1: 10000 -8608.593 0.110 0.127
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001506 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.06 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57084.050 1.000 1.000
Chain 1: 200 -17416.317 1.639 2.278
Chain 1: 300 -8733.392 1.424 1.000
Chain 1: 400 -8400.804 1.078 1.000
Chain 1: 500 -8411.654 0.863 0.994
Chain 1: 600 -8578.139 0.722 0.994
Chain 1: 700 -8014.232 0.629 0.070
Chain 1: 800 -8968.559 0.564 0.106
Chain 1: 900 -7975.121 0.515 0.106
Chain 1: 1000 -7934.346 0.464 0.106
Chain 1: 1100 -8174.990 0.367 0.070
Chain 1: 1200 -7605.888 0.147 0.070
Chain 1: 1300 -7929.576 0.051 0.041
Chain 1: 1400 -7937.908 0.047 0.041
Chain 1: 1500 -7636.515 0.051 0.041
Chain 1: 1600 -7713.793 0.050 0.041
Chain 1: 1700 -7555.274 0.045 0.039
Chain 1: 1800 -7635.892 0.036 0.029
Chain 1: 1900 -7596.729 0.024 0.021
Chain 1: 2000 -7672.713 0.024 0.021
Chain 1: 2100 -7625.110 0.022 0.011
Chain 1: 2200 -7733.063 0.016 0.011
Chain 1: 2300 -7641.667 0.013 0.011
Chain 1: 2400 -7682.107 0.013 0.011
Chain 1: 2500 -7616.982 0.010 0.010
Chain 1: 2600 -7575.502 0.010 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.005211 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 52.11 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86139.230 1.000 1.000
Chain 1: 200 -13445.312 3.203 5.407
Chain 1: 300 -9850.619 2.257 1.000
Chain 1: 400 -10780.708 1.714 1.000
Chain 1: 500 -8796.762 1.417 0.365
Chain 1: 600 -8360.471 1.189 0.365
Chain 1: 700 -8469.255 1.021 0.226
Chain 1: 800 -9053.172 0.902 0.226
Chain 1: 900 -8638.737 0.807 0.086
Chain 1: 1000 -8472.409 0.728 0.086
Chain 1: 1100 -8696.278 0.631 0.064 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8325.336 0.094 0.052
Chain 1: 1300 -8556.765 0.061 0.048
Chain 1: 1400 -8557.253 0.052 0.045
Chain 1: 1500 -8452.255 0.031 0.027
Chain 1: 1600 -8553.800 0.027 0.026
Chain 1: 1700 -8643.230 0.026 0.026
Chain 1: 1800 -8238.776 0.025 0.026
Chain 1: 1900 -8337.953 0.021 0.020
Chain 1: 2000 -8309.321 0.020 0.012
Chain 1: 2100 -8429.118 0.018 0.012
Chain 1: 2200 -8223.457 0.017 0.012
Chain 1: 2300 -8371.366 0.016 0.012
Chain 1: 2400 -8248.816 0.017 0.014
Chain 1: 2500 -8313.045 0.017 0.014
Chain 1: 2600 -8336.177 0.016 0.014
Chain 1: 2700 -8254.801 0.016 0.014
Chain 1: 2800 -8227.771 0.011 0.012
Chain 1: 2900 -8283.168 0.011 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003729 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 37.29 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8408858.037 1.000 1.000
Chain 1: 200 -1587116.580 2.649 4.298
Chain 1: 300 -891753.894 2.026 1.000
Chain 1: 400 -457774.180 1.756 1.000
Chain 1: 500 -357882.144 1.461 0.948
Chain 1: 600 -232728.761 1.307 0.948
Chain 1: 700 -119043.326 1.257 0.948
Chain 1: 800 -86287.851 1.147 0.948
Chain 1: 900 -66654.266 1.052 0.780
Chain 1: 1000 -51472.379 0.977 0.780
Chain 1: 1100 -38967.462 0.909 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38144.973 0.481 0.380
Chain 1: 1300 -26120.976 0.449 0.380
Chain 1: 1400 -25840.747 0.355 0.321
Chain 1: 1500 -22433.522 0.343 0.321
Chain 1: 1600 -21651.490 0.293 0.295
Chain 1: 1700 -20527.744 0.203 0.295
Chain 1: 1800 -20472.378 0.165 0.152
Chain 1: 1900 -20798.325 0.137 0.055
Chain 1: 2000 -19311.271 0.115 0.055
Chain 1: 2100 -19549.446 0.084 0.036
Chain 1: 2200 -19775.641 0.083 0.036
Chain 1: 2300 -19393.173 0.039 0.020
Chain 1: 2400 -19165.385 0.039 0.020
Chain 1: 2500 -18967.360 0.025 0.016
Chain 1: 2600 -18597.803 0.024 0.016
Chain 1: 2700 -18554.870 0.018 0.012
Chain 1: 2800 -18271.814 0.020 0.015
Chain 1: 2900 -18552.936 0.020 0.015
Chain 1: 3000 -18539.148 0.012 0.012
Chain 1: 3100 -18624.104 0.011 0.012
Chain 1: 3200 -18314.937 0.012 0.015
Chain 1: 3300 -18519.557 0.011 0.012
Chain 1: 3400 -17994.737 0.013 0.015
Chain 1: 3500 -18606.187 0.015 0.015
Chain 1: 3600 -17913.467 0.017 0.015
Chain 1: 3700 -18299.800 0.019 0.017
Chain 1: 3800 -17260.404 0.023 0.021
Chain 1: 3900 -17256.586 0.022 0.021
Chain 1: 4000 -17373.892 0.022 0.021
Chain 1: 4100 -17287.684 0.022 0.021
Chain 1: 4200 -17104.154 0.022 0.021
Chain 1: 4300 -17242.394 0.021 0.021
Chain 1: 4400 -17199.373 0.019 0.011
Chain 1: 4500 -17101.950 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001455 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.55 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49269.116 1.000 1.000
Chain 1: 200 -21151.120 1.165 1.329
Chain 1: 300 -13609.653 0.961 1.000
Chain 1: 400 -12697.689 0.739 1.000
Chain 1: 500 -17507.053 0.646 0.554
Chain 1: 600 -16002.842 0.554 0.554
Chain 1: 700 -12428.375 0.516 0.288
Chain 1: 800 -13296.064 0.460 0.288
Chain 1: 900 -11756.986 0.423 0.275
Chain 1: 1000 -13316.780 0.392 0.275
Chain 1: 1100 -19000.413 0.322 0.275
Chain 1: 1200 -13816.172 0.227 0.275
Chain 1: 1300 -13430.246 0.174 0.131
Chain 1: 1400 -12938.778 0.171 0.131
Chain 1: 1500 -23348.977 0.188 0.131
Chain 1: 1600 -21332.858 0.188 0.131
Chain 1: 1700 -10335.409 0.266 0.131
Chain 1: 1800 -11744.879 0.271 0.131
Chain 1: 1900 -11396.573 0.261 0.120
Chain 1: 2000 -10519.616 0.258 0.120
Chain 1: 2100 -10694.428 0.230 0.095
Chain 1: 2200 -18500.315 0.234 0.095
Chain 1: 2300 -11466.459 0.293 0.120
Chain 1: 2400 -9314.941 0.312 0.231
Chain 1: 2500 -10948.418 0.282 0.149
Chain 1: 2600 -9407.145 0.289 0.164
Chain 1: 2700 -9639.492 0.185 0.149
Chain 1: 2800 -12969.550 0.199 0.164
Chain 1: 2900 -10118.259 0.224 0.231
Chain 1: 3000 -13437.879 0.241 0.247
Chain 1: 3100 -13987.269 0.243 0.247
Chain 1: 3200 -11580.329 0.221 0.231
Chain 1: 3300 -11881.409 0.163 0.208
Chain 1: 3400 -17365.038 0.171 0.208
Chain 1: 3500 -10289.193 0.225 0.247
Chain 1: 3600 -9479.562 0.217 0.247
Chain 1: 3700 -12453.686 0.239 0.247
Chain 1: 3800 -9009.317 0.251 0.247
Chain 1: 3900 -13381.206 0.256 0.247
Chain 1: 4000 -9401.338 0.273 0.316
Chain 1: 4100 -9966.682 0.275 0.316
Chain 1: 4200 -11378.344 0.267 0.316
Chain 1: 4300 -13739.335 0.281 0.316
Chain 1: 4400 -12577.741 0.259 0.239
Chain 1: 4500 -9356.519 0.225 0.239
Chain 1: 4600 -12523.259 0.241 0.253
Chain 1: 4700 -15364.439 0.236 0.253
Chain 1: 4800 -9018.209 0.268 0.253
Chain 1: 4900 -8970.025 0.236 0.185
Chain 1: 5000 -10008.918 0.204 0.172
Chain 1: 5100 -9140.514 0.208 0.172
Chain 1: 5200 -9742.396 0.202 0.172
Chain 1: 5300 -12908.695 0.209 0.185
Chain 1: 5400 -9477.444 0.236 0.245
Chain 1: 5500 -13904.785 0.233 0.245
Chain 1: 5600 -12359.491 0.221 0.185
Chain 1: 5700 -13398.860 0.210 0.125
Chain 1: 5800 -9028.752 0.188 0.125
Chain 1: 5900 -9002.536 0.188 0.125
Chain 1: 6000 -9832.619 0.186 0.125
Chain 1: 6100 -9369.358 0.181 0.125
Chain 1: 6200 -14324.132 0.210 0.245
Chain 1: 6300 -9339.637 0.238 0.318
Chain 1: 6400 -15053.334 0.240 0.318
Chain 1: 6500 -9010.985 0.275 0.346
Chain 1: 6600 -9577.948 0.269 0.346
Chain 1: 6700 -9055.963 0.267 0.346
Chain 1: 6800 -9297.978 0.221 0.084
Chain 1: 6900 -11750.139 0.242 0.209
Chain 1: 7000 -8788.866 0.267 0.337
Chain 1: 7100 -8433.709 0.266 0.337
Chain 1: 7200 -9019.558 0.238 0.209
Chain 1: 7300 -9582.077 0.190 0.065
Chain 1: 7400 -9819.085 0.155 0.059
Chain 1: 7500 -10129.935 0.091 0.059
Chain 1: 7600 -8669.249 0.102 0.059
Chain 1: 7700 -8925.581 0.099 0.059
Chain 1: 7800 -8772.030 0.098 0.059
Chain 1: 7900 -8652.194 0.079 0.042
Chain 1: 8000 -8744.005 0.046 0.031
Chain 1: 8100 -11773.575 0.067 0.031
Chain 1: 8200 -8704.228 0.096 0.031
Chain 1: 8300 -8482.123 0.093 0.029
Chain 1: 8400 -11387.703 0.116 0.031
Chain 1: 8500 -8576.528 0.146 0.168
Chain 1: 8600 -12363.273 0.160 0.255
Chain 1: 8700 -8769.666 0.198 0.257
Chain 1: 8800 -10851.495 0.215 0.257
Chain 1: 8900 -11102.559 0.216 0.257
Chain 1: 9000 -8929.022 0.239 0.257
Chain 1: 9100 -9848.114 0.223 0.255
Chain 1: 9200 -8978.199 0.197 0.243
Chain 1: 9300 -9501.053 0.200 0.243
Chain 1: 9400 -9183.971 0.178 0.192
Chain 1: 9500 -9440.448 0.148 0.097
Chain 1: 9600 -9160.131 0.121 0.093
Chain 1: 9700 -8484.987 0.087 0.080
Chain 1: 9800 -9958.458 0.083 0.080
Chain 1: 9900 -11824.906 0.097 0.093
Chain 1: 10000 -10678.650 0.083 0.093
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001497 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.97 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -51454.577 1.000 1.000
Chain 1: 200 -16641.185 1.546 2.092
Chain 1: 300 -8942.605 1.318 1.000
Chain 1: 400 -9247.784 0.996 1.000
Chain 1: 500 -8590.992 0.812 0.861
Chain 1: 600 -9358.265 0.691 0.861
Chain 1: 700 -8302.925 0.610 0.127
Chain 1: 800 -8386.393 0.535 0.127
Chain 1: 900 -7941.647 0.482 0.082
Chain 1: 1000 -7964.073 0.434 0.082
Chain 1: 1100 -7753.131 0.337 0.076
Chain 1: 1200 -7970.024 0.130 0.056
Chain 1: 1300 -7823.330 0.046 0.033
Chain 1: 1400 -7731.005 0.044 0.027
Chain 1: 1500 -7622.422 0.038 0.027
Chain 1: 1600 -7964.084 0.034 0.027
Chain 1: 1700 -7572.073 0.026 0.027
Chain 1: 1800 -7694.204 0.027 0.027
Chain 1: 1900 -7661.543 0.022 0.019
Chain 1: 2000 -7780.153 0.023 0.019
Chain 1: 2100 -7655.647 0.022 0.016
Chain 1: 2200 -7836.984 0.021 0.016
Chain 1: 2300 -7616.488 0.022 0.016
Chain 1: 2400 -7745.854 0.023 0.017
Chain 1: 2500 -7703.552 0.022 0.017
Chain 1: 2600 -7635.147 0.019 0.016
Chain 1: 2700 -7640.197 0.014 0.016
Chain 1: 2800 -7592.043 0.013 0.015
Chain 1: 2900 -7493.898 0.013 0.015
Chain 1: 3000 -7617.282 0.014 0.016
Chain 1: 3100 -7603.146 0.012 0.013
Chain 1: 3200 -7802.113 0.012 0.013
Chain 1: 3300 -7526.481 0.013 0.013
Chain 1: 3400 -7746.867 0.014 0.013
Chain 1: 3500 -7510.229 0.017 0.016
Chain 1: 3600 -7576.761 0.017 0.016
Chain 1: 3700 -7525.207 0.018 0.016
Chain 1: 3800 -7524.033 0.017 0.016
Chain 1: 3900 -7491.509 0.016 0.016
Chain 1: 4000 -7486.394 0.014 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003593 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.93 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86281.564 1.000 1.000
Chain 1: 200 -13846.555 3.116 5.231
Chain 1: 300 -10172.022 2.198 1.000
Chain 1: 400 -11205.340 1.671 1.000
Chain 1: 500 -9155.551 1.382 0.361
Chain 1: 600 -8639.129 1.161 0.361
Chain 1: 700 -8974.011 1.001 0.224
Chain 1: 800 -9446.377 0.882 0.224
Chain 1: 900 -8890.480 0.791 0.092
Chain 1: 1000 -8759.802 0.713 0.092
Chain 1: 1100 -8812.920 0.614 0.063 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8603.738 0.093 0.060
Chain 1: 1300 -8843.934 0.060 0.050
Chain 1: 1400 -8870.988 0.051 0.037
Chain 1: 1500 -8713.045 0.030 0.027
Chain 1: 1600 -8828.290 0.026 0.024
Chain 1: 1700 -8901.598 0.023 0.018
Chain 1: 1800 -8475.055 0.023 0.018
Chain 1: 1900 -8577.684 0.018 0.015
Chain 1: 2000 -8552.551 0.017 0.013
Chain 1: 2100 -8679.672 0.017 0.015
Chain 1: 2200 -8478.624 0.017 0.015
Chain 1: 2300 -8572.987 0.016 0.013
Chain 1: 2400 -8640.923 0.016 0.013
Chain 1: 2500 -8587.136 0.015 0.012
Chain 1: 2600 -8589.572 0.014 0.011
Chain 1: 2700 -8505.762 0.014 0.011
Chain 1: 2800 -8464.327 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003837 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 38.37 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8390926.955 1.000 1.000
Chain 1: 200 -1585438.787 2.646 4.292
Chain 1: 300 -892089.251 2.023 1.000
Chain 1: 400 -457956.603 1.754 1.000
Chain 1: 500 -358236.588 1.459 0.948
Chain 1: 600 -233303.327 1.305 0.948
Chain 1: 700 -119592.573 1.255 0.948
Chain 1: 800 -86767.217 1.145 0.948
Chain 1: 900 -67129.384 1.050 0.777
Chain 1: 1000 -51936.402 0.975 0.777
Chain 1: 1100 -39411.205 0.906 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38592.669 0.479 0.378
Chain 1: 1300 -26548.716 0.447 0.378
Chain 1: 1400 -26269.136 0.353 0.318
Chain 1: 1500 -22854.534 0.340 0.318
Chain 1: 1600 -22070.399 0.290 0.293
Chain 1: 1700 -20944.283 0.201 0.293
Chain 1: 1800 -20888.567 0.163 0.149
Chain 1: 1900 -21214.856 0.135 0.054
Chain 1: 2000 -19725.353 0.114 0.054
Chain 1: 2100 -19964.045 0.083 0.036
Chain 1: 2200 -20190.396 0.082 0.036
Chain 1: 2300 -19807.644 0.039 0.019
Chain 1: 2400 -19579.662 0.039 0.019
Chain 1: 2500 -19381.468 0.025 0.015
Chain 1: 2600 -19011.745 0.023 0.015
Chain 1: 2700 -18968.727 0.018 0.012
Chain 1: 2800 -18685.364 0.019 0.015
Chain 1: 2900 -18966.723 0.019 0.015
Chain 1: 3000 -18953.054 0.012 0.012
Chain 1: 3100 -19037.994 0.011 0.012
Chain 1: 3200 -18728.615 0.011 0.015
Chain 1: 3300 -18933.385 0.011 0.012
Chain 1: 3400 -18408.081 0.012 0.015
Chain 1: 3500 -19020.211 0.014 0.015
Chain 1: 3600 -18326.639 0.016 0.015
Chain 1: 3700 -18713.589 0.018 0.017
Chain 1: 3800 -17672.786 0.023 0.021
Chain 1: 3900 -17668.884 0.021 0.021
Chain 1: 4000 -17786.247 0.022 0.021
Chain 1: 4100 -17699.908 0.022 0.021
Chain 1: 4200 -17516.086 0.021 0.021
Chain 1: 4300 -17654.578 0.021 0.021
Chain 1: 4400 -17611.325 0.018 0.010
Chain 1: 4500 -17513.820 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001304 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.04 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12593.846 1.000 1.000
Chain 1: 200 -9477.453 0.664 1.000
Chain 1: 300 -7997.924 0.505 0.329
Chain 1: 400 -8261.341 0.386 0.329
Chain 1: 500 -8149.732 0.312 0.185
Chain 1: 600 -7988.270 0.263 0.185
Chain 1: 700 -7875.707 0.228 0.032
Chain 1: 800 -7880.187 0.199 0.032
Chain 1: 900 -7857.831 0.177 0.020
Chain 1: 1000 -7941.608 0.161 0.020
Chain 1: 1100 -7975.593 0.061 0.014
Chain 1: 1200 -7927.001 0.029 0.014
Chain 1: 1300 -7851.534 0.011 0.011
Chain 1: 1400 -7873.007 0.008 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001513 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.13 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58478.297 1.000 1.000
Chain 1: 200 -17956.358 1.628 2.257
Chain 1: 300 -8788.448 1.433 1.043
Chain 1: 400 -8173.201 1.094 1.043
Chain 1: 500 -8866.831 0.891 1.000
Chain 1: 600 -8565.007 0.748 1.000
Chain 1: 700 -7777.135 0.656 0.101
Chain 1: 800 -8227.482 0.581 0.101
Chain 1: 900 -7993.348 0.519 0.078
Chain 1: 1000 -7889.206 0.469 0.078
Chain 1: 1100 -7827.695 0.370 0.075
Chain 1: 1200 -7876.372 0.144 0.055
Chain 1: 1300 -7804.246 0.041 0.035
Chain 1: 1400 -7841.396 0.034 0.029
Chain 1: 1500 -7572.457 0.030 0.029
Chain 1: 1600 -7737.298 0.028 0.021
Chain 1: 1700 -7553.995 0.021 0.021
Chain 1: 1800 -7588.033 0.016 0.013
Chain 1: 1900 -7604.048 0.013 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002969 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.69 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87393.964 1.000 1.000
Chain 1: 200 -13704.386 3.189 5.377
Chain 1: 300 -9975.693 2.250 1.000
Chain 1: 400 -11295.091 1.717 1.000
Chain 1: 500 -8899.933 1.427 0.374
Chain 1: 600 -8765.300 1.192 0.374
Chain 1: 700 -8547.563 1.025 0.269
Chain 1: 800 -8294.126 0.901 0.269
Chain 1: 900 -8297.709 0.801 0.117
Chain 1: 1000 -8598.453 0.724 0.117
Chain 1: 1100 -8750.720 0.626 0.035 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8339.777 0.093 0.035
Chain 1: 1300 -8654.941 0.060 0.035
Chain 1: 1400 -8617.682 0.048 0.031
Chain 1: 1500 -8492.019 0.023 0.025
Chain 1: 1600 -8595.345 0.023 0.025
Chain 1: 1700 -8660.357 0.021 0.017
Chain 1: 1800 -8222.515 0.023 0.017
Chain 1: 1900 -8327.886 0.024 0.017
Chain 1: 2000 -8305.464 0.021 0.015
Chain 1: 2100 -8448.236 0.021 0.015
Chain 1: 2200 -8234.455 0.019 0.015
Chain 1: 2300 -8393.018 0.017 0.015
Chain 1: 2400 -8229.873 0.018 0.017
Chain 1: 2500 -8302.380 0.018 0.017
Chain 1: 2600 -8214.053 0.018 0.017
Chain 1: 2700 -8247.904 0.017 0.017
Chain 1: 2800 -8207.322 0.013 0.013
Chain 1: 2900 -8301.664 0.012 0.011
Chain 1: 3000 -8137.842 0.014 0.017
Chain 1: 3100 -8290.417 0.014 0.018
Chain 1: 3200 -8161.746 0.013 0.016
Chain 1: 3300 -8172.449 0.012 0.011
Chain 1: 3400 -8338.743 0.012 0.011
Chain 1: 3500 -8349.043 0.011 0.011
Chain 1: 3600 -8118.072 0.013 0.016
Chain 1: 3700 -8265.229 0.014 0.018
Chain 1: 3800 -8124.088 0.015 0.018
Chain 1: 3900 -8058.170 0.015 0.018
Chain 1: 4000 -8137.111 0.014 0.017
Chain 1: 4100 -8129.670 0.012 0.016
Chain 1: 4200 -8114.534 0.011 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003504 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.04 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8437331.031 1.000 1.000
Chain 1: 200 -1589953.629 2.653 4.307
Chain 1: 300 -891061.827 2.030 1.000
Chain 1: 400 -457568.989 1.760 1.000
Chain 1: 500 -357326.601 1.464 0.947
Chain 1: 600 -232253.025 1.310 0.947
Chain 1: 700 -118927.811 1.259 0.947
Chain 1: 800 -86248.679 1.149 0.947
Chain 1: 900 -66696.400 1.054 0.784
Chain 1: 1000 -51585.972 0.978 0.784
Chain 1: 1100 -39142.350 0.909 0.539 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38332.483 0.481 0.379
Chain 1: 1300 -26366.349 0.448 0.379
Chain 1: 1400 -26094.134 0.354 0.318
Chain 1: 1500 -22701.295 0.341 0.318
Chain 1: 1600 -21924.277 0.291 0.293
Chain 1: 1700 -20807.027 0.201 0.293
Chain 1: 1800 -20753.548 0.163 0.149
Chain 1: 1900 -21080.099 0.135 0.054
Chain 1: 2000 -19595.041 0.114 0.054
Chain 1: 2100 -19833.360 0.083 0.035
Chain 1: 2200 -20059.301 0.082 0.035
Chain 1: 2300 -19676.838 0.039 0.019
Chain 1: 2400 -19448.858 0.039 0.019
Chain 1: 2500 -19250.432 0.025 0.015
Chain 1: 2600 -18880.637 0.023 0.015
Chain 1: 2700 -18837.587 0.018 0.012
Chain 1: 2800 -18554.088 0.019 0.015
Chain 1: 2900 -18835.405 0.019 0.015
Chain 1: 3000 -18821.649 0.012 0.012
Chain 1: 3100 -18906.708 0.011 0.012
Chain 1: 3200 -18597.181 0.012 0.015
Chain 1: 3300 -18802.052 0.011 0.012
Chain 1: 3400 -18276.484 0.012 0.015
Chain 1: 3500 -18888.965 0.015 0.015
Chain 1: 3600 -18194.781 0.016 0.015
Chain 1: 3700 -18582.160 0.018 0.017
Chain 1: 3800 -17540.458 0.023 0.021
Chain 1: 3900 -17536.485 0.021 0.021
Chain 1: 4000 -17653.866 0.022 0.021
Chain 1: 4100 -17567.549 0.022 0.021
Chain 1: 4200 -17383.455 0.021 0.021
Chain 1: 4300 -17522.140 0.021 0.021
Chain 1: 4400 -17478.718 0.018 0.011
Chain 1: 4500 -17381.126 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001333 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.33 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49182.866 1.000 1.000
Chain 1: 200 -18594.363 1.323 1.645
Chain 1: 300 -14914.771 0.964 1.000
Chain 1: 400 -20740.377 0.793 1.000
Chain 1: 500 -16320.783 0.689 0.281
Chain 1: 600 -14199.635 0.599 0.281
Chain 1: 700 -18286.517 0.545 0.271
Chain 1: 800 -15811.838 0.497 0.271
Chain 1: 900 -14284.921 0.453 0.247
Chain 1: 1000 -13268.663 0.416 0.247
Chain 1: 1100 -13695.198 0.319 0.223
Chain 1: 1200 -11842.391 0.170 0.157
Chain 1: 1300 -14081.876 0.161 0.157
Chain 1: 1400 -10656.493 0.165 0.157
Chain 1: 1500 -12691.906 0.154 0.157
Chain 1: 1600 -12201.432 0.143 0.157
Chain 1: 1700 -19332.772 0.158 0.157
Chain 1: 1800 -13223.678 0.188 0.159
Chain 1: 1900 -10936.038 0.199 0.160
Chain 1: 2000 -17813.522 0.229 0.209
Chain 1: 2100 -10005.179 0.304 0.321
Chain 1: 2200 -10471.939 0.293 0.321
Chain 1: 2300 -9492.851 0.288 0.321
Chain 1: 2400 -9532.978 0.256 0.209
Chain 1: 2500 -9842.067 0.243 0.209
Chain 1: 2600 -10353.598 0.244 0.209
Chain 1: 2700 -11857.174 0.220 0.127
Chain 1: 2800 -10501.503 0.186 0.127
Chain 1: 2900 -9458.330 0.177 0.110
Chain 1: 3000 -9338.415 0.139 0.103
Chain 1: 3100 -11667.982 0.081 0.103
Chain 1: 3200 -11667.261 0.077 0.103
Chain 1: 3300 -11545.586 0.067 0.049
Chain 1: 3400 -17042.180 0.099 0.110
Chain 1: 3500 -9147.216 0.182 0.127
Chain 1: 3600 -10798.045 0.193 0.129
Chain 1: 3700 -10513.424 0.183 0.129
Chain 1: 3800 -10123.311 0.174 0.110
Chain 1: 3900 -13912.743 0.190 0.153
Chain 1: 4000 -14871.596 0.195 0.153
Chain 1: 4100 -9191.200 0.237 0.153
Chain 1: 4200 -8918.626 0.240 0.153
Chain 1: 4300 -10918.597 0.257 0.183
Chain 1: 4400 -9155.682 0.244 0.183
Chain 1: 4500 -9020.818 0.159 0.153
Chain 1: 4600 -11564.363 0.166 0.183
Chain 1: 4700 -9324.784 0.187 0.193
Chain 1: 4800 -8713.504 0.191 0.193
Chain 1: 4900 -9873.056 0.175 0.183
Chain 1: 5000 -12258.587 0.188 0.193
Chain 1: 5100 -12516.109 0.128 0.183
Chain 1: 5200 -16257.920 0.148 0.193
Chain 1: 5300 -10849.990 0.180 0.195
Chain 1: 5400 -9223.945 0.178 0.195
Chain 1: 5500 -14569.186 0.213 0.220
Chain 1: 5600 -8922.698 0.255 0.230
Chain 1: 5700 -14691.949 0.270 0.230
Chain 1: 5800 -11085.560 0.296 0.325
Chain 1: 5900 -8624.803 0.312 0.325
Chain 1: 6000 -9103.641 0.298 0.325
Chain 1: 6100 -9834.845 0.303 0.325
Chain 1: 6200 -13257.166 0.306 0.325
Chain 1: 6300 -9811.765 0.292 0.325
Chain 1: 6400 -14681.478 0.307 0.332
Chain 1: 6500 -9641.010 0.323 0.332
Chain 1: 6600 -9031.399 0.266 0.325
Chain 1: 6700 -10869.910 0.244 0.285
Chain 1: 6800 -8595.798 0.238 0.265
Chain 1: 6900 -12700.050 0.242 0.265
Chain 1: 7000 -11596.181 0.246 0.265
Chain 1: 7100 -11735.676 0.240 0.265
Chain 1: 7200 -8880.480 0.246 0.322
Chain 1: 7300 -9166.760 0.214 0.265
Chain 1: 7400 -8558.622 0.188 0.169
Chain 1: 7500 -10858.761 0.157 0.169
Chain 1: 7600 -8975.694 0.171 0.210
Chain 1: 7700 -10645.661 0.170 0.210
Chain 1: 7800 -10730.001 0.144 0.157
Chain 1: 7900 -9467.128 0.125 0.133
Chain 1: 8000 -11693.838 0.135 0.157
Chain 1: 8100 -11067.413 0.139 0.157
Chain 1: 8200 -8836.256 0.132 0.157
Chain 1: 8300 -11627.638 0.153 0.190
Chain 1: 8400 -8739.896 0.179 0.210
Chain 1: 8500 -9803.877 0.169 0.190
Chain 1: 8600 -12481.131 0.169 0.190
Chain 1: 8700 -8359.386 0.203 0.215
Chain 1: 8800 -8788.564 0.207 0.215
Chain 1: 8900 -9780.069 0.204 0.215
Chain 1: 9000 -10110.695 0.188 0.215
Chain 1: 9100 -9380.265 0.190 0.215
Chain 1: 9200 -9136.515 0.167 0.109
Chain 1: 9300 -8745.820 0.148 0.101
Chain 1: 9400 -10336.083 0.130 0.101
Chain 1: 9500 -9339.748 0.130 0.101
Chain 1: 9600 -10360.412 0.118 0.099
Chain 1: 9700 -12885.813 0.089 0.099
Chain 1: 9800 -11478.770 0.096 0.101
Chain 1: 9900 -8948.013 0.114 0.107
Chain 1: 10000 -11285.542 0.132 0.123
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001631 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 16.31 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57346.559 1.000 1.000
Chain 1: 200 -17863.558 1.605 2.210
Chain 1: 300 -8936.071 1.403 1.000
Chain 1: 400 -8250.145 1.073 1.000
Chain 1: 500 -8815.705 0.871 0.999
Chain 1: 600 -8663.818 0.729 0.999
Chain 1: 700 -8029.347 0.636 0.083
Chain 1: 800 -8309.408 0.561 0.083
Chain 1: 900 -8180.177 0.500 0.079
Chain 1: 1000 -7876.227 0.454 0.079
Chain 1: 1100 -7623.759 0.357 0.064
Chain 1: 1200 -7775.843 0.138 0.039
Chain 1: 1300 -7950.912 0.041 0.034
Chain 1: 1400 -7921.108 0.033 0.033
Chain 1: 1500 -7652.001 0.030 0.033
Chain 1: 1600 -7872.012 0.031 0.033
Chain 1: 1700 -7542.590 0.027 0.033
Chain 1: 1800 -7674.985 0.026 0.028
Chain 1: 1900 -7782.586 0.025 0.028
Chain 1: 2000 -7749.557 0.022 0.022
Chain 1: 2100 -7659.127 0.020 0.020
Chain 1: 2200 -7817.721 0.020 0.020
Chain 1: 2300 -7604.572 0.021 0.020
Chain 1: 2400 -7637.830 0.021 0.020
Chain 1: 2500 -7669.651 0.018 0.017
Chain 1: 2600 -7581.159 0.016 0.014
Chain 1: 2700 -7581.179 0.012 0.012
Chain 1: 2800 -7573.277 0.010 0.012 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003548 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.48 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86272.602 1.000 1.000
Chain 1: 200 -13867.197 3.111 5.221
Chain 1: 300 -10178.236 2.195 1.000
Chain 1: 400 -11347.843 1.672 1.000
Chain 1: 500 -9172.233 1.385 0.362
Chain 1: 600 -8621.442 1.165 0.362
Chain 1: 700 -8725.959 1.000 0.237
Chain 1: 800 -9039.161 0.879 0.237
Chain 1: 900 -8908.021 0.783 0.103
Chain 1: 1000 -8893.850 0.705 0.103
Chain 1: 1100 -8862.313 0.605 0.064 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8463.343 0.088 0.047
Chain 1: 1300 -8851.483 0.056 0.044
Chain 1: 1400 -8800.781 0.046 0.035
Chain 1: 1500 -8721.309 0.024 0.015
Chain 1: 1600 -8824.542 0.018 0.012
Chain 1: 1700 -8891.931 0.018 0.012
Chain 1: 1800 -8460.299 0.020 0.012
Chain 1: 1900 -8564.392 0.019 0.012
Chain 1: 2000 -8539.767 0.019 0.012
Chain 1: 2100 -8676.651 0.021 0.012
Chain 1: 2200 -8469.739 0.018 0.012
Chain 1: 2300 -8568.853 0.015 0.012
Chain 1: 2400 -8631.553 0.015 0.012
Chain 1: 2500 -8572.037 0.015 0.012
Chain 1: 2600 -8577.872 0.014 0.012
Chain 1: 2700 -8492.561 0.014 0.012
Chain 1: 2800 -8448.927 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003381 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.81 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8399802.184 1.000 1.000
Chain 1: 200 -1585404.572 2.649 4.298
Chain 1: 300 -892353.254 2.025 1.000
Chain 1: 400 -458687.206 1.755 1.000
Chain 1: 500 -359019.663 1.460 0.945
Chain 1: 600 -233730.109 1.306 0.945
Chain 1: 700 -119767.813 1.255 0.945
Chain 1: 800 -86933.936 1.145 0.945
Chain 1: 900 -67244.954 1.051 0.777
Chain 1: 1000 -52026.687 0.975 0.777
Chain 1: 1100 -39484.377 0.907 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38660.831 0.479 0.378
Chain 1: 1300 -26592.122 0.447 0.378
Chain 1: 1400 -26309.698 0.353 0.318
Chain 1: 1500 -22890.445 0.340 0.318
Chain 1: 1600 -22105.422 0.290 0.293
Chain 1: 1700 -20975.845 0.201 0.293
Chain 1: 1800 -20919.472 0.163 0.149
Chain 1: 1900 -21245.850 0.135 0.054
Chain 1: 2000 -19754.888 0.114 0.054
Chain 1: 2100 -19993.316 0.083 0.036
Chain 1: 2200 -20220.275 0.082 0.036
Chain 1: 2300 -19836.989 0.039 0.019
Chain 1: 2400 -19608.980 0.039 0.019
Chain 1: 2500 -19411.095 0.025 0.015
Chain 1: 2600 -19040.892 0.023 0.015
Chain 1: 2700 -18997.773 0.018 0.012
Chain 1: 2800 -18714.550 0.019 0.015
Chain 1: 2900 -18995.955 0.019 0.015
Chain 1: 3000 -18982.086 0.012 0.012
Chain 1: 3100 -19067.125 0.011 0.012
Chain 1: 3200 -18757.613 0.011 0.015
Chain 1: 3300 -18962.499 0.011 0.012
Chain 1: 3400 -18437.099 0.012 0.015
Chain 1: 3500 -19049.517 0.014 0.015
Chain 1: 3600 -18355.539 0.016 0.015
Chain 1: 3700 -18742.818 0.018 0.017
Chain 1: 3800 -17701.540 0.023 0.021
Chain 1: 3900 -17697.692 0.021 0.021
Chain 1: 4000 -17814.963 0.022 0.021
Chain 1: 4100 -17728.685 0.022 0.021
Chain 1: 4200 -17544.731 0.021 0.021
Chain 1: 4300 -17683.251 0.021 0.021
Chain 1: 4400 -17639.894 0.018 0.010
Chain 1: 4500 -17542.429 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001519 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.19 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48161.376 1.000 1.000
Chain 1: 200 -14353.416 1.678 2.355
Chain 1: 300 -19660.549 1.208 1.000
Chain 1: 400 -18589.176 0.921 1.000
Chain 1: 500 -14492.179 0.793 0.283
Chain 1: 600 -18958.302 0.700 0.283
Chain 1: 700 -14431.645 0.645 0.283
Chain 1: 800 -10801.126 0.606 0.314
Chain 1: 900 -13694.159 0.562 0.283
Chain 1: 1000 -11640.094 0.524 0.283
Chain 1: 1100 -10541.943 0.434 0.270
Chain 1: 1200 -20199.873 0.247 0.270
Chain 1: 1300 -10839.611 0.306 0.283
Chain 1: 1400 -10697.888 0.301 0.283
Chain 1: 1500 -12231.086 0.286 0.236
Chain 1: 1600 -13705.912 0.273 0.211
Chain 1: 1700 -16387.018 0.258 0.176
Chain 1: 1800 -9891.202 0.290 0.176
Chain 1: 1900 -13474.294 0.295 0.176
Chain 1: 2000 -16064.426 0.294 0.164
Chain 1: 2100 -14540.954 0.294 0.164
Chain 1: 2200 -8904.099 0.310 0.164
Chain 1: 2300 -16369.131 0.269 0.164
Chain 1: 2400 -9147.716 0.346 0.266
Chain 1: 2500 -9313.828 0.336 0.266
Chain 1: 2600 -11455.034 0.344 0.266
Chain 1: 2700 -9145.225 0.352 0.266
Chain 1: 2800 -8722.922 0.292 0.253
Chain 1: 2900 -14392.274 0.304 0.253
Chain 1: 3000 -8469.999 0.358 0.394
Chain 1: 3100 -8988.188 0.354 0.394
Chain 1: 3200 -12405.803 0.318 0.275
Chain 1: 3300 -10875.354 0.286 0.253
Chain 1: 3400 -9863.247 0.218 0.187
Chain 1: 3500 -8866.626 0.227 0.187
Chain 1: 3600 -9495.369 0.215 0.141
Chain 1: 3700 -9109.485 0.194 0.112
Chain 1: 3800 -8933.272 0.191 0.112
Chain 1: 3900 -8476.688 0.157 0.103
Chain 1: 4000 -10944.654 0.110 0.103
Chain 1: 4100 -11790.464 0.111 0.103
Chain 1: 4200 -14143.206 0.100 0.103
Chain 1: 4300 -8773.908 0.147 0.103
Chain 1: 4400 -9084.304 0.140 0.072
Chain 1: 4500 -8693.718 0.134 0.066
Chain 1: 4600 -8772.956 0.128 0.054
Chain 1: 4700 -10854.834 0.143 0.072
Chain 1: 4800 -13698.307 0.162 0.166
Chain 1: 4900 -13053.324 0.161 0.166
Chain 1: 5000 -10297.324 0.165 0.166
Chain 1: 5100 -8436.696 0.180 0.192
Chain 1: 5200 -11986.037 0.193 0.208
Chain 1: 5300 -14134.261 0.147 0.192
Chain 1: 5400 -10356.562 0.180 0.208
Chain 1: 5500 -8979.244 0.191 0.208
Chain 1: 5600 -11777.072 0.214 0.221
Chain 1: 5700 -12561.347 0.201 0.221
Chain 1: 5800 -12979.954 0.184 0.221
Chain 1: 5900 -8516.240 0.231 0.238
Chain 1: 6000 -10244.200 0.221 0.221
Chain 1: 6100 -8313.509 0.222 0.232
Chain 1: 6200 -10156.287 0.211 0.181
Chain 1: 6300 -11712.916 0.209 0.181
Chain 1: 6400 -13102.354 0.183 0.169
Chain 1: 6500 -9979.549 0.199 0.181
Chain 1: 6600 -8439.370 0.194 0.181
Chain 1: 6700 -10728.075 0.209 0.182
Chain 1: 6800 -8128.490 0.237 0.213
Chain 1: 6900 -9767.933 0.202 0.182
Chain 1: 7000 -7957.129 0.208 0.213
Chain 1: 7100 -8178.350 0.187 0.182
Chain 1: 7200 -11178.076 0.196 0.213
Chain 1: 7300 -8383.841 0.216 0.228
Chain 1: 7400 -10972.624 0.229 0.236
Chain 1: 7500 -8283.898 0.230 0.236
Chain 1: 7600 -8246.214 0.212 0.236
Chain 1: 7700 -7966.198 0.194 0.236
Chain 1: 7800 -9542.243 0.179 0.228
Chain 1: 7900 -8074.804 0.180 0.228
Chain 1: 8000 -8853.880 0.166 0.182
Chain 1: 8100 -8227.951 0.171 0.182
Chain 1: 8200 -10233.025 0.164 0.182
Chain 1: 8300 -8068.888 0.158 0.182
Chain 1: 8400 -8555.911 0.140 0.165
Chain 1: 8500 -8099.814 0.113 0.088
Chain 1: 8600 -9579.833 0.128 0.154
Chain 1: 8700 -8446.474 0.138 0.154
Chain 1: 8800 -8051.947 0.126 0.134
Chain 1: 8900 -8129.411 0.109 0.088
Chain 1: 9000 -9569.391 0.115 0.134
Chain 1: 9100 -8355.148 0.122 0.145
Chain 1: 9200 -8322.135 0.103 0.134
Chain 1: 9300 -8027.985 0.080 0.057
Chain 1: 9400 -11325.994 0.103 0.134
Chain 1: 9500 -8005.075 0.139 0.145
Chain 1: 9600 -8070.008 0.124 0.134
Chain 1: 9700 -10364.438 0.133 0.145
Chain 1: 9800 -7967.881 0.158 0.150
Chain 1: 9900 -9684.896 0.175 0.177
Chain 1: 10000 -8003.808 0.181 0.210
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001431 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.31 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -45614.000 1.000 1.000
Chain 1: 200 -14997.516 1.521 2.041
Chain 1: 300 -8479.982 1.270 1.000
Chain 1: 400 -8311.906 0.958 1.000
Chain 1: 500 -8172.097 0.769 0.769
Chain 1: 600 -7847.358 0.648 0.769
Chain 1: 700 -7830.991 0.556 0.041
Chain 1: 800 -8192.712 0.492 0.044
Chain 1: 900 -7914.605 0.441 0.041
Chain 1: 1000 -7934.997 0.397 0.041
Chain 1: 1100 -7641.316 0.301 0.038
Chain 1: 1200 -7762.016 0.099 0.035
Chain 1: 1300 -7703.090 0.022 0.020
Chain 1: 1400 -7849.849 0.022 0.019
Chain 1: 1500 -7632.399 0.023 0.028
Chain 1: 1600 -7545.778 0.020 0.019
Chain 1: 1700 -7526.699 0.020 0.019
Chain 1: 1800 -7548.825 0.016 0.016
Chain 1: 1900 -7620.359 0.014 0.011
Chain 1: 2000 -7621.417 0.014 0.011
Chain 1: 2100 -7676.357 0.010 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003041 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.41 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85753.946 1.000 1.000
Chain 1: 200 -13003.729 3.297 5.595
Chain 1: 300 -9520.180 2.320 1.000
Chain 1: 400 -10189.304 1.757 1.000
Chain 1: 500 -8415.428 1.447 0.366
Chain 1: 600 -8431.014 1.206 0.366
Chain 1: 700 -8457.436 1.035 0.211
Chain 1: 800 -8564.475 0.907 0.211
Chain 1: 900 -8437.820 0.808 0.066
Chain 1: 1000 -8182.415 0.730 0.066
Chain 1: 1100 -8457.923 0.633 0.033 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8220.583 0.077 0.031
Chain 1: 1300 -8303.217 0.041 0.029
Chain 1: 1400 -8297.414 0.035 0.015
Chain 1: 1500 -8207.785 0.015 0.012
Chain 1: 1600 -8290.930 0.015 0.012
Chain 1: 1700 -8388.823 0.016 0.012
Chain 1: 1800 -8017.954 0.020 0.015
Chain 1: 1900 -8114.375 0.019 0.012
Chain 1: 2000 -8085.738 0.017 0.012
Chain 1: 2100 -8233.429 0.015 0.012
Chain 1: 2200 -8009.489 0.015 0.012
Chain 1: 2300 -8092.227 0.015 0.012
Chain 1: 2400 -8160.511 0.016 0.012
Chain 1: 2500 -8121.883 0.015 0.012
Chain 1: 2600 -8115.190 0.014 0.012
Chain 1: 2700 -8027.499 0.014 0.011
Chain 1: 2800 -8013.051 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003177 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.77 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8378888.681 1.000 1.000
Chain 1: 200 -1581603.105 2.649 4.298
Chain 1: 300 -890628.200 2.025 1.000
Chain 1: 400 -457359.884 1.755 1.000
Chain 1: 500 -357786.555 1.460 0.947
Chain 1: 600 -232835.004 1.306 0.947
Chain 1: 700 -118902.534 1.256 0.947
Chain 1: 800 -86035.216 1.147 0.947
Chain 1: 900 -66344.660 1.053 0.776
Chain 1: 1000 -51100.070 0.977 0.776
Chain 1: 1100 -38546.116 0.910 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37711.249 0.482 0.382
Chain 1: 1300 -25656.747 0.452 0.382
Chain 1: 1400 -25370.173 0.358 0.326
Chain 1: 1500 -21954.535 0.346 0.326
Chain 1: 1600 -21168.489 0.296 0.298
Chain 1: 1700 -20042.246 0.206 0.297
Chain 1: 1800 -19985.680 0.168 0.156
Chain 1: 1900 -20310.848 0.140 0.056
Chain 1: 2000 -18823.517 0.118 0.056
Chain 1: 2100 -19061.890 0.086 0.037
Chain 1: 2200 -19287.619 0.085 0.037
Chain 1: 2300 -18905.675 0.040 0.020
Chain 1: 2400 -18678.081 0.040 0.020
Chain 1: 2500 -18480.071 0.026 0.016
Chain 1: 2600 -18111.297 0.024 0.016
Chain 1: 2700 -18068.557 0.019 0.013
Chain 1: 2800 -17785.816 0.020 0.016
Chain 1: 2900 -18066.649 0.020 0.016
Chain 1: 3000 -18052.956 0.012 0.013
Chain 1: 3100 -18137.759 0.011 0.012
Chain 1: 3200 -17829.110 0.012 0.016
Chain 1: 3300 -18033.301 0.011 0.012
Chain 1: 3400 -17509.377 0.013 0.016
Chain 1: 3500 -18119.546 0.015 0.016
Chain 1: 3600 -17428.500 0.017 0.016
Chain 1: 3700 -17813.603 0.019 0.017
Chain 1: 3800 -16776.826 0.024 0.022
Chain 1: 3900 -16773.058 0.022 0.022
Chain 1: 4000 -16890.359 0.023 0.022
Chain 1: 4100 -16804.268 0.023 0.022
Chain 1: 4200 -16621.295 0.022 0.022
Chain 1: 4300 -16759.158 0.022 0.022
Chain 1: 4400 -16716.619 0.019 0.011
Chain 1: 4500 -16619.252 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001305 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.05 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48901.575 1.000 1.000
Chain 1: 200 -15860.593 1.542 2.083
Chain 1: 300 -15025.839 1.046 1.000
Chain 1: 400 -14433.187 0.795 1.000
Chain 1: 500 -26670.831 0.728 0.459
Chain 1: 600 -12142.439 0.806 1.000
Chain 1: 700 -15600.137 0.722 0.459
Chain 1: 800 -13057.862 0.656 0.459
Chain 1: 900 -11668.179 0.597 0.222
Chain 1: 1000 -13148.657 0.548 0.222
Chain 1: 1100 -13105.504 0.449 0.195
Chain 1: 1200 -11664.841 0.253 0.124
Chain 1: 1300 -12662.453 0.255 0.124
Chain 1: 1400 -15125.141 0.267 0.163
Chain 1: 1500 -17134.021 0.233 0.124
Chain 1: 1600 -10212.796 0.181 0.124
Chain 1: 1700 -11711.316 0.172 0.124
Chain 1: 1800 -15922.013 0.179 0.124
Chain 1: 1900 -18684.965 0.182 0.128
Chain 1: 2000 -12472.351 0.220 0.148
Chain 1: 2100 -10012.602 0.244 0.163
Chain 1: 2200 -10792.766 0.239 0.163
Chain 1: 2300 -10477.960 0.234 0.163
Chain 1: 2400 -9317.764 0.231 0.148
Chain 1: 2500 -9339.705 0.219 0.148
Chain 1: 2600 -9677.111 0.155 0.128
Chain 1: 2700 -9956.888 0.145 0.125
Chain 1: 2800 -10237.412 0.121 0.072
Chain 1: 2900 -9973.934 0.109 0.035
Chain 1: 3000 -9176.698 0.068 0.035
Chain 1: 3100 -9290.492 0.045 0.030
Chain 1: 3200 -15889.731 0.079 0.030
Chain 1: 3300 -10726.173 0.124 0.035
Chain 1: 3400 -9810.792 0.121 0.035
Chain 1: 3500 -9331.668 0.126 0.051
Chain 1: 3600 -9171.703 0.124 0.051
Chain 1: 3700 -9535.565 0.125 0.051
Chain 1: 3800 -10995.107 0.136 0.087
Chain 1: 3900 -10416.407 0.138 0.087
Chain 1: 4000 -9760.565 0.136 0.067
Chain 1: 4100 -8909.335 0.145 0.093
Chain 1: 4200 -16004.869 0.148 0.093
Chain 1: 4300 -9778.571 0.163 0.093
Chain 1: 4400 -8891.332 0.164 0.096
Chain 1: 4500 -10050.942 0.170 0.100
Chain 1: 4600 -9479.966 0.174 0.100
Chain 1: 4700 -9212.443 0.174 0.100
Chain 1: 4800 -8682.625 0.166 0.096
Chain 1: 4900 -15170.397 0.204 0.100
Chain 1: 5000 -10714.467 0.238 0.115
Chain 1: 5100 -9899.077 0.237 0.115
Chain 1: 5200 -9155.054 0.201 0.100
Chain 1: 5300 -10330.630 0.149 0.100
Chain 1: 5400 -8433.113 0.161 0.114
Chain 1: 5500 -11240.802 0.175 0.114
Chain 1: 5600 -11502.064 0.171 0.114
Chain 1: 5700 -10365.768 0.179 0.114
Chain 1: 5800 -10798.440 0.177 0.114
Chain 1: 5900 -8550.544 0.160 0.114
Chain 1: 6000 -11675.296 0.146 0.114
Chain 1: 6100 -12995.335 0.147 0.114
Chain 1: 6200 -8624.240 0.190 0.225
Chain 1: 6300 -8781.366 0.180 0.225
Chain 1: 6400 -10452.905 0.174 0.160
Chain 1: 6500 -9176.563 0.163 0.139
Chain 1: 6600 -8881.222 0.164 0.139
Chain 1: 6700 -8497.823 0.157 0.139
Chain 1: 6800 -11548.633 0.180 0.160
Chain 1: 6900 -12723.216 0.163 0.139
Chain 1: 7000 -12323.381 0.139 0.102
Chain 1: 7100 -8500.455 0.174 0.139
Chain 1: 7200 -8511.691 0.124 0.092
Chain 1: 7300 -10706.599 0.142 0.139
Chain 1: 7400 -9577.238 0.138 0.118
Chain 1: 7500 -11164.484 0.138 0.118
Chain 1: 7600 -9193.022 0.156 0.142
Chain 1: 7700 -9973.942 0.160 0.142
Chain 1: 7800 -11291.042 0.145 0.118
Chain 1: 7900 -9822.013 0.151 0.142
Chain 1: 8000 -8864.834 0.158 0.142
Chain 1: 8100 -8858.377 0.113 0.118
Chain 1: 8200 -9252.484 0.118 0.118
Chain 1: 8300 -8399.437 0.107 0.117
Chain 1: 8400 -12269.601 0.127 0.117
Chain 1: 8500 -8413.372 0.159 0.117
Chain 1: 8600 -11523.847 0.164 0.117
Chain 1: 8700 -8421.001 0.193 0.150
Chain 1: 8800 -8887.388 0.187 0.150
Chain 1: 8900 -9806.072 0.181 0.108
Chain 1: 9000 -8717.710 0.183 0.125
Chain 1: 9100 -8626.501 0.184 0.125
Chain 1: 9200 -8596.280 0.180 0.125
Chain 1: 9300 -8485.773 0.171 0.125
Chain 1: 9400 -11828.779 0.168 0.125
Chain 1: 9500 -8242.901 0.165 0.125
Chain 1: 9600 -10595.139 0.161 0.125
Chain 1: 9700 -9214.131 0.139 0.125
Chain 1: 9800 -9242.449 0.134 0.125
Chain 1: 9900 -8878.269 0.129 0.125
Chain 1: 10000 -8163.260 0.125 0.088
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001437 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.37 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57133.972 1.000 1.000
Chain 1: 200 -17600.706 1.623 2.246
Chain 1: 300 -8829.296 1.413 1.000
Chain 1: 400 -8416.099 1.072 1.000
Chain 1: 500 -8728.645 0.865 0.993
Chain 1: 600 -8528.828 0.725 0.993
Chain 1: 700 -8292.125 0.625 0.049
Chain 1: 800 -8139.845 0.549 0.049
Chain 1: 900 -7982.506 0.491 0.036
Chain 1: 1000 -7735.094 0.445 0.036
Chain 1: 1100 -7687.554 0.345 0.032
Chain 1: 1200 -7727.482 0.121 0.029
Chain 1: 1300 -7800.215 0.023 0.023
Chain 1: 1400 -7849.004 0.019 0.020
Chain 1: 1500 -7588.508 0.018 0.020
Chain 1: 1600 -7750.379 0.018 0.020
Chain 1: 1700 -7562.541 0.018 0.020
Chain 1: 1800 -7693.338 0.018 0.020
Chain 1: 1900 -7566.705 0.017 0.017
Chain 1: 2000 -7598.430 0.014 0.017
Chain 1: 2100 -7620.317 0.014 0.017
Chain 1: 2200 -7743.370 0.015 0.017
Chain 1: 2300 -7629.922 0.016 0.017
Chain 1: 2400 -7674.997 0.016 0.017
Chain 1: 2500 -7592.937 0.013 0.016
Chain 1: 2600 -7552.063 0.012 0.015
Chain 1: 2700 -7538.333 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002565 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.65 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86620.313 1.000 1.000
Chain 1: 200 -13666.911 3.169 5.338
Chain 1: 300 -10009.365 2.234 1.000
Chain 1: 400 -10869.232 1.696 1.000
Chain 1: 500 -9005.271 1.398 0.365
Chain 1: 600 -8513.512 1.175 0.365
Chain 1: 700 -8561.397 1.008 0.207
Chain 1: 800 -8905.292 0.886 0.207
Chain 1: 900 -8805.283 0.789 0.079
Chain 1: 1000 -8670.315 0.712 0.079
Chain 1: 1100 -8783.960 0.613 0.058 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8341.774 0.085 0.053
Chain 1: 1300 -8688.431 0.052 0.040
Chain 1: 1400 -8689.143 0.044 0.039
Chain 1: 1500 -8560.454 0.025 0.016
Chain 1: 1600 -8668.840 0.020 0.015
Chain 1: 1700 -8745.792 0.021 0.015
Chain 1: 1800 -8322.192 0.022 0.015
Chain 1: 1900 -8422.882 0.022 0.015
Chain 1: 2000 -8397.301 0.021 0.013
Chain 1: 2100 -8522.844 0.021 0.015
Chain 1: 2200 -8325.962 0.018 0.015
Chain 1: 2300 -8417.687 0.015 0.013
Chain 1: 2400 -8486.464 0.016 0.013
Chain 1: 2500 -8432.716 0.015 0.012
Chain 1: 2600 -8434.067 0.014 0.011
Chain 1: 2700 -8350.800 0.014 0.011
Chain 1: 2800 -8310.688 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003383 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.83 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8386586.239 1.000 1.000
Chain 1: 200 -1581301.785 2.652 4.304
Chain 1: 300 -890111.381 2.027 1.000
Chain 1: 400 -456925.736 1.757 1.000
Chain 1: 500 -357469.212 1.461 0.948
Chain 1: 600 -232640.875 1.307 0.948
Chain 1: 700 -119210.276 1.256 0.948
Chain 1: 800 -86470.583 1.147 0.948
Chain 1: 900 -66867.079 1.052 0.777
Chain 1: 1000 -51698.764 0.976 0.777
Chain 1: 1100 -39199.934 0.908 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38385.221 0.480 0.379
Chain 1: 1300 -26358.286 0.448 0.379
Chain 1: 1400 -26080.287 0.354 0.319
Chain 1: 1500 -22670.741 0.341 0.319
Chain 1: 1600 -21888.778 0.291 0.293
Chain 1: 1700 -20763.978 0.201 0.293
Chain 1: 1800 -20708.792 0.164 0.150
Chain 1: 1900 -21035.147 0.136 0.054
Chain 1: 2000 -19546.750 0.114 0.054
Chain 1: 2100 -19785.193 0.083 0.036
Chain 1: 2200 -20011.549 0.082 0.036
Chain 1: 2300 -19628.800 0.039 0.019
Chain 1: 2400 -19400.863 0.039 0.019
Chain 1: 2500 -19202.745 0.025 0.016
Chain 1: 2600 -18832.917 0.023 0.016
Chain 1: 2700 -18789.955 0.018 0.012
Chain 1: 2800 -18506.660 0.019 0.015
Chain 1: 2900 -18787.973 0.019 0.015
Chain 1: 3000 -18774.205 0.012 0.012
Chain 1: 3100 -18859.181 0.011 0.012
Chain 1: 3200 -18549.801 0.012 0.015
Chain 1: 3300 -18754.623 0.011 0.012
Chain 1: 3400 -18229.368 0.012 0.015
Chain 1: 3500 -18841.438 0.015 0.015
Chain 1: 3600 -18147.924 0.016 0.015
Chain 1: 3700 -18534.837 0.018 0.017
Chain 1: 3800 -17494.172 0.023 0.021
Chain 1: 3900 -17490.315 0.021 0.021
Chain 1: 4000 -17607.629 0.022 0.021
Chain 1: 4100 -17521.312 0.022 0.021
Chain 1: 4200 -17337.565 0.021 0.021
Chain 1: 4300 -17475.998 0.021 0.021
Chain 1: 4400 -17432.774 0.018 0.011
Chain 1: 4500 -17335.281 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001438 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.38 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -11950.494 1.000 1.000
Chain 1: 200 -8937.848 0.669 1.000
Chain 1: 300 -7817.002 0.493 0.337
Chain 1: 400 -7959.296 0.375 0.337
Chain 1: 500 -7791.500 0.304 0.143
Chain 1: 600 -7720.810 0.255 0.143
Chain 1: 700 -7653.239 0.220 0.022
Chain 1: 800 -7604.281 0.193 0.022
Chain 1: 900 -7765.977 0.174 0.021
Chain 1: 1000 -7681.839 0.158 0.021
Chain 1: 1100 -7744.504 0.058 0.018
Chain 1: 1200 -7680.287 0.026 0.011
Chain 1: 1300 -7629.502 0.012 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001467 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.67 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56545.351 1.000 1.000
Chain 1: 200 -16951.668 1.668 2.336
Chain 1: 300 -8461.577 1.446 1.003
Chain 1: 400 -8657.923 1.090 1.003
Chain 1: 500 -8344.796 0.880 1.000
Chain 1: 600 -8808.572 0.742 1.000
Chain 1: 700 -7815.808 0.654 0.127
Chain 1: 800 -8060.901 0.576 0.127
Chain 1: 900 -7780.877 0.516 0.053
Chain 1: 1000 -7623.866 0.467 0.053
Chain 1: 1100 -7571.373 0.367 0.038
Chain 1: 1200 -7511.312 0.135 0.036
Chain 1: 1300 -7607.397 0.035 0.030
Chain 1: 1400 -7740.305 0.035 0.030
Chain 1: 1500 -7508.928 0.034 0.030
Chain 1: 1600 -7446.790 0.030 0.021
Chain 1: 1700 -7432.450 0.017 0.017
Chain 1: 1800 -7463.000 0.015 0.013
Chain 1: 1900 -7502.013 0.012 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002707 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 27.07 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86406.495 1.000 1.000
Chain 1: 200 -13023.548 3.317 5.635
Chain 1: 300 -9498.276 2.335 1.000
Chain 1: 400 -10363.712 1.772 1.000
Chain 1: 500 -8392.455 1.465 0.371
Chain 1: 600 -8101.839 1.227 0.371
Chain 1: 700 -8154.186 1.052 0.235
Chain 1: 800 -8365.030 0.924 0.235
Chain 1: 900 -8382.954 0.822 0.084
Chain 1: 1000 -8590.596 0.742 0.084
Chain 1: 1100 -8238.851 0.646 0.043 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8052.967 0.085 0.036
Chain 1: 1300 -8102.454 0.048 0.025
Chain 1: 1400 -8095.842 0.040 0.024
Chain 1: 1500 -8126.604 0.017 0.023
Chain 1: 1600 -8132.848 0.014 0.006 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002687 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 26.87 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8414407.473 1.000 1.000
Chain 1: 200 -1586259.035 2.652 4.305
Chain 1: 300 -890240.118 2.029 1.000
Chain 1: 400 -456684.646 1.759 1.000
Chain 1: 500 -356779.429 1.463 0.949
Chain 1: 600 -231817.722 1.309 0.949
Chain 1: 700 -118375.881 1.259 0.949
Chain 1: 800 -85674.630 1.149 0.949
Chain 1: 900 -66083.938 1.055 0.782
Chain 1: 1000 -50924.191 0.979 0.782
Chain 1: 1100 -38448.057 0.911 0.539 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37624.468 0.483 0.382
Chain 1: 1300 -25638.079 0.452 0.382
Chain 1: 1400 -25358.885 0.358 0.324
Chain 1: 1500 -21962.091 0.345 0.324
Chain 1: 1600 -21182.394 0.295 0.298
Chain 1: 1700 -20063.716 0.205 0.296
Chain 1: 1800 -20009.150 0.167 0.155
Chain 1: 1900 -20334.681 0.139 0.056
Chain 1: 2000 -18851.069 0.117 0.056
Chain 1: 2100 -19089.040 0.086 0.037
Chain 1: 2200 -19314.517 0.085 0.037
Chain 1: 2300 -18932.758 0.040 0.020
Chain 1: 2400 -18705.195 0.040 0.020
Chain 1: 2500 -18507.014 0.026 0.016
Chain 1: 2600 -18138.119 0.024 0.016
Chain 1: 2700 -18095.345 0.019 0.012
Chain 1: 2800 -17812.487 0.020 0.016
Chain 1: 2900 -18093.311 0.020 0.016
Chain 1: 3000 -18079.551 0.012 0.012
Chain 1: 3100 -18164.449 0.011 0.012
Chain 1: 3200 -17855.636 0.012 0.016
Chain 1: 3300 -18059.968 0.011 0.012
Chain 1: 3400 -17535.752 0.013 0.016
Chain 1: 3500 -18146.263 0.015 0.016
Chain 1: 3600 -17454.692 0.017 0.016
Chain 1: 3700 -17840.194 0.019 0.017
Chain 1: 3800 -16802.566 0.024 0.022
Chain 1: 3900 -16798.762 0.022 0.022
Chain 1: 4000 -16916.077 0.023 0.022
Chain 1: 4100 -16829.989 0.023 0.022
Chain 1: 4200 -16646.813 0.022 0.022
Chain 1: 4300 -16784.816 0.022 0.022
Chain 1: 4400 -16742.122 0.019 0.011
Chain 1: 4500 -16644.721 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001395 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.95 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48823.206 1.000 1.000
Chain 1: 200 -15621.324 1.563 2.125
Chain 1: 300 -15363.311 1.047 1.000
Chain 1: 400 -11175.012 0.879 1.000
Chain 1: 500 -14820.102 0.753 0.375
Chain 1: 600 -13332.996 0.646 0.375
Chain 1: 700 -14102.481 0.561 0.246
Chain 1: 800 -11100.643 0.525 0.270
Chain 1: 900 -12263.571 0.477 0.246
Chain 1: 1000 -13019.541 0.435 0.246
Chain 1: 1100 -27908.618 0.389 0.246
Chain 1: 1200 -11441.676 0.320 0.246
Chain 1: 1300 -12139.965 0.324 0.246
Chain 1: 1400 -10757.201 0.299 0.129
Chain 1: 1500 -11531.604 0.282 0.112
Chain 1: 1600 -12875.588 0.281 0.104
Chain 1: 1700 -9300.632 0.314 0.129
Chain 1: 1800 -18240.115 0.336 0.129
Chain 1: 1900 -10732.593 0.396 0.384
Chain 1: 2000 -10137.827 0.396 0.384
Chain 1: 2100 -11093.697 0.352 0.129
Chain 1: 2200 -9584.679 0.223 0.129
Chain 1: 2300 -9672.008 0.219 0.129
Chain 1: 2400 -9049.642 0.213 0.104
Chain 1: 2500 -16533.114 0.251 0.157
Chain 1: 2600 -9802.558 0.309 0.384
Chain 1: 2700 -9676.067 0.272 0.157
Chain 1: 2800 -9335.417 0.227 0.086
Chain 1: 2900 -9611.500 0.160 0.069
Chain 1: 3000 -8977.045 0.161 0.071
Chain 1: 3100 -8540.907 0.157 0.069
Chain 1: 3200 -9194.667 0.149 0.069
Chain 1: 3300 -10102.228 0.157 0.071
Chain 1: 3400 -9141.269 0.161 0.071
Chain 1: 3500 -10487.508 0.128 0.071
Chain 1: 3600 -9746.993 0.067 0.071
Chain 1: 3700 -8715.089 0.078 0.076
Chain 1: 3800 -8711.848 0.074 0.076
Chain 1: 3900 -9236.179 0.077 0.076
Chain 1: 4000 -9764.545 0.075 0.076
Chain 1: 4100 -9166.795 0.077 0.076
Chain 1: 4200 -12826.206 0.098 0.090
Chain 1: 4300 -13374.003 0.093 0.076
Chain 1: 4400 -9305.403 0.126 0.076
Chain 1: 4500 -9131.911 0.115 0.065
Chain 1: 4600 -12205.515 0.133 0.065
Chain 1: 4700 -11206.673 0.130 0.065
Chain 1: 4800 -8573.438 0.161 0.089
Chain 1: 4900 -8586.333 0.155 0.089
Chain 1: 5000 -9094.874 0.155 0.089
Chain 1: 5100 -11249.567 0.168 0.192
Chain 1: 5200 -9891.031 0.153 0.137
Chain 1: 5300 -12000.037 0.167 0.176
Chain 1: 5400 -16355.953 0.150 0.176
Chain 1: 5500 -10810.435 0.199 0.192
Chain 1: 5600 -8645.569 0.199 0.192
Chain 1: 5700 -14498.986 0.230 0.250
Chain 1: 5800 -8709.371 0.266 0.250
Chain 1: 5900 -8210.919 0.272 0.250
Chain 1: 6000 -9050.749 0.276 0.250
Chain 1: 6100 -9693.218 0.263 0.250
Chain 1: 6200 -11056.641 0.262 0.250
Chain 1: 6300 -8362.332 0.276 0.266
Chain 1: 6400 -8607.158 0.253 0.250
Chain 1: 6500 -11634.398 0.227 0.250
Chain 1: 6600 -10221.758 0.216 0.138
Chain 1: 6700 -9220.144 0.187 0.123
Chain 1: 6800 -8212.330 0.132 0.123
Chain 1: 6900 -14740.750 0.171 0.123
Chain 1: 7000 -9180.499 0.222 0.138
Chain 1: 7100 -8672.405 0.221 0.138
Chain 1: 7200 -8835.635 0.211 0.138
Chain 1: 7300 -9854.219 0.189 0.123
Chain 1: 7400 -10531.517 0.192 0.123
Chain 1: 7500 -9957.343 0.172 0.109
Chain 1: 7600 -8526.571 0.175 0.109
Chain 1: 7700 -8177.115 0.168 0.103
Chain 1: 7800 -12067.358 0.188 0.103
Chain 1: 7900 -8574.423 0.185 0.103
Chain 1: 8000 -8495.859 0.125 0.064
Chain 1: 8100 -8682.068 0.121 0.064
Chain 1: 8200 -9462.750 0.128 0.083
Chain 1: 8300 -10000.697 0.123 0.064
Chain 1: 8400 -10593.211 0.122 0.058
Chain 1: 8500 -8181.091 0.146 0.083
Chain 1: 8600 -9618.478 0.144 0.083
Chain 1: 8700 -8018.344 0.160 0.149
Chain 1: 8800 -8323.713 0.131 0.083
Chain 1: 8900 -9057.588 0.098 0.081
Chain 1: 9000 -8178.773 0.108 0.083
Chain 1: 9100 -10527.205 0.128 0.107
Chain 1: 9200 -8382.763 0.146 0.149
Chain 1: 9300 -8291.313 0.141 0.149
Chain 1: 9400 -8946.743 0.143 0.149
Chain 1: 9500 -10635.658 0.130 0.149
Chain 1: 9600 -8982.392 0.133 0.159
Chain 1: 9700 -10228.757 0.125 0.122
Chain 1: 9800 -10794.826 0.127 0.122
Chain 1: 9900 -10994.941 0.121 0.122
Chain 1: 10000 -7999.748 0.147 0.159
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001386 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.86 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57052.448 1.000 1.000
Chain 1: 200 -17457.118 1.634 2.268
Chain 1: 300 -8745.975 1.421 1.000
Chain 1: 400 -8378.348 1.077 1.000
Chain 1: 500 -8748.744 0.870 0.996
Chain 1: 600 -8726.874 0.725 0.996
Chain 1: 700 -8007.608 0.635 0.090
Chain 1: 800 -8070.296 0.556 0.090
Chain 1: 900 -7939.752 0.496 0.044
Chain 1: 1000 -7684.149 0.450 0.044
Chain 1: 1100 -7810.005 0.352 0.042
Chain 1: 1200 -7712.171 0.126 0.033
Chain 1: 1300 -7691.051 0.027 0.016
Chain 1: 1400 -7917.858 0.025 0.016
Chain 1: 1500 -7629.106 0.025 0.016
Chain 1: 1600 -7756.836 0.026 0.016
Chain 1: 1700 -7565.274 0.020 0.016
Chain 1: 1800 -7581.905 0.019 0.016
Chain 1: 1900 -7612.514 0.018 0.016
Chain 1: 2000 -7667.433 0.015 0.016
Chain 1: 2100 -7528.354 0.016 0.016
Chain 1: 2200 -7734.798 0.017 0.018
Chain 1: 2300 -7603.078 0.018 0.018
Chain 1: 2400 -7548.914 0.016 0.017
Chain 1: 2500 -7678.060 0.014 0.017
Chain 1: 2600 -7540.553 0.014 0.017
Chain 1: 2700 -7578.785 0.012 0.017
Chain 1: 2800 -7573.418 0.012 0.017
Chain 1: 2900 -7427.599 0.014 0.017
Chain 1: 3000 -7555.556 0.015 0.017
Chain 1: 3100 -7550.140 0.013 0.017
Chain 1: 3200 -7729.759 0.013 0.017
Chain 1: 3300 -7492.742 0.014 0.017
Chain 1: 3400 -7679.170 0.016 0.018
Chain 1: 3500 -7466.987 0.017 0.020
Chain 1: 3600 -7525.817 0.016 0.020
Chain 1: 3700 -7478.271 0.016 0.020
Chain 1: 3800 -7490.942 0.016 0.020
Chain 1: 3900 -7472.231 0.014 0.017
Chain 1: 4000 -7444.470 0.013 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004257 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 42.57 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86560.673 1.000 1.000
Chain 1: 200 -13499.954 3.206 5.412
Chain 1: 300 -9848.174 2.261 1.000
Chain 1: 400 -10719.170 1.716 1.000
Chain 1: 500 -8802.395 1.416 0.371
Chain 1: 600 -8579.594 1.185 0.371
Chain 1: 700 -8501.221 1.017 0.218
Chain 1: 800 -9379.645 0.901 0.218
Chain 1: 900 -8596.830 0.811 0.094
Chain 1: 1000 -8386.507 0.733 0.094
Chain 1: 1100 -8682.667 0.636 0.091 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8254.068 0.100 0.081
Chain 1: 1300 -8511.846 0.066 0.052
Chain 1: 1400 -8491.025 0.058 0.034
Chain 1: 1500 -8385.675 0.038 0.030
Chain 1: 1600 -8492.442 0.036 0.030
Chain 1: 1700 -8567.868 0.036 0.030
Chain 1: 1800 -8142.742 0.032 0.030
Chain 1: 1900 -8244.477 0.024 0.025
Chain 1: 2000 -8219.198 0.022 0.013
Chain 1: 2100 -8345.683 0.020 0.013
Chain 1: 2200 -8145.852 0.017 0.013
Chain 1: 2300 -8239.593 0.016 0.013
Chain 1: 2400 -8307.871 0.016 0.013
Chain 1: 2500 -8254.098 0.015 0.012
Chain 1: 2600 -8256.126 0.014 0.011
Chain 1: 2700 -8172.554 0.014 0.011
Chain 1: 2800 -8131.557 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002686 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 26.86 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8399924.304 1.000 1.000
Chain 1: 200 -1584961.392 2.650 4.300
Chain 1: 300 -890201.014 2.027 1.000
Chain 1: 400 -457030.168 1.757 1.000
Chain 1: 500 -357443.874 1.461 0.948
Chain 1: 600 -232441.665 1.307 0.948
Chain 1: 700 -118963.660 1.257 0.948
Chain 1: 800 -86266.283 1.147 0.948
Chain 1: 900 -66661.739 1.052 0.780
Chain 1: 1000 -51501.520 0.977 0.780
Chain 1: 1100 -39016.798 0.909 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38198.354 0.481 0.379
Chain 1: 1300 -26186.540 0.449 0.379
Chain 1: 1400 -25908.995 0.355 0.320
Chain 1: 1500 -22504.932 0.342 0.320
Chain 1: 1600 -21724.441 0.292 0.294
Chain 1: 1700 -20601.651 0.202 0.294
Chain 1: 1800 -20546.714 0.164 0.151
Chain 1: 1900 -20873.090 0.137 0.054
Chain 1: 2000 -19385.801 0.115 0.054
Chain 1: 2100 -19624.047 0.084 0.036
Chain 1: 2200 -19850.437 0.083 0.036
Chain 1: 2300 -19467.637 0.039 0.020
Chain 1: 2400 -19239.715 0.039 0.020
Chain 1: 2500 -19041.635 0.025 0.016
Chain 1: 2600 -18671.792 0.023 0.016
Chain 1: 2700 -18628.730 0.018 0.012
Chain 1: 2800 -18345.545 0.020 0.015
Chain 1: 2900 -18626.763 0.019 0.015
Chain 1: 3000 -18612.954 0.012 0.012
Chain 1: 3100 -18697.988 0.011 0.012
Chain 1: 3200 -18388.584 0.012 0.015
Chain 1: 3300 -18593.370 0.011 0.012
Chain 1: 3400 -18068.136 0.013 0.015
Chain 1: 3500 -18680.231 0.015 0.015
Chain 1: 3600 -17986.572 0.017 0.015
Chain 1: 3700 -18373.630 0.019 0.017
Chain 1: 3800 -17332.826 0.023 0.021
Chain 1: 3900 -17328.939 0.021 0.021
Chain 1: 4000 -17446.255 0.022 0.021
Chain 1: 4100 -17360.028 0.022 0.021
Chain 1: 4200 -17176.135 0.022 0.021
Chain 1: 4300 -17314.638 0.021 0.021
Chain 1: 4400 -17271.369 0.019 0.011
Chain 1: 4500 -17173.867 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001189 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 11.89 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12724.421 1.000 1.000
Chain 1: 200 -9592.758 0.663 1.000
Chain 1: 300 -8207.269 0.498 0.326
Chain 1: 400 -8426.622 0.380 0.326
Chain 1: 500 -8408.635 0.305 0.169
Chain 1: 600 -8186.084 0.258 0.169
Chain 1: 700 -8147.033 0.222 0.027
Chain 1: 800 -8112.395 0.195 0.027
Chain 1: 900 -8091.751 0.174 0.026
Chain 1: 1000 -8190.966 0.157 0.026
Chain 1: 1100 -8212.382 0.058 0.012
Chain 1: 1200 -8153.306 0.026 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001397 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.97 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -52134.166 1.000 1.000
Chain 1: 200 -16783.354 1.553 2.106
Chain 1: 300 -8957.607 1.327 1.000
Chain 1: 400 -7855.169 1.030 1.000
Chain 1: 500 -8318.815 0.835 0.874
Chain 1: 600 -9113.694 0.711 0.874
Chain 1: 700 -8217.135 0.625 0.140
Chain 1: 800 -8377.413 0.549 0.140
Chain 1: 900 -8094.171 0.492 0.109
Chain 1: 1000 -7899.785 0.445 0.109
Chain 1: 1100 -7798.046 0.346 0.087
Chain 1: 1200 -7713.040 0.137 0.056
Chain 1: 1300 -7745.367 0.050 0.035
Chain 1: 1400 -7716.055 0.036 0.025
Chain 1: 1500 -7605.077 0.032 0.019
Chain 1: 1600 -7766.829 0.026 0.019
Chain 1: 1700 -7685.019 0.016 0.015
Chain 1: 1800 -7713.481 0.014 0.013
Chain 1: 1900 -7752.228 0.011 0.011
Chain 1: 2000 -7686.456 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002686 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 26.86 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86469.018 1.000 1.000
Chain 1: 200 -13892.545 3.112 5.224
Chain 1: 300 -10178.549 2.196 1.000
Chain 1: 400 -11552.243 1.677 1.000
Chain 1: 500 -9079.735 1.396 0.365
Chain 1: 600 -8947.947 1.166 0.365
Chain 1: 700 -8766.793 1.002 0.272
Chain 1: 800 -8452.655 0.882 0.272
Chain 1: 900 -8624.115 0.786 0.119
Chain 1: 1000 -8783.686 0.709 0.119
Chain 1: 1100 -8959.536 0.611 0.037 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8507.080 0.094 0.037
Chain 1: 1300 -8821.207 0.061 0.036
Chain 1: 1400 -8800.362 0.049 0.021
Chain 1: 1500 -8688.383 0.023 0.020
Chain 1: 1600 -8790.279 0.023 0.020
Chain 1: 1700 -8853.531 0.022 0.020
Chain 1: 1800 -8416.693 0.023 0.020
Chain 1: 1900 -8521.396 0.022 0.018
Chain 1: 2000 -8498.019 0.021 0.013
Chain 1: 2100 -8640.122 0.021 0.013
Chain 1: 2200 -8427.294 0.018 0.013
Chain 1: 2300 -8586.694 0.016 0.013
Chain 1: 2400 -8426.170 0.018 0.016
Chain 1: 2500 -8495.753 0.017 0.016
Chain 1: 2600 -8408.246 0.017 0.016
Chain 1: 2700 -8441.207 0.017 0.016
Chain 1: 2800 -8400.799 0.012 0.012
Chain 1: 2900 -8494.909 0.012 0.011
Chain 1: 3000 -8330.277 0.014 0.016
Chain 1: 3100 -8483.756 0.014 0.018
Chain 1: 3200 -8355.339 0.013 0.015
Chain 1: 3300 -8364.832 0.011 0.011
Chain 1: 3400 -8528.269 0.011 0.011
Chain 1: 3500 -8540.128 0.011 0.011
Chain 1: 3600 -8311.683 0.012 0.015
Chain 1: 3700 -8458.675 0.014 0.017
Chain 1: 3800 -8317.885 0.015 0.017
Chain 1: 3900 -8252.083 0.014 0.017
Chain 1: 4000 -8329.956 0.013 0.017
Chain 1: 4100 -8323.471 0.012 0.015
Chain 1: 4200 -8308.031 0.010 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00335 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.5 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8403823.490 1.000 1.000
Chain 1: 200 -1583050.853 2.654 4.309
Chain 1: 300 -891120.119 2.028 1.000
Chain 1: 400 -458237.642 1.757 1.000
Chain 1: 500 -358614.359 1.462 0.945
Chain 1: 600 -233612.072 1.307 0.945
Chain 1: 700 -119744.041 1.256 0.945
Chain 1: 800 -86911.916 1.146 0.945
Chain 1: 900 -67239.187 1.052 0.776
Chain 1: 1000 -52029.732 0.976 0.776
Chain 1: 1100 -39494.362 0.907 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38675.453 0.479 0.378
Chain 1: 1300 -26614.586 0.446 0.378
Chain 1: 1400 -26334.840 0.353 0.317
Chain 1: 1500 -22916.785 0.340 0.317
Chain 1: 1600 -22132.216 0.290 0.293
Chain 1: 1700 -21003.549 0.200 0.292
Chain 1: 1800 -20947.583 0.163 0.149
Chain 1: 1900 -21274.080 0.135 0.054
Chain 1: 2000 -19783.260 0.113 0.054
Chain 1: 2100 -20021.826 0.083 0.035
Chain 1: 2200 -20248.621 0.082 0.035
Chain 1: 2300 -19865.458 0.038 0.019
Chain 1: 2400 -19637.376 0.039 0.019
Chain 1: 2500 -19439.363 0.025 0.015
Chain 1: 2600 -19069.086 0.023 0.015
Chain 1: 2700 -19026.003 0.018 0.012
Chain 1: 2800 -18742.581 0.019 0.015
Chain 1: 2900 -19024.126 0.019 0.015
Chain 1: 3000 -19010.322 0.012 0.012
Chain 1: 3100 -19095.301 0.011 0.012
Chain 1: 3200 -18785.707 0.011 0.015
Chain 1: 3300 -18990.691 0.011 0.012
Chain 1: 3400 -18465.053 0.012 0.015
Chain 1: 3500 -19077.682 0.014 0.015
Chain 1: 3600 -18383.490 0.016 0.015
Chain 1: 3700 -18770.899 0.018 0.016
Chain 1: 3800 -17729.122 0.022 0.021
Chain 1: 3900 -17725.234 0.021 0.021
Chain 1: 4000 -17842.564 0.022 0.021
Chain 1: 4100 -17756.159 0.022 0.021
Chain 1: 4200 -17572.155 0.021 0.021
Chain 1: 4300 -17710.741 0.021 0.021
Chain 1: 4400 -17667.304 0.018 0.010
Chain 1: 4500 -17569.802 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001317 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.17 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12244.797 1.000 1.000
Chain 1: 200 -9140.557 0.670 1.000
Chain 1: 300 -7920.615 0.498 0.340
Chain 1: 400 -7971.406 0.375 0.340
Chain 1: 500 -7938.698 0.301 0.154
Chain 1: 600 -7785.396 0.254 0.154
Chain 1: 700 -7695.119 0.219 0.020
Chain 1: 800 -7703.332 0.192 0.020
Chain 1: 900 -7625.251 0.172 0.012
Chain 1: 1000 -7804.568 0.157 0.020
Chain 1: 1100 -7830.472 0.057 0.012
Chain 1: 1200 -7728.012 0.025 0.012
Chain 1: 1300 -7662.866 0.010 0.010
Chain 1: 1400 -7686.417 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001673 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 16.73 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57891.945 1.000 1.000
Chain 1: 200 -17483.210 1.656 2.311
Chain 1: 300 -8549.503 1.452 1.045
Chain 1: 400 -8082.641 1.103 1.045
Chain 1: 500 -7992.688 0.885 1.000
Chain 1: 600 -8002.948 0.738 1.000
Chain 1: 700 -7774.810 0.637 0.058
Chain 1: 800 -8090.863 0.562 0.058
Chain 1: 900 -7867.270 0.503 0.039
Chain 1: 1000 -7704.353 0.454 0.039
Chain 1: 1100 -7566.065 0.356 0.029
Chain 1: 1200 -7538.601 0.126 0.028
Chain 1: 1300 -7541.869 0.021 0.021
Chain 1: 1400 -7767.685 0.018 0.021
Chain 1: 1500 -7497.136 0.021 0.028
Chain 1: 1600 -7536.757 0.021 0.028
Chain 1: 1700 -7448.482 0.019 0.021
Chain 1: 1800 -7456.012 0.016 0.018
Chain 1: 1900 -7488.079 0.013 0.012
Chain 1: 2000 -7524.552 0.011 0.005 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003098 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.98 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86540.541 1.000 1.000
Chain 1: 200 -13314.093 3.250 5.500
Chain 1: 300 -9686.642 2.291 1.000
Chain 1: 400 -10442.245 1.737 1.000
Chain 1: 500 -8659.812 1.431 0.374
Chain 1: 600 -8118.421 1.203 0.374
Chain 1: 700 -8460.193 1.037 0.206
Chain 1: 800 -9177.294 0.917 0.206
Chain 1: 900 -8533.476 0.824 0.078
Chain 1: 1000 -8263.391 0.745 0.078
Chain 1: 1100 -8461.313 0.647 0.075 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8029.719 0.102 0.072
Chain 1: 1300 -8313.916 0.068 0.067
Chain 1: 1400 -8331.859 0.061 0.054
Chain 1: 1500 -8266.278 0.041 0.040
Chain 1: 1600 -8364.484 0.036 0.034
Chain 1: 1700 -8434.839 0.033 0.033
Chain 1: 1800 -8024.240 0.030 0.033
Chain 1: 1900 -8120.068 0.024 0.023
Chain 1: 2000 -8093.318 0.021 0.012
Chain 1: 2100 -8216.054 0.020 0.012
Chain 1: 2200 -8036.100 0.017 0.012
Chain 1: 2300 -8115.210 0.014 0.012
Chain 1: 2400 -8184.847 0.015 0.012
Chain 1: 2500 -8130.203 0.015 0.012
Chain 1: 2600 -8129.585 0.014 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003632 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 36.32 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8425787.739 1.000 1.000
Chain 1: 200 -1588482.199 2.652 4.304
Chain 1: 300 -890249.372 2.030 1.000
Chain 1: 400 -456913.408 1.759 1.000
Chain 1: 500 -356980.596 1.463 0.948
Chain 1: 600 -232180.280 1.309 0.948
Chain 1: 700 -118726.304 1.259 0.948
Chain 1: 800 -85988.634 1.149 0.948
Chain 1: 900 -66403.430 1.054 0.784
Chain 1: 1000 -51249.075 0.978 0.784
Chain 1: 1100 -38774.165 0.910 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37956.679 0.482 0.381
Chain 1: 1300 -25966.447 0.450 0.381
Chain 1: 1400 -25690.696 0.356 0.322
Chain 1: 1500 -22290.898 0.343 0.322
Chain 1: 1600 -21511.164 0.293 0.296
Chain 1: 1700 -20391.312 0.203 0.295
Chain 1: 1800 -20336.885 0.165 0.153
Chain 1: 1900 -20663.049 0.137 0.055
Chain 1: 2000 -19177.359 0.116 0.055
Chain 1: 2100 -19415.793 0.085 0.036
Chain 1: 2200 -19641.570 0.084 0.036
Chain 1: 2300 -19259.328 0.039 0.020
Chain 1: 2400 -19031.455 0.040 0.020
Chain 1: 2500 -18833.165 0.025 0.016
Chain 1: 2600 -18463.798 0.024 0.016
Chain 1: 2700 -18420.840 0.018 0.012
Chain 1: 2800 -18137.567 0.020 0.016
Chain 1: 2900 -18418.708 0.020 0.015
Chain 1: 3000 -18405.050 0.012 0.012
Chain 1: 3100 -18490.010 0.011 0.012
Chain 1: 3200 -18180.805 0.012 0.015
Chain 1: 3300 -18385.411 0.011 0.012
Chain 1: 3400 -17860.420 0.013 0.015
Chain 1: 3500 -18472.077 0.015 0.016
Chain 1: 3600 -17778.975 0.017 0.016
Chain 1: 3700 -18165.579 0.019 0.017
Chain 1: 3800 -17125.575 0.023 0.021
Chain 1: 3900 -17121.645 0.022 0.021
Chain 1: 4000 -17239.024 0.022 0.021
Chain 1: 4100 -17152.783 0.022 0.021
Chain 1: 4200 -16969.055 0.022 0.021
Chain 1: 4300 -17107.485 0.021 0.021
Chain 1: 4400 -17064.377 0.019 0.011
Chain 1: 4500 -16966.834 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001306 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.06 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49722.904 1.000 1.000
Chain 1: 200 -23826.535 1.043 1.087
Chain 1: 300 -16323.578 0.849 1.000
Chain 1: 400 -20242.992 0.685 1.000
Chain 1: 500 -13071.771 0.658 0.549
Chain 1: 600 -18000.524 0.594 0.549
Chain 1: 700 -15733.630 0.530 0.460
Chain 1: 800 -16530.125 0.469 0.460
Chain 1: 900 -13089.359 0.446 0.274
Chain 1: 1000 -11172.938 0.419 0.274
Chain 1: 1100 -15403.345 0.346 0.274
Chain 1: 1200 -18839.818 0.256 0.263
Chain 1: 1300 -11298.287 0.277 0.263
Chain 1: 1400 -12528.257 0.267 0.263
Chain 1: 1500 -16622.839 0.237 0.246
Chain 1: 1600 -12263.332 0.245 0.246
Chain 1: 1700 -13816.214 0.242 0.246
Chain 1: 1800 -15953.363 0.251 0.246
Chain 1: 1900 -10421.048 0.277 0.246
Chain 1: 2000 -18016.362 0.302 0.275
Chain 1: 2100 -10875.089 0.341 0.355
Chain 1: 2200 -13119.717 0.339 0.355
Chain 1: 2300 -10857.861 0.293 0.246
Chain 1: 2400 -12015.577 0.293 0.246
Chain 1: 2500 -9908.881 0.290 0.213
Chain 1: 2600 -12090.675 0.272 0.208
Chain 1: 2700 -12049.877 0.262 0.208
Chain 1: 2800 -10143.080 0.267 0.208
Chain 1: 2900 -10557.043 0.218 0.188
Chain 1: 3000 -10007.257 0.181 0.180
Chain 1: 3100 -12969.290 0.138 0.180
Chain 1: 3200 -10638.561 0.143 0.188
Chain 1: 3300 -15563.954 0.154 0.188
Chain 1: 3400 -13231.597 0.162 0.188
Chain 1: 3500 -9966.308 0.173 0.188
Chain 1: 3600 -9714.183 0.158 0.188
Chain 1: 3700 -18455.374 0.205 0.219
Chain 1: 3800 -12223.174 0.237 0.228
Chain 1: 3900 -10944.658 0.245 0.228
Chain 1: 4000 -14939.681 0.266 0.267
Chain 1: 4100 -9825.102 0.295 0.316
Chain 1: 4200 -9717.502 0.275 0.316
Chain 1: 4300 -13113.195 0.269 0.267
Chain 1: 4400 -14684.219 0.262 0.267
Chain 1: 4500 -14654.973 0.229 0.259
Chain 1: 4600 -9273.479 0.285 0.267
Chain 1: 4700 -9360.076 0.238 0.259
Chain 1: 4800 -13432.674 0.218 0.259
Chain 1: 4900 -11880.287 0.219 0.259
Chain 1: 5000 -11131.855 0.199 0.131
Chain 1: 5100 -11090.861 0.147 0.107
Chain 1: 5200 -9741.058 0.160 0.131
Chain 1: 5300 -10100.706 0.138 0.107
Chain 1: 5400 -9134.242 0.138 0.106
Chain 1: 5500 -11497.153 0.158 0.131
Chain 1: 5600 -9229.965 0.125 0.131
Chain 1: 5700 -13114.694 0.153 0.139
Chain 1: 5800 -9874.931 0.156 0.139
Chain 1: 5900 -11758.325 0.159 0.160
Chain 1: 6000 -12559.025 0.158 0.160
Chain 1: 6100 -9678.473 0.188 0.206
Chain 1: 6200 -10502.211 0.182 0.206
Chain 1: 6300 -9275.583 0.191 0.206
Chain 1: 6400 -15725.535 0.222 0.246
Chain 1: 6500 -10480.047 0.251 0.296
Chain 1: 6600 -9416.415 0.238 0.296
Chain 1: 6700 -9022.689 0.213 0.160
Chain 1: 6800 -10490.681 0.194 0.140
Chain 1: 6900 -9152.016 0.193 0.140
Chain 1: 7000 -11993.461 0.210 0.146
Chain 1: 7100 -13799.986 0.193 0.140
Chain 1: 7200 -9138.824 0.236 0.146
Chain 1: 7300 -11903.244 0.246 0.232
Chain 1: 7400 -13667.410 0.218 0.146
Chain 1: 7500 -9061.490 0.219 0.146
Chain 1: 7600 -10312.971 0.220 0.146
Chain 1: 7700 -9242.455 0.227 0.146
Chain 1: 7800 -12619.795 0.240 0.232
Chain 1: 7900 -8890.160 0.267 0.237
Chain 1: 8000 -8828.271 0.244 0.232
Chain 1: 8100 -9044.299 0.233 0.232
Chain 1: 8200 -10160.484 0.193 0.129
Chain 1: 8300 -9652.264 0.176 0.121
Chain 1: 8400 -12572.371 0.186 0.121
Chain 1: 8500 -8890.811 0.176 0.121
Chain 1: 8600 -10152.018 0.177 0.124
Chain 1: 8700 -11500.269 0.177 0.124
Chain 1: 8800 -9868.382 0.167 0.124
Chain 1: 8900 -11621.747 0.140 0.124
Chain 1: 9000 -9481.686 0.162 0.151
Chain 1: 9100 -9003.613 0.165 0.151
Chain 1: 9200 -9262.196 0.156 0.151
Chain 1: 9300 -8920.867 0.155 0.151
Chain 1: 9400 -12493.406 0.160 0.151
Chain 1: 9500 -9280.037 0.153 0.151
Chain 1: 9600 -10788.098 0.155 0.151
Chain 1: 9700 -11974.064 0.153 0.151
Chain 1: 9800 -8808.403 0.173 0.151
Chain 1: 9900 -11824.980 0.183 0.226
Chain 1: 10000 -11522.245 0.163 0.140
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001436 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.36 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -59016.669 1.000 1.000
Chain 1: 200 -18620.728 1.585 2.169
Chain 1: 300 -9094.445 1.406 1.047
Chain 1: 400 -8109.192 1.085 1.047
Chain 1: 500 -8218.040 0.870 1.000
Chain 1: 600 -8750.752 0.735 1.000
Chain 1: 700 -8552.390 0.634 0.121
Chain 1: 800 -8293.106 0.558 0.121
Chain 1: 900 -7930.630 0.501 0.061
Chain 1: 1000 -8216.661 0.455 0.061
Chain 1: 1100 -7708.064 0.361 0.061
Chain 1: 1200 -7765.132 0.145 0.046
Chain 1: 1300 -7886.363 0.042 0.035
Chain 1: 1400 -7680.361 0.032 0.031
Chain 1: 1500 -7585.053 0.032 0.031
Chain 1: 1600 -7703.585 0.028 0.027
Chain 1: 1700 -7528.929 0.028 0.027
Chain 1: 1800 -7594.801 0.026 0.023
Chain 1: 1900 -7529.832 0.022 0.015
Chain 1: 2000 -7736.338 0.021 0.015
Chain 1: 2100 -7556.679 0.017 0.015
Chain 1: 2200 -7836.719 0.020 0.023
Chain 1: 2300 -7630.374 0.021 0.024
Chain 1: 2400 -7765.946 0.020 0.023
Chain 1: 2500 -7548.910 0.022 0.024
Chain 1: 2600 -7570.497 0.020 0.024
Chain 1: 2700 -7460.309 0.019 0.024
Chain 1: 2800 -7682.577 0.021 0.027
Chain 1: 2900 -7453.285 0.024 0.027
Chain 1: 3000 -7574.942 0.023 0.027
Chain 1: 3100 -7556.096 0.020 0.027
Chain 1: 3200 -7762.028 0.020 0.027
Chain 1: 3300 -7420.767 0.021 0.027
Chain 1: 3400 -7546.094 0.021 0.027
Chain 1: 3500 -7472.965 0.019 0.017
Chain 1: 3600 -7488.963 0.019 0.017
Chain 1: 3700 -7413.299 0.019 0.017
Chain 1: 3800 -7457.755 0.017 0.016
Chain 1: 3900 -7453.022 0.014 0.010
Chain 1: 4000 -7409.834 0.013 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003122 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.22 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86632.582 1.000 1.000
Chain 1: 200 -14331.837 3.022 5.045
Chain 1: 300 -10588.872 2.133 1.000
Chain 1: 400 -12076.678 1.630 1.000
Chain 1: 500 -9571.631 1.357 0.353
Chain 1: 600 -9065.304 1.140 0.353
Chain 1: 700 -9819.877 0.988 0.262
Chain 1: 800 -8834.050 0.878 0.262
Chain 1: 900 -8981.484 0.783 0.123
Chain 1: 1000 -9295.120 0.708 0.123
Chain 1: 1100 -9367.055 0.609 0.112 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8920.944 0.109 0.077
Chain 1: 1300 -9251.152 0.077 0.056
Chain 1: 1400 -9019.803 0.068 0.050
Chain 1: 1500 -9078.597 0.042 0.036
Chain 1: 1600 -9184.636 0.038 0.034
Chain 1: 1700 -9241.979 0.031 0.026
Chain 1: 1800 -8794.318 0.024 0.026
Chain 1: 1900 -8902.430 0.024 0.026
Chain 1: 2000 -8886.778 0.021 0.012
Chain 1: 2100 -9024.485 0.022 0.015
Chain 1: 2200 -8797.611 0.019 0.015
Chain 1: 2300 -8894.739 0.017 0.012
Chain 1: 2400 -8969.932 0.015 0.012
Chain 1: 2500 -8909.953 0.015 0.012
Chain 1: 2600 -8926.706 0.014 0.011
Chain 1: 2700 -8832.869 0.014 0.011
Chain 1: 2800 -8778.529 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002882 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 28.82 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8379539.335 1.000 1.000
Chain 1: 200 -1584842.538 2.644 4.287
Chain 1: 300 -892972.849 2.021 1.000
Chain 1: 400 -459688.412 1.751 1.000
Chain 1: 500 -360222.131 1.456 0.943
Chain 1: 600 -234949.526 1.302 0.943
Chain 1: 700 -120628.058 1.252 0.943
Chain 1: 800 -87708.084 1.142 0.943
Chain 1: 900 -67949.338 1.048 0.775
Chain 1: 1000 -52673.119 0.972 0.775
Chain 1: 1100 -40075.966 0.903 0.533 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39246.072 0.477 0.375
Chain 1: 1300 -27119.082 0.444 0.375
Chain 1: 1400 -26832.284 0.351 0.314
Chain 1: 1500 -23397.718 0.338 0.314
Chain 1: 1600 -22608.105 0.288 0.291
Chain 1: 1700 -21471.566 0.198 0.290
Chain 1: 1800 -21413.569 0.161 0.147
Chain 1: 1900 -21740.261 0.134 0.053
Chain 1: 2000 -20244.713 0.112 0.053
Chain 1: 2100 -20483.553 0.082 0.035
Chain 1: 2200 -20711.301 0.081 0.035
Chain 1: 2300 -20327.169 0.038 0.019
Chain 1: 2400 -20098.944 0.038 0.019
Chain 1: 2500 -19901.267 0.024 0.015
Chain 1: 2600 -19530.633 0.023 0.015
Chain 1: 2700 -19487.240 0.018 0.012
Chain 1: 2800 -19204.013 0.019 0.015
Chain 1: 2900 -19485.611 0.019 0.014
Chain 1: 3000 -19471.702 0.011 0.012
Chain 1: 3100 -19556.834 0.011 0.011
Chain 1: 3200 -19247.036 0.011 0.014
Chain 1: 3300 -19452.076 0.010 0.011
Chain 1: 3400 -18926.305 0.012 0.014
Chain 1: 3500 -19539.403 0.014 0.015
Chain 1: 3600 -18844.478 0.016 0.015
Chain 1: 3700 -19232.562 0.018 0.016
Chain 1: 3800 -18189.891 0.022 0.020
Chain 1: 3900 -18185.994 0.021 0.020
Chain 1: 4000 -18303.261 0.021 0.020
Chain 1: 4100 -18216.981 0.021 0.020
Chain 1: 4200 -18032.623 0.021 0.020
Chain 1: 4300 -18171.404 0.020 0.020
Chain 1: 4400 -18127.807 0.018 0.010
Chain 1: 4500 -18030.280 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001168 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 11.68 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49411.159 1.000 1.000
Chain 1: 200 -23591.514 1.047 1.094
Chain 1: 300 -15179.538 0.883 1.000
Chain 1: 400 -20970.809 0.731 1.000
Chain 1: 500 -17119.083 0.630 0.554
Chain 1: 600 -42530.733 0.625 0.597
Chain 1: 700 -11346.925 0.928 0.597
Chain 1: 800 -16361.556 0.850 0.597
Chain 1: 900 -12041.133 0.796 0.554
Chain 1: 1000 -14203.648 0.731 0.554
Chain 1: 1100 -11640.290 0.653 0.359 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -11052.903 0.549 0.306 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1300 -11455.416 0.497 0.276
Chain 1: 1400 -11984.374 0.474 0.225
Chain 1: 1500 -12874.004 0.458 0.220
Chain 1: 1600 -12003.484 0.406 0.152
Chain 1: 1700 -10250.836 0.148 0.152
Chain 1: 1800 -10308.818 0.118 0.073
Chain 1: 1900 -11773.316 0.095 0.073
Chain 1: 2000 -12546.834 0.086 0.069
Chain 1: 2100 -9751.435 0.092 0.069
Chain 1: 2200 -11486.203 0.102 0.073
Chain 1: 2300 -10082.689 0.113 0.124
Chain 1: 2400 -15937.706 0.145 0.139
Chain 1: 2500 -10865.949 0.185 0.151
Chain 1: 2600 -11749.088 0.185 0.151
Chain 1: 2700 -10963.488 0.175 0.139
Chain 1: 2800 -11185.065 0.176 0.139
Chain 1: 2900 -9738.014 0.179 0.149
Chain 1: 3000 -9066.618 0.180 0.149
Chain 1: 3100 -10034.320 0.161 0.139
Chain 1: 3200 -17808.779 0.190 0.139
Chain 1: 3300 -9658.177 0.260 0.149
Chain 1: 3400 -8964.973 0.231 0.096
Chain 1: 3500 -10025.876 0.195 0.096
Chain 1: 3600 -9980.006 0.188 0.096
Chain 1: 3700 -10821.877 0.188 0.096
Chain 1: 3800 -9479.816 0.201 0.106
Chain 1: 3900 -15990.162 0.227 0.106
Chain 1: 4000 -9526.821 0.287 0.142
Chain 1: 4100 -12922.683 0.304 0.263
Chain 1: 4200 -10520.078 0.283 0.228
Chain 1: 4300 -11832.256 0.209 0.142
Chain 1: 4400 -8771.436 0.237 0.228
Chain 1: 4500 -8846.221 0.227 0.228
Chain 1: 4600 -8712.830 0.228 0.228
Chain 1: 4700 -8911.530 0.222 0.228
Chain 1: 4800 -9109.202 0.210 0.228
Chain 1: 4900 -9309.785 0.172 0.111
Chain 1: 5000 -14328.711 0.139 0.111
Chain 1: 5100 -9168.543 0.169 0.111
Chain 1: 5200 -15981.718 0.189 0.111
Chain 1: 5300 -11278.171 0.219 0.349
Chain 1: 5400 -15846.296 0.213 0.288
Chain 1: 5500 -10243.966 0.267 0.350
Chain 1: 5600 -12524.485 0.284 0.350
Chain 1: 5700 -9064.216 0.320 0.382
Chain 1: 5800 -10069.022 0.328 0.382
Chain 1: 5900 -17564.965 0.368 0.417
Chain 1: 6000 -11131.732 0.391 0.426
Chain 1: 6100 -9553.865 0.351 0.417
Chain 1: 6200 -8986.472 0.315 0.382
Chain 1: 6300 -15154.429 0.314 0.382
Chain 1: 6400 -11530.402 0.316 0.382
Chain 1: 6500 -10879.084 0.268 0.314
Chain 1: 6600 -10774.616 0.251 0.314
Chain 1: 6700 -11425.726 0.218 0.165
Chain 1: 6800 -8482.581 0.243 0.314
Chain 1: 6900 -11158.889 0.224 0.240
Chain 1: 7000 -9635.128 0.182 0.165
Chain 1: 7100 -8267.061 0.182 0.165
Chain 1: 7200 -10901.318 0.200 0.240
Chain 1: 7300 -11451.568 0.164 0.165
Chain 1: 7400 -10958.081 0.137 0.158
Chain 1: 7500 -9428.020 0.147 0.162
Chain 1: 7600 -8742.689 0.154 0.162
Chain 1: 7700 -8779.969 0.149 0.162
Chain 1: 7800 -11523.512 0.138 0.162
Chain 1: 7900 -8683.023 0.147 0.162
Chain 1: 8000 -8512.047 0.133 0.162
Chain 1: 8100 -9489.892 0.127 0.103
Chain 1: 8200 -9492.390 0.103 0.078
Chain 1: 8300 -8363.157 0.111 0.103
Chain 1: 8400 -8159.001 0.109 0.103
Chain 1: 8500 -11340.072 0.121 0.103
Chain 1: 8600 -8731.236 0.143 0.135
Chain 1: 8700 -8796.976 0.144 0.135
Chain 1: 8800 -12015.767 0.147 0.135
Chain 1: 8900 -10611.366 0.127 0.132
Chain 1: 9000 -9111.279 0.142 0.135
Chain 1: 9100 -9077.856 0.132 0.135
Chain 1: 9200 -8424.768 0.139 0.135
Chain 1: 9300 -8522.469 0.127 0.132
Chain 1: 9400 -8240.526 0.128 0.132
Chain 1: 9500 -11942.311 0.131 0.132
Chain 1: 9600 -8366.371 0.144 0.132
Chain 1: 9700 -9045.024 0.150 0.132
Chain 1: 9800 -8436.605 0.131 0.078
Chain 1: 9900 -9404.163 0.128 0.078
Chain 1: 10000 -8091.188 0.128 0.078
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001423 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.23 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58721.680 1.000 1.000
Chain 1: 200 -18052.459 1.626 2.253
Chain 1: 300 -8822.641 1.433 1.046
Chain 1: 400 -8115.328 1.097 1.046
Chain 1: 500 -9137.196 0.900 1.000
Chain 1: 600 -8088.813 0.771 1.000
Chain 1: 700 -8037.172 0.662 0.130
Chain 1: 800 -7873.521 0.582 0.130
Chain 1: 900 -7988.555 0.519 0.112
Chain 1: 1000 -7948.855 0.467 0.112
Chain 1: 1100 -7675.515 0.371 0.087
Chain 1: 1200 -7753.793 0.147 0.036
Chain 1: 1300 -7710.076 0.043 0.021
Chain 1: 1400 -7894.195 0.036 0.021
Chain 1: 1500 -7586.978 0.029 0.021
Chain 1: 1600 -7732.409 0.018 0.019
Chain 1: 1700 -7585.215 0.019 0.019
Chain 1: 1800 -7643.934 0.018 0.019
Chain 1: 1900 -7589.536 0.017 0.019
Chain 1: 2000 -7658.900 0.018 0.019
Chain 1: 2100 -7562.936 0.015 0.013
Chain 1: 2200 -7717.631 0.016 0.019
Chain 1: 2300 -7544.140 0.018 0.019
Chain 1: 2400 -7513.059 0.016 0.019
Chain 1: 2500 -7465.344 0.013 0.013
Chain 1: 2600 -7513.664 0.012 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00327 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.7 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85811.814 1.000 1.000
Chain 1: 200 -13787.928 3.112 5.224
Chain 1: 300 -10039.645 2.199 1.000
Chain 1: 400 -11749.068 1.686 1.000
Chain 1: 500 -8604.312 1.422 0.373
Chain 1: 600 -8744.716 1.187 0.373
Chain 1: 700 -9145.050 1.024 0.365
Chain 1: 800 -9591.770 0.902 0.365
Chain 1: 900 -8695.232 0.813 0.145
Chain 1: 1000 -8853.239 0.734 0.145
Chain 1: 1100 -8788.614 0.634 0.103 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8338.310 0.117 0.054
Chain 1: 1300 -8692.539 0.084 0.047
Chain 1: 1400 -8608.059 0.070 0.044
Chain 1: 1500 -8515.276 0.035 0.041
Chain 1: 1600 -8626.750 0.035 0.041
Chain 1: 1700 -8687.618 0.031 0.018
Chain 1: 1800 -8236.991 0.032 0.018
Chain 1: 1900 -8346.813 0.023 0.013
Chain 1: 2000 -8336.619 0.021 0.013
Chain 1: 2100 -8511.024 0.022 0.013
Chain 1: 2200 -8242.973 0.020 0.013
Chain 1: 2300 -8426.890 0.018 0.013
Chain 1: 2400 -8243.556 0.020 0.020
Chain 1: 2500 -8320.436 0.020 0.020
Chain 1: 2600 -8352.693 0.019 0.020
Chain 1: 2700 -8272.056 0.019 0.020
Chain 1: 2800 -8223.522 0.014 0.013
Chain 1: 2900 -8332.299 0.014 0.013
Chain 1: 3000 -8243.112 0.015 0.013
Chain 1: 3100 -8207.952 0.013 0.011
Chain 1: 3200 -8182.711 0.010 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003107 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.07 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8397776.825 1.000 1.000
Chain 1: 200 -1583769.402 2.651 4.302
Chain 1: 300 -892424.329 2.026 1.000
Chain 1: 400 -458968.503 1.755 1.000
Chain 1: 500 -359192.865 1.460 0.944
Chain 1: 600 -233987.494 1.306 0.944
Chain 1: 700 -119890.787 1.255 0.944
Chain 1: 800 -86996.483 1.146 0.944
Chain 1: 900 -67279.091 1.051 0.775
Chain 1: 1000 -52041.172 0.975 0.775
Chain 1: 1100 -39476.430 0.907 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38654.840 0.479 0.378
Chain 1: 1300 -26560.553 0.447 0.378
Chain 1: 1400 -26278.224 0.353 0.318
Chain 1: 1500 -22851.228 0.341 0.318
Chain 1: 1600 -22064.524 0.291 0.293
Chain 1: 1700 -20931.507 0.201 0.293
Chain 1: 1800 -20874.610 0.163 0.150
Chain 1: 1900 -21201.350 0.136 0.054
Chain 1: 2000 -19707.847 0.114 0.054
Chain 1: 2100 -19946.493 0.083 0.036
Chain 1: 2200 -20173.890 0.082 0.036
Chain 1: 2300 -19790.153 0.039 0.019
Chain 1: 2400 -19561.955 0.039 0.019
Chain 1: 2500 -19364.111 0.025 0.015
Chain 1: 2600 -18993.472 0.023 0.015
Chain 1: 2700 -18950.247 0.018 0.012
Chain 1: 2800 -18666.814 0.019 0.015
Chain 1: 2900 -18948.497 0.019 0.015
Chain 1: 3000 -18934.622 0.012 0.012
Chain 1: 3100 -19019.663 0.011 0.012
Chain 1: 3200 -18709.905 0.011 0.015
Chain 1: 3300 -18915.009 0.011 0.012
Chain 1: 3400 -18389.118 0.012 0.015
Chain 1: 3500 -19002.238 0.015 0.015
Chain 1: 3600 -18307.421 0.016 0.015
Chain 1: 3700 -18695.323 0.018 0.017
Chain 1: 3800 -17652.662 0.023 0.021
Chain 1: 3900 -17648.787 0.021 0.021
Chain 1: 4000 -17766.084 0.022 0.021
Chain 1: 4100 -17679.679 0.022 0.021
Chain 1: 4200 -17495.455 0.021 0.021
Chain 1: 4300 -17634.169 0.021 0.021
Chain 1: 4400 -17590.571 0.018 0.011
Chain 1: 4500 -17493.074 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001302 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.02 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49650.180 1.000 1.000
Chain 1: 200 -16676.164 1.489 1.977
Chain 1: 300 -21526.600 1.068 1.000
Chain 1: 400 -22932.762 0.816 1.000
Chain 1: 500 -16346.953 0.733 0.403
Chain 1: 600 -13743.566 0.643 0.403
Chain 1: 700 -14973.342 0.563 0.225
Chain 1: 800 -13774.502 0.503 0.225
Chain 1: 900 -14178.002 0.450 0.189
Chain 1: 1000 -13054.966 0.414 0.189
Chain 1: 1100 -13255.109 0.316 0.087
Chain 1: 1200 -11793.329 0.130 0.087
Chain 1: 1300 -13192.554 0.118 0.087
Chain 1: 1400 -10803.158 0.134 0.106
Chain 1: 1500 -11108.877 0.097 0.087
Chain 1: 1600 -13021.986 0.092 0.087
Chain 1: 1700 -15734.201 0.101 0.106
Chain 1: 1800 -11640.620 0.128 0.124
Chain 1: 1900 -10377.866 0.137 0.124
Chain 1: 2000 -10500.289 0.130 0.124
Chain 1: 2100 -10045.413 0.133 0.124
Chain 1: 2200 -10527.777 0.125 0.122
Chain 1: 2300 -15429.243 0.146 0.147
Chain 1: 2400 -9714.865 0.183 0.147
Chain 1: 2500 -11662.673 0.197 0.167
Chain 1: 2600 -10001.778 0.199 0.167
Chain 1: 2700 -11448.583 0.194 0.166
Chain 1: 2800 -20513.497 0.203 0.166
Chain 1: 2900 -10072.523 0.295 0.167
Chain 1: 3000 -15234.584 0.327 0.318
Chain 1: 3100 -10288.860 0.371 0.339
Chain 1: 3200 -9344.757 0.376 0.339
Chain 1: 3300 -9907.536 0.350 0.339
Chain 1: 3400 -14153.010 0.322 0.300
Chain 1: 3500 -16536.988 0.319 0.300
Chain 1: 3600 -14882.404 0.314 0.300
Chain 1: 3700 -10362.755 0.345 0.339
Chain 1: 3800 -9226.925 0.313 0.300
Chain 1: 3900 -9457.936 0.212 0.144
Chain 1: 4000 -9440.682 0.178 0.123
Chain 1: 4100 -9901.545 0.135 0.111
Chain 1: 4200 -9273.742 0.131 0.111
Chain 1: 4300 -9685.505 0.130 0.111
Chain 1: 4400 -9706.936 0.100 0.068
Chain 1: 4500 -9156.098 0.092 0.060
Chain 1: 4600 -15251.486 0.120 0.060
Chain 1: 4700 -9550.467 0.137 0.060
Chain 1: 4800 -9233.682 0.128 0.047
Chain 1: 4900 -17716.839 0.173 0.060
Chain 1: 5000 -9288.802 0.264 0.068
Chain 1: 5100 -9662.654 0.263 0.068
Chain 1: 5200 -9164.945 0.261 0.060
Chain 1: 5300 -13818.602 0.291 0.337
Chain 1: 5400 -10538.021 0.322 0.337
Chain 1: 5500 -9954.686 0.322 0.337
Chain 1: 5600 -12956.285 0.305 0.311
Chain 1: 5700 -9128.886 0.287 0.311
Chain 1: 5800 -9057.657 0.284 0.311
Chain 1: 5900 -14901.581 0.276 0.311
Chain 1: 6000 -9482.430 0.242 0.311
Chain 1: 6100 -9708.116 0.241 0.311
Chain 1: 6200 -8897.824 0.244 0.311
Chain 1: 6300 -15460.855 0.253 0.311
Chain 1: 6400 -9722.568 0.281 0.392
Chain 1: 6500 -10364.499 0.281 0.392
Chain 1: 6600 -9313.516 0.269 0.392
Chain 1: 6700 -8819.941 0.233 0.113
Chain 1: 6800 -9061.447 0.235 0.113
Chain 1: 6900 -10448.821 0.209 0.113
Chain 1: 7000 -9979.438 0.157 0.091
Chain 1: 7100 -9776.201 0.156 0.091
Chain 1: 7200 -8801.184 0.158 0.111
Chain 1: 7300 -12010.899 0.143 0.111
Chain 1: 7400 -12236.762 0.085 0.062
Chain 1: 7500 -11872.171 0.082 0.056
Chain 1: 7600 -8937.870 0.104 0.056
Chain 1: 7700 -11593.397 0.121 0.111
Chain 1: 7800 -13609.529 0.133 0.133
Chain 1: 7900 -9339.242 0.166 0.148
Chain 1: 8000 -12207.131 0.185 0.229
Chain 1: 8100 -9292.079 0.214 0.235
Chain 1: 8200 -8668.004 0.210 0.235
Chain 1: 8300 -8639.520 0.184 0.229
Chain 1: 8400 -12628.050 0.213 0.235
Chain 1: 8500 -8663.624 0.256 0.314
Chain 1: 8600 -8741.390 0.224 0.235
Chain 1: 8700 -11078.152 0.222 0.235
Chain 1: 8800 -9154.329 0.228 0.235
Chain 1: 8900 -10692.536 0.197 0.211
Chain 1: 9000 -10849.954 0.175 0.210
Chain 1: 9100 -9577.815 0.157 0.144
Chain 1: 9200 -8793.434 0.159 0.144
Chain 1: 9300 -11673.346 0.183 0.210
Chain 1: 9400 -8809.262 0.184 0.210
Chain 1: 9500 -9191.340 0.142 0.144
Chain 1: 9600 -9452.430 0.144 0.144
Chain 1: 9700 -8464.955 0.135 0.133
Chain 1: 9800 -12010.803 0.143 0.133
Chain 1: 9900 -10622.562 0.142 0.131
Chain 1: 10000 -8622.437 0.164 0.133
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001392 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.92 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -52016.291 1.000 1.000
Chain 1: 200 -16988.608 1.531 2.062
Chain 1: 300 -9073.532 1.311 1.000
Chain 1: 400 -9662.196 0.999 1.000
Chain 1: 500 -9230.709 0.808 0.872
Chain 1: 600 -9029.545 0.677 0.872
Chain 1: 700 -8598.263 0.588 0.061
Chain 1: 800 -8492.119 0.516 0.061
Chain 1: 900 -8068.319 0.464 0.053
Chain 1: 1000 -7886.722 0.420 0.053
Chain 1: 1100 -7690.722 0.323 0.050
Chain 1: 1200 -7690.538 0.117 0.047
Chain 1: 1300 -7737.583 0.030 0.025
Chain 1: 1400 -7813.896 0.025 0.023
Chain 1: 1500 -7579.479 0.023 0.023
Chain 1: 1600 -7536.402 0.022 0.023
Chain 1: 1700 -7654.045 0.018 0.015
Chain 1: 1800 -7720.380 0.018 0.015
Chain 1: 1900 -7604.838 0.014 0.015
Chain 1: 2000 -7649.344 0.012 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002735 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 27.35 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87817.137 1.000 1.000
Chain 1: 200 -14174.308 3.098 5.196
Chain 1: 300 -10447.043 2.184 1.000
Chain 1: 400 -11801.663 1.667 1.000
Chain 1: 500 -9355.750 1.386 0.357
Chain 1: 600 -9170.827 1.158 0.357
Chain 1: 700 -9805.212 1.002 0.261
Chain 1: 800 -8756.427 0.892 0.261
Chain 1: 900 -8865.708 0.794 0.120
Chain 1: 1000 -9106.142 0.717 0.120
Chain 1: 1100 -9214.935 0.618 0.115 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8811.288 0.103 0.065
Chain 1: 1300 -9120.206 0.071 0.046
Chain 1: 1400 -9002.021 0.061 0.034
Chain 1: 1500 -8955.065 0.035 0.026
Chain 1: 1600 -9063.180 0.034 0.026
Chain 1: 1700 -9127.513 0.029 0.013
Chain 1: 1800 -8689.305 0.022 0.013
Chain 1: 1900 -8793.764 0.022 0.013
Chain 1: 2000 -8771.743 0.019 0.012
Chain 1: 2100 -8748.644 0.018 0.012
Chain 1: 2200 -8713.635 0.014 0.012
Chain 1: 2300 -8848.674 0.012 0.012
Chain 1: 2400 -8692.381 0.013 0.012
Chain 1: 2500 -8763.524 0.013 0.012
Chain 1: 2600 -8677.093 0.013 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002527 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.27 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8414662.096 1.000 1.000
Chain 1: 200 -1585881.872 2.653 4.306
Chain 1: 300 -891431.784 2.028 1.000
Chain 1: 400 -457917.653 1.758 1.000
Chain 1: 500 -358237.993 1.462 0.947
Chain 1: 600 -233208.776 1.308 0.947
Chain 1: 700 -119698.926 1.256 0.947
Chain 1: 800 -86957.644 1.146 0.947
Chain 1: 900 -67352.551 1.051 0.779
Chain 1: 1000 -52191.922 0.975 0.779
Chain 1: 1100 -39701.566 0.907 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38886.615 0.478 0.377
Chain 1: 1300 -26868.704 0.445 0.377
Chain 1: 1400 -26591.949 0.351 0.315
Chain 1: 1500 -23185.232 0.338 0.315
Chain 1: 1600 -22404.033 0.288 0.291
Chain 1: 1700 -21280.473 0.199 0.290
Chain 1: 1800 -21225.510 0.161 0.147
Chain 1: 1900 -21552.138 0.134 0.053
Chain 1: 2000 -20063.749 0.112 0.053
Chain 1: 2100 -20302.182 0.082 0.035
Chain 1: 2200 -20528.695 0.081 0.035
Chain 1: 2300 -20145.726 0.038 0.019
Chain 1: 2400 -19917.695 0.038 0.019
Chain 1: 2500 -19719.474 0.024 0.015
Chain 1: 2600 -19349.442 0.023 0.015
Chain 1: 2700 -19306.335 0.018 0.012
Chain 1: 2800 -19022.940 0.019 0.015
Chain 1: 2900 -19304.318 0.019 0.015
Chain 1: 3000 -19290.545 0.011 0.012
Chain 1: 3100 -19375.587 0.011 0.011
Chain 1: 3200 -19066.011 0.011 0.015
Chain 1: 3300 -19270.930 0.010 0.011
Chain 1: 3400 -18745.323 0.012 0.015
Chain 1: 3500 -19357.936 0.014 0.015
Chain 1: 3600 -18663.643 0.016 0.015
Chain 1: 3700 -19051.152 0.018 0.016
Chain 1: 3800 -18009.294 0.022 0.020
Chain 1: 3900 -18005.359 0.021 0.020
Chain 1: 4000 -18122.701 0.021 0.020
Chain 1: 4100 -18036.387 0.021 0.020
Chain 1: 4200 -17852.277 0.021 0.020
Chain 1: 4300 -17990.958 0.020 0.020
Chain 1: 4400 -17947.510 0.018 0.010
Chain 1: 4500 -17849.954 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001325 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.25 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49413.413 1.000 1.000
Chain 1: 200 -14117.140 1.750 2.500
Chain 1: 300 -15639.746 1.199 1.000
Chain 1: 400 -19925.633 0.953 1.000
Chain 1: 500 -17437.056 0.791 0.215
Chain 1: 600 -26378.606 0.716 0.339
Chain 1: 700 -16652.802 0.697 0.339
Chain 1: 800 -14570.060 0.628 0.339
Chain 1: 900 -17952.708 0.579 0.215
Chain 1: 1000 -21542.746 0.538 0.215
Chain 1: 1100 -10760.536 0.538 0.215 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -11564.926 0.295 0.188
Chain 1: 1300 -12842.283 0.295 0.188
Chain 1: 1400 -18643.041 0.305 0.188
Chain 1: 1500 -11964.883 0.346 0.311
Chain 1: 1600 -17024.298 0.342 0.297
Chain 1: 1700 -14985.868 0.297 0.188
Chain 1: 1800 -10862.984 0.321 0.297
Chain 1: 1900 -11285.371 0.306 0.297
Chain 1: 2000 -15559.619 0.317 0.297
Chain 1: 2100 -11102.279 0.256 0.297
Chain 1: 2200 -10820.324 0.252 0.297
Chain 1: 2300 -9566.385 0.255 0.297
Chain 1: 2400 -10087.260 0.229 0.275
Chain 1: 2500 -10292.106 0.176 0.136
Chain 1: 2600 -11219.141 0.154 0.131
Chain 1: 2700 -16514.165 0.173 0.131
Chain 1: 2800 -9617.284 0.206 0.131
Chain 1: 2900 -16490.253 0.244 0.275
Chain 1: 3000 -9794.902 0.285 0.321
Chain 1: 3100 -10649.910 0.253 0.131
Chain 1: 3200 -9458.360 0.263 0.131
Chain 1: 3300 -10021.894 0.255 0.126
Chain 1: 3400 -9342.554 0.258 0.126
Chain 1: 3500 -12083.426 0.278 0.227
Chain 1: 3600 -9370.243 0.299 0.290
Chain 1: 3700 -10792.792 0.280 0.227
Chain 1: 3800 -9097.477 0.227 0.186
Chain 1: 3900 -15249.939 0.226 0.186
Chain 1: 4000 -15542.318 0.159 0.132
Chain 1: 4100 -9631.476 0.213 0.186
Chain 1: 4200 -12973.368 0.226 0.227
Chain 1: 4300 -9043.118 0.264 0.258
Chain 1: 4400 -13749.636 0.291 0.290
Chain 1: 4500 -9101.723 0.319 0.342
Chain 1: 4600 -9527.080 0.294 0.342
Chain 1: 4700 -8929.158 0.288 0.342
Chain 1: 4800 -8940.518 0.269 0.342
Chain 1: 4900 -9298.233 0.233 0.258
Chain 1: 5000 -14573.270 0.267 0.342
Chain 1: 5100 -9524.839 0.259 0.342
Chain 1: 5200 -15906.972 0.273 0.362
Chain 1: 5300 -8956.389 0.307 0.362
Chain 1: 5400 -16091.640 0.317 0.401
Chain 1: 5500 -8938.364 0.346 0.401
Chain 1: 5600 -9119.124 0.344 0.401
Chain 1: 5700 -9186.877 0.338 0.401
Chain 1: 5800 -8967.788 0.340 0.401
Chain 1: 5900 -9565.438 0.343 0.401
Chain 1: 6000 -9055.641 0.312 0.401
Chain 1: 6100 -9370.161 0.262 0.062
Chain 1: 6200 -9360.497 0.222 0.056
Chain 1: 6300 -9393.925 0.145 0.034
Chain 1: 6400 -9331.498 0.102 0.024
Chain 1: 6500 -9551.286 0.024 0.023
Chain 1: 6600 -9281.018 0.025 0.024
Chain 1: 6700 -10111.984 0.032 0.029
Chain 1: 6800 -9334.029 0.038 0.034
Chain 1: 6900 -12122.237 0.055 0.034
Chain 1: 7000 -8993.894 0.084 0.034
Chain 1: 7100 -9000.095 0.081 0.029
Chain 1: 7200 -8848.124 0.082 0.029
Chain 1: 7300 -9413.207 0.088 0.060
Chain 1: 7400 -14267.550 0.121 0.082
Chain 1: 7500 -9954.481 0.162 0.083
Chain 1: 7600 -10928.893 0.168 0.089
Chain 1: 7700 -8794.986 0.184 0.230
Chain 1: 7800 -13610.487 0.211 0.243
Chain 1: 7900 -9723.551 0.228 0.340
Chain 1: 8000 -9253.464 0.199 0.243
Chain 1: 8100 -10566.614 0.211 0.243
Chain 1: 8200 -10013.009 0.215 0.243
Chain 1: 8300 -8963.546 0.221 0.243
Chain 1: 8400 -8563.289 0.191 0.124
Chain 1: 8500 -8943.922 0.152 0.117
Chain 1: 8600 -10943.614 0.162 0.124
Chain 1: 8700 -8626.706 0.164 0.124
Chain 1: 8800 -8845.469 0.131 0.117
Chain 1: 8900 -9210.701 0.095 0.055
Chain 1: 9000 -10679.707 0.104 0.117
Chain 1: 9100 -8646.459 0.115 0.117
Chain 1: 9200 -12537.190 0.141 0.138
Chain 1: 9300 -8540.787 0.176 0.183
Chain 1: 9400 -9307.079 0.179 0.183
Chain 1: 9500 -8508.669 0.184 0.183
Chain 1: 9600 -11407.070 0.191 0.235
Chain 1: 9700 -11754.000 0.168 0.138
Chain 1: 9800 -10502.369 0.177 0.138
Chain 1: 9900 -11213.225 0.179 0.138
Chain 1: 10000 -8583.513 0.196 0.235
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001471 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.71 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57598.067 1.000 1.000
Chain 1: 200 -17885.067 1.610 2.220
Chain 1: 300 -8963.851 1.405 1.000
Chain 1: 400 -8217.640 1.077 1.000
Chain 1: 500 -8624.907 0.871 0.995
Chain 1: 600 -9069.860 0.734 0.995
Chain 1: 700 -8099.960 0.646 0.120
Chain 1: 800 -8019.482 0.567 0.120
Chain 1: 900 -8118.778 0.505 0.091
Chain 1: 1000 -7797.360 0.459 0.091
Chain 1: 1100 -8267.821 0.364 0.057
Chain 1: 1200 -7655.715 0.150 0.057
Chain 1: 1300 -7884.307 0.054 0.049
Chain 1: 1400 -7890.931 0.045 0.047
Chain 1: 1500 -7604.369 0.044 0.041
Chain 1: 1600 -7794.618 0.041 0.038
Chain 1: 1700 -7542.623 0.033 0.033
Chain 1: 1800 -7694.180 0.034 0.033
Chain 1: 1900 -7669.442 0.033 0.033
Chain 1: 2000 -7710.475 0.029 0.029
Chain 1: 2100 -7612.969 0.025 0.024
Chain 1: 2200 -7778.305 0.019 0.021
Chain 1: 2300 -7623.719 0.018 0.020
Chain 1: 2400 -7598.272 0.018 0.020
Chain 1: 2500 -7691.171 0.016 0.020
Chain 1: 2600 -7586.570 0.015 0.014
Chain 1: 2700 -7498.437 0.012 0.013
Chain 1: 2800 -7679.157 0.013 0.013
Chain 1: 2900 -7436.680 0.016 0.014
Chain 1: 3000 -7594.454 0.017 0.020
Chain 1: 3100 -7578.700 0.016 0.020
Chain 1: 3200 -7793.726 0.017 0.020
Chain 1: 3300 -7504.212 0.019 0.021
Chain 1: 3400 -7746.797 0.021 0.024
Chain 1: 3500 -7493.035 0.024 0.028
Chain 1: 3600 -7558.251 0.023 0.028
Chain 1: 3700 -7509.305 0.023 0.028
Chain 1: 3800 -7508.805 0.020 0.028
Chain 1: 3900 -7469.011 0.017 0.021
Chain 1: 4000 -7460.760 0.016 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002623 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 26.23 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87092.942 1.000 1.000
Chain 1: 200 -13970.200 3.117 5.234
Chain 1: 300 -10293.080 2.197 1.000
Chain 1: 400 -11378.549 1.672 1.000
Chain 1: 500 -9196.947 1.385 0.357
Chain 1: 600 -8771.779 1.162 0.357
Chain 1: 700 -8790.075 0.996 0.237
Chain 1: 800 -9193.787 0.877 0.237
Chain 1: 900 -9026.704 0.782 0.095
Chain 1: 1000 -9032.199 0.704 0.095
Chain 1: 1100 -9112.040 0.605 0.048 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8615.402 0.087 0.048
Chain 1: 1300 -8952.290 0.055 0.044
Chain 1: 1400 -8955.216 0.046 0.038
Chain 1: 1500 -8843.423 0.023 0.019
Chain 1: 1600 -8957.311 0.019 0.013
Chain 1: 1700 -9028.227 0.020 0.013
Chain 1: 1800 -8601.343 0.021 0.013
Chain 1: 1900 -8703.675 0.020 0.013
Chain 1: 2000 -8678.671 0.020 0.013
Chain 1: 2100 -8805.547 0.021 0.013
Chain 1: 2200 -8604.825 0.017 0.013
Chain 1: 2300 -8699.018 0.015 0.013
Chain 1: 2400 -8767.069 0.015 0.013
Chain 1: 2500 -8713.333 0.015 0.012
Chain 1: 2600 -8715.641 0.013 0.011
Chain 1: 2700 -8631.877 0.014 0.011
Chain 1: 2800 -8590.586 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002848 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 28.48 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8405706.148 1.000 1.000
Chain 1: 200 -1583247.829 2.655 4.309
Chain 1: 300 -890790.745 2.029 1.000
Chain 1: 400 -458210.233 1.758 1.000
Chain 1: 500 -358724.380 1.462 0.944
Chain 1: 600 -233592.227 1.307 0.944
Chain 1: 700 -119733.420 1.256 0.944
Chain 1: 800 -86958.381 1.146 0.944
Chain 1: 900 -67290.593 1.052 0.777
Chain 1: 1000 -52084.296 0.976 0.777
Chain 1: 1100 -39557.530 0.907 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38735.045 0.478 0.377
Chain 1: 1300 -26679.897 0.446 0.377
Chain 1: 1400 -26398.443 0.353 0.317
Chain 1: 1500 -22983.518 0.340 0.317
Chain 1: 1600 -22199.994 0.290 0.292
Chain 1: 1700 -21071.868 0.200 0.292
Chain 1: 1800 -21015.763 0.162 0.149
Chain 1: 1900 -21342.146 0.135 0.054
Chain 1: 2000 -19852.317 0.113 0.054
Chain 1: 2100 -20090.540 0.083 0.035
Chain 1: 2200 -20317.426 0.082 0.035
Chain 1: 2300 -19934.215 0.038 0.019
Chain 1: 2400 -19706.282 0.038 0.019
Chain 1: 2500 -19508.445 0.025 0.015
Chain 1: 2600 -19138.270 0.023 0.015
Chain 1: 2700 -19095.141 0.018 0.012
Chain 1: 2800 -18812.039 0.019 0.015
Chain 1: 2900 -19093.370 0.019 0.015
Chain 1: 3000 -19079.456 0.012 0.012
Chain 1: 3100 -19164.510 0.011 0.012
Chain 1: 3200 -18855.016 0.011 0.015
Chain 1: 3300 -19059.897 0.011 0.012
Chain 1: 3400 -18534.610 0.012 0.015
Chain 1: 3500 -19146.825 0.014 0.015
Chain 1: 3600 -18453.074 0.016 0.015
Chain 1: 3700 -18840.240 0.018 0.016
Chain 1: 3800 -17799.290 0.022 0.021
Chain 1: 3900 -17795.457 0.021 0.021
Chain 1: 4000 -17912.719 0.022 0.021
Chain 1: 4100 -17826.493 0.022 0.021
Chain 1: 4200 -17642.593 0.021 0.021
Chain 1: 4300 -17781.058 0.021 0.021
Chain 1: 4400 -17737.770 0.018 0.010
Chain 1: 4500 -17640.323 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001462 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.62 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12584.131 1.000 1.000
Chain 1: 200 -9514.443 0.661 1.000
Chain 1: 300 -8120.575 0.498 0.323
Chain 1: 400 -8307.221 0.379 0.323
Chain 1: 500 -8272.651 0.304 0.172
Chain 1: 600 -8064.080 0.258 0.172
Chain 1: 700 -7953.148 0.223 0.026
Chain 1: 800 -7984.467 0.196 0.026
Chain 1: 900 -8079.241 0.175 0.022
Chain 1: 1000 -8018.578 0.158 0.022
Chain 1: 1100 -8087.094 0.059 0.014
Chain 1: 1200 -7967.984 0.028 0.014
Chain 1: 1300 -7911.495 0.012 0.012
Chain 1: 1400 -7945.522 0.010 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001399 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.99 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56811.887 1.000 1.000
Chain 1: 200 -17680.175 1.607 2.213
Chain 1: 300 -8826.368 1.405 1.003
Chain 1: 400 -8263.313 1.071 1.003
Chain 1: 500 -8883.774 0.871 1.000
Chain 1: 600 -8331.543 0.737 1.000
Chain 1: 700 -8352.309 0.632 0.070
Chain 1: 800 -8044.120 0.558 0.070
Chain 1: 900 -7936.183 0.497 0.068
Chain 1: 1000 -7828.534 0.449 0.068
Chain 1: 1100 -7772.069 0.350 0.066
Chain 1: 1200 -7640.195 0.130 0.038
Chain 1: 1300 -7774.425 0.031 0.017
Chain 1: 1400 -7897.648 0.026 0.017
Chain 1: 1500 -7631.144 0.023 0.017
Chain 1: 1600 -7737.633 0.017 0.016
Chain 1: 1700 -7563.046 0.019 0.017
Chain 1: 1800 -7677.641 0.017 0.016
Chain 1: 1900 -7677.060 0.016 0.016
Chain 1: 2000 -7610.958 0.015 0.016
Chain 1: 2100 -7570.155 0.015 0.016
Chain 1: 2200 -7882.021 0.017 0.016
Chain 1: 2300 -7528.770 0.020 0.016
Chain 1: 2400 -7531.652 0.019 0.015
Chain 1: 2500 -7665.324 0.017 0.015
Chain 1: 2600 -7528.396 0.017 0.017
Chain 1: 2700 -7594.718 0.016 0.015
Chain 1: 2800 -7521.512 0.016 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003095 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.95 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86784.645 1.000 1.000
Chain 1: 200 -13732.873 3.160 5.319
Chain 1: 300 -10023.025 2.230 1.000
Chain 1: 400 -11189.291 1.698 1.000
Chain 1: 500 -9028.987 1.407 0.370
Chain 1: 600 -8634.066 1.180 0.370
Chain 1: 700 -8618.490 1.012 0.239
Chain 1: 800 -8946.698 0.890 0.239
Chain 1: 900 -8776.446 0.793 0.104
Chain 1: 1000 -8583.299 0.716 0.104
Chain 1: 1100 -8801.319 0.618 0.046 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8304.711 0.092 0.046
Chain 1: 1300 -8691.468 0.060 0.044
Chain 1: 1400 -8644.557 0.050 0.037
Chain 1: 1500 -8556.502 0.027 0.025
Chain 1: 1600 -8661.106 0.024 0.023
Chain 1: 1700 -8723.867 0.024 0.023
Chain 1: 1800 -8290.090 0.026 0.023
Chain 1: 1900 -8394.511 0.025 0.023
Chain 1: 2000 -8370.011 0.023 0.012
Chain 1: 2100 -8327.197 0.021 0.012
Chain 1: 2200 -8313.071 0.015 0.010
Chain 1: 2300 -8448.381 0.013 0.010
Chain 1: 2400 -8296.369 0.014 0.012
Chain 1: 2500 -8364.615 0.014 0.012
Chain 1: 2600 -8283.706 0.013 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003363 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.63 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8375073.339 1.000 1.000
Chain 1: 200 -1580302.582 2.650 4.300
Chain 1: 300 -891328.238 2.024 1.000
Chain 1: 400 -458204.262 1.754 1.000
Chain 1: 500 -359016.030 1.459 0.945
Chain 1: 600 -233852.592 1.305 0.945
Chain 1: 700 -119824.777 1.254 0.945
Chain 1: 800 -86928.093 1.145 0.945
Chain 1: 900 -67221.475 1.050 0.773
Chain 1: 1000 -51971.626 0.975 0.773
Chain 1: 1100 -39398.278 0.907 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38572.430 0.479 0.378
Chain 1: 1300 -26476.012 0.447 0.378
Chain 1: 1400 -26191.295 0.354 0.319
Chain 1: 1500 -22763.697 0.341 0.319
Chain 1: 1600 -21975.845 0.291 0.293
Chain 1: 1700 -20843.299 0.201 0.293
Chain 1: 1800 -20786.094 0.164 0.151
Chain 1: 1900 -21112.592 0.136 0.054
Chain 1: 2000 -19619.414 0.114 0.054
Chain 1: 2100 -19858.239 0.084 0.036
Chain 1: 2200 -20085.365 0.083 0.036
Chain 1: 2300 -19701.833 0.039 0.019
Chain 1: 2400 -19473.739 0.039 0.019
Chain 1: 2500 -19275.797 0.025 0.015
Chain 1: 2600 -18905.658 0.023 0.015
Chain 1: 2700 -18862.457 0.018 0.012
Chain 1: 2800 -18579.202 0.019 0.015
Chain 1: 2900 -18860.671 0.019 0.015
Chain 1: 3000 -18846.838 0.012 0.012
Chain 1: 3100 -18931.881 0.011 0.012
Chain 1: 3200 -18622.333 0.012 0.015
Chain 1: 3300 -18827.206 0.011 0.012
Chain 1: 3400 -18301.734 0.012 0.015
Chain 1: 3500 -18914.307 0.015 0.015
Chain 1: 3600 -18220.099 0.016 0.015
Chain 1: 3700 -18607.615 0.018 0.017
Chain 1: 3800 -17565.961 0.023 0.021
Chain 1: 3900 -17562.063 0.021 0.021
Chain 1: 4000 -17679.357 0.022 0.021
Chain 1: 4100 -17593.084 0.022 0.021
Chain 1: 4200 -17408.985 0.021 0.021
Chain 1: 4300 -17547.624 0.021 0.021
Chain 1: 4400 -17504.235 0.018 0.011
Chain 1: 4500 -17406.698 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002455 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 24.55 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49070.529 1.000 1.000
Chain 1: 200 -20833.761 1.178 1.355
Chain 1: 300 -18451.361 0.828 1.000
Chain 1: 400 -22487.573 0.666 1.000
Chain 1: 500 -14299.358 0.647 0.573
Chain 1: 600 -16167.388 0.559 0.573
Chain 1: 700 -13713.344 0.504 0.179
Chain 1: 800 -28498.415 0.506 0.519
Chain 1: 900 -11711.214 0.609 0.519
Chain 1: 1000 -12904.320 0.558 0.519
Chain 1: 1100 -27802.436 0.511 0.519 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -10716.275 0.535 0.519 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1300 -13253.144 0.541 0.519 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1400 -10863.428 0.545 0.519 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1500 -12067.106 0.498 0.220
Chain 1: 1600 -10357.761 0.503 0.220 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1700 -12573.908 0.503 0.220 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1800 -11860.152 0.457 0.191
Chain 1: 1900 -10971.627 0.322 0.176
Chain 1: 2000 -11771.347 0.319 0.176
Chain 1: 2100 -12240.510 0.269 0.165
Chain 1: 2200 -11605.119 0.115 0.100
Chain 1: 2300 -9396.248 0.120 0.100
Chain 1: 2400 -10697.598 0.110 0.100
Chain 1: 2500 -10926.147 0.102 0.081
Chain 1: 2600 -9477.403 0.101 0.081
Chain 1: 2700 -14124.922 0.116 0.081
Chain 1: 2800 -9411.050 0.160 0.122
Chain 1: 2900 -13962.316 0.185 0.153
Chain 1: 3000 -9571.942 0.224 0.235
Chain 1: 3100 -16802.872 0.263 0.326
Chain 1: 3200 -13705.857 0.280 0.326
Chain 1: 3300 -15966.429 0.271 0.326
Chain 1: 3400 -12073.345 0.291 0.326
Chain 1: 3500 -10611.376 0.303 0.326
Chain 1: 3600 -9901.644 0.294 0.326
Chain 1: 3700 -10561.476 0.268 0.322
Chain 1: 3800 -13354.517 0.239 0.226
Chain 1: 3900 -9419.575 0.248 0.226
Chain 1: 4000 -9630.189 0.204 0.209
Chain 1: 4100 -14020.432 0.192 0.209
Chain 1: 4200 -10207.891 0.207 0.209
Chain 1: 4300 -9983.555 0.195 0.209
Chain 1: 4400 -11344.601 0.175 0.138
Chain 1: 4500 -9478.726 0.181 0.197
Chain 1: 4600 -11977.013 0.195 0.209
Chain 1: 4700 -9320.313 0.217 0.209
Chain 1: 4800 -14487.548 0.232 0.285
Chain 1: 4900 -9025.978 0.250 0.285
Chain 1: 5000 -9774.486 0.256 0.285
Chain 1: 5100 -8622.890 0.238 0.209
Chain 1: 5200 -8924.783 0.204 0.197
Chain 1: 5300 -14279.302 0.239 0.209
Chain 1: 5400 -9953.220 0.271 0.285
Chain 1: 5500 -11185.642 0.262 0.285
Chain 1: 5600 -8647.317 0.270 0.294
Chain 1: 5700 -9106.221 0.247 0.294
Chain 1: 5800 -8673.695 0.216 0.134
Chain 1: 5900 -9408.633 0.164 0.110
Chain 1: 6000 -9417.542 0.156 0.110
Chain 1: 6100 -8590.050 0.152 0.096
Chain 1: 6200 -8445.047 0.151 0.096
Chain 1: 6300 -9253.169 0.122 0.087
Chain 1: 6400 -11372.457 0.097 0.087
Chain 1: 6500 -10972.289 0.090 0.078
Chain 1: 6600 -8505.087 0.089 0.078
Chain 1: 6700 -8523.322 0.084 0.078
Chain 1: 6800 -13889.496 0.118 0.087
Chain 1: 6900 -13518.751 0.113 0.087
Chain 1: 7000 -12692.833 0.119 0.087
Chain 1: 7100 -9327.628 0.146 0.087
Chain 1: 7200 -8687.843 0.152 0.087
Chain 1: 7300 -11179.243 0.165 0.186
Chain 1: 7400 -8407.648 0.179 0.223
Chain 1: 7500 -8319.639 0.177 0.223
Chain 1: 7600 -8668.875 0.152 0.074
Chain 1: 7700 -12235.786 0.181 0.223
Chain 1: 7800 -10504.334 0.159 0.165
Chain 1: 7900 -11024.324 0.161 0.165
Chain 1: 8000 -9860.127 0.166 0.165
Chain 1: 8100 -11928.398 0.147 0.165
Chain 1: 8200 -9902.731 0.160 0.173
Chain 1: 8300 -8627.676 0.153 0.165
Chain 1: 8400 -10845.441 0.140 0.165
Chain 1: 8500 -8467.707 0.167 0.173
Chain 1: 8600 -8686.079 0.166 0.173
Chain 1: 8700 -8721.244 0.137 0.165
Chain 1: 8800 -8877.000 0.122 0.148
Chain 1: 8900 -9704.391 0.126 0.148
Chain 1: 9000 -8459.327 0.129 0.148
Chain 1: 9100 -8788.461 0.115 0.147
Chain 1: 9200 -10969.726 0.115 0.147
Chain 1: 9300 -8436.471 0.130 0.147
Chain 1: 9400 -9288.495 0.119 0.092
Chain 1: 9500 -8680.636 0.098 0.085
Chain 1: 9600 -9748.464 0.106 0.092
Chain 1: 9700 -11043.710 0.118 0.110
Chain 1: 9800 -9558.221 0.131 0.117
Chain 1: 9900 -10877.412 0.135 0.121
Chain 1: 10000 -8359.558 0.150 0.121
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001409 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.09 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57492.198 1.000 1.000
Chain 1: 200 -17723.470 1.622 2.244
Chain 1: 300 -8827.805 1.417 1.008
Chain 1: 400 -8172.972 1.083 1.008
Chain 1: 500 -8750.806 0.880 1.000
Chain 1: 600 -8137.178 0.746 1.000
Chain 1: 700 -8306.859 0.642 0.080
Chain 1: 800 -7876.454 0.569 0.080
Chain 1: 900 -7940.731 0.506 0.075
Chain 1: 1000 -7914.330 0.456 0.075
Chain 1: 1100 -7745.881 0.358 0.066
Chain 1: 1200 -7591.034 0.136 0.055
Chain 1: 1300 -7792.164 0.038 0.026
Chain 1: 1400 -7915.696 0.031 0.022
Chain 1: 1500 -7561.638 0.029 0.022
Chain 1: 1600 -7756.903 0.024 0.022
Chain 1: 1700 -7513.520 0.025 0.025
Chain 1: 1800 -7626.149 0.021 0.022
Chain 1: 1900 -7591.613 0.021 0.022
Chain 1: 2000 -7643.298 0.021 0.022
Chain 1: 2100 -7607.000 0.020 0.020
Chain 1: 2200 -7749.672 0.020 0.018
Chain 1: 2300 -7520.097 0.020 0.018
Chain 1: 2400 -7558.013 0.019 0.018
Chain 1: 2500 -7570.811 0.014 0.015
Chain 1: 2600 -7522.043 0.013 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003408 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.08 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86886.899 1.000 1.000
Chain 1: 200 -13704.357 3.170 5.340
Chain 1: 300 -10071.738 2.234 1.000
Chain 1: 400 -11053.807 1.697 1.000
Chain 1: 500 -8970.909 1.404 0.361
Chain 1: 600 -8626.124 1.177 0.361
Chain 1: 700 -8925.891 1.014 0.232
Chain 1: 800 -9472.464 0.894 0.232
Chain 1: 900 -8927.436 0.802 0.089
Chain 1: 1000 -8687.613 0.724 0.089
Chain 1: 1100 -8861.677 0.626 0.061 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8432.251 0.097 0.058
Chain 1: 1300 -8637.708 0.064 0.051
Chain 1: 1400 -8778.964 0.056 0.040
Chain 1: 1500 -8635.185 0.035 0.034
Chain 1: 1600 -8750.959 0.032 0.028
Chain 1: 1700 -8831.154 0.030 0.024
Chain 1: 1800 -8418.997 0.029 0.024
Chain 1: 1900 -8514.864 0.024 0.020
Chain 1: 2000 -8488.147 0.021 0.017
Chain 1: 2100 -8610.682 0.021 0.016
Chain 1: 2200 -8430.757 0.018 0.016
Chain 1: 2300 -8509.913 0.016 0.014
Chain 1: 2400 -8579.583 0.016 0.013
Chain 1: 2500 -8524.983 0.015 0.011
Chain 1: 2600 -8524.343 0.013 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003154 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.54 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8407266.859 1.000 1.000
Chain 1: 200 -1584914.568 2.652 4.305
Chain 1: 300 -890363.473 2.028 1.000
Chain 1: 400 -457382.609 1.758 1.000
Chain 1: 500 -357802.754 1.462 0.947
Chain 1: 600 -232783.763 1.308 0.947
Chain 1: 700 -119227.783 1.257 0.947
Chain 1: 800 -86489.610 1.147 0.947
Chain 1: 900 -66873.019 1.052 0.780
Chain 1: 1000 -51702.380 0.976 0.780
Chain 1: 1100 -39207.578 0.908 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38387.966 0.480 0.379
Chain 1: 1300 -26374.810 0.448 0.379
Chain 1: 1400 -26096.008 0.354 0.319
Chain 1: 1500 -22691.887 0.341 0.319
Chain 1: 1600 -21910.779 0.291 0.293
Chain 1: 1700 -20788.236 0.201 0.293
Chain 1: 1800 -20733.337 0.164 0.150
Chain 1: 1900 -21059.472 0.136 0.054
Chain 1: 2000 -19572.687 0.114 0.054
Chain 1: 2100 -19810.898 0.083 0.036
Chain 1: 2200 -20037.060 0.082 0.036
Chain 1: 2300 -19654.555 0.039 0.019
Chain 1: 2400 -19426.728 0.039 0.019
Chain 1: 2500 -19228.592 0.025 0.015
Chain 1: 2600 -18858.969 0.023 0.015
Chain 1: 2700 -18815.966 0.018 0.012
Chain 1: 2800 -18532.805 0.019 0.015
Chain 1: 2900 -18814.011 0.019 0.015
Chain 1: 3000 -18800.186 0.012 0.012
Chain 1: 3100 -18885.178 0.011 0.012
Chain 1: 3200 -18575.908 0.012 0.015
Chain 1: 3300 -18780.604 0.011 0.012
Chain 1: 3400 -18255.575 0.012 0.015
Chain 1: 3500 -18867.309 0.015 0.015
Chain 1: 3600 -18174.174 0.016 0.015
Chain 1: 3700 -18560.844 0.018 0.017
Chain 1: 3800 -17520.781 0.023 0.021
Chain 1: 3900 -17516.920 0.021 0.021
Chain 1: 4000 -17634.244 0.022 0.021
Chain 1: 4100 -17548.018 0.022 0.021
Chain 1: 4200 -17364.311 0.021 0.021
Chain 1: 4300 -17502.686 0.021 0.021
Chain 1: 4400 -17459.554 0.018 0.011
Chain 1: 4500 -17362.091 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001398 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.98 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48581.918 1.000 1.000
Chain 1: 200 -28590.818 0.850 1.000
Chain 1: 300 -18600.000 0.745 0.699
Chain 1: 400 -13946.429 0.643 0.699
Chain 1: 500 -17364.861 0.553 0.537
Chain 1: 600 -11203.296 0.553 0.550
Chain 1: 700 -12397.778 0.488 0.537
Chain 1: 800 -13595.856 0.438 0.537
Chain 1: 900 -11202.815 0.413 0.334
Chain 1: 1000 -10265.315 0.381 0.334
Chain 1: 1100 -9734.677 0.286 0.214
Chain 1: 1200 -10429.039 0.223 0.197
Chain 1: 1300 -18305.556 0.212 0.197
Chain 1: 1400 -10156.728 0.259 0.197
Chain 1: 1500 -9132.323 0.251 0.112
Chain 1: 1600 -11176.606 0.214 0.112
Chain 1: 1700 -9531.013 0.221 0.173
Chain 1: 1800 -11493.409 0.230 0.173
Chain 1: 1900 -10575.103 0.217 0.171
Chain 1: 2000 -19659.056 0.254 0.173
Chain 1: 2100 -9771.396 0.350 0.183
Chain 1: 2200 -9790.584 0.343 0.183
Chain 1: 2300 -9256.242 0.306 0.173
Chain 1: 2400 -9011.685 0.229 0.171
Chain 1: 2500 -10542.401 0.232 0.171
Chain 1: 2600 -9509.922 0.224 0.145
Chain 1: 2700 -9033.333 0.212 0.109
Chain 1: 2800 -10030.173 0.205 0.099
Chain 1: 2900 -9367.882 0.204 0.099
Chain 1: 3000 -8540.972 0.167 0.097
Chain 1: 3100 -17443.907 0.117 0.097
Chain 1: 3200 -18114.314 0.121 0.097
Chain 1: 3300 -13506.155 0.149 0.099
Chain 1: 3400 -9180.588 0.193 0.109
Chain 1: 3500 -17299.406 0.226 0.109
Chain 1: 3600 -8547.721 0.317 0.341
Chain 1: 3700 -10345.300 0.329 0.341
Chain 1: 3800 -9834.908 0.325 0.341
Chain 1: 3900 -8559.040 0.332 0.341
Chain 1: 4000 -8651.891 0.324 0.341
Chain 1: 4100 -8394.874 0.276 0.174
Chain 1: 4200 -10148.912 0.289 0.174
Chain 1: 4300 -12965.128 0.277 0.174
Chain 1: 4400 -9584.531 0.265 0.174
Chain 1: 4500 -11826.092 0.237 0.174
Chain 1: 4600 -13525.310 0.147 0.173
Chain 1: 4700 -8321.433 0.193 0.173
Chain 1: 4800 -8490.462 0.189 0.173
Chain 1: 4900 -13706.928 0.213 0.190
Chain 1: 5000 -8734.839 0.268 0.217
Chain 1: 5100 -8390.578 0.269 0.217
Chain 1: 5200 -8952.087 0.258 0.217
Chain 1: 5300 -9069.755 0.238 0.190
Chain 1: 5400 -8332.279 0.212 0.126
Chain 1: 5500 -11072.891 0.217 0.126
Chain 1: 5600 -11075.754 0.205 0.089
Chain 1: 5700 -8433.866 0.174 0.089
Chain 1: 5800 -8522.256 0.173 0.089
Chain 1: 5900 -9989.954 0.149 0.089
Chain 1: 6000 -10554.960 0.098 0.063
Chain 1: 6100 -9049.636 0.110 0.089
Chain 1: 6200 -8185.469 0.115 0.106
Chain 1: 6300 -11766.626 0.144 0.147
Chain 1: 6400 -9335.587 0.161 0.166
Chain 1: 6500 -12539.165 0.162 0.166
Chain 1: 6600 -12802.506 0.164 0.166
Chain 1: 6700 -10314.771 0.156 0.166
Chain 1: 6800 -7920.482 0.186 0.241
Chain 1: 6900 -11237.582 0.200 0.255
Chain 1: 7000 -8208.979 0.232 0.260
Chain 1: 7100 -8017.860 0.218 0.260
Chain 1: 7200 -10405.807 0.230 0.260
Chain 1: 7300 -7897.811 0.231 0.260
Chain 1: 7400 -8502.653 0.213 0.255
Chain 1: 7500 -10678.571 0.207 0.241
Chain 1: 7600 -10203.400 0.210 0.241
Chain 1: 7700 -8365.973 0.208 0.229
Chain 1: 7800 -10269.973 0.196 0.220
Chain 1: 7900 -8528.753 0.187 0.204
Chain 1: 8000 -10352.024 0.168 0.204
Chain 1: 8100 -7909.283 0.196 0.204
Chain 1: 8200 -8098.845 0.176 0.204
Chain 1: 8300 -7967.087 0.146 0.185
Chain 1: 8400 -8016.234 0.139 0.185
Chain 1: 8500 -8915.252 0.129 0.176
Chain 1: 8600 -9455.389 0.130 0.176
Chain 1: 8700 -9621.212 0.110 0.101
Chain 1: 8800 -9822.564 0.093 0.057
Chain 1: 8900 -12004.711 0.091 0.057
Chain 1: 9000 -9571.310 0.099 0.057
Chain 1: 9100 -8854.538 0.076 0.057
Chain 1: 9200 -8395.047 0.079 0.057
Chain 1: 9300 -7906.729 0.084 0.062
Chain 1: 9400 -8414.494 0.089 0.062
Chain 1: 9500 -8292.938 0.080 0.060
Chain 1: 9600 -7961.536 0.079 0.060
Chain 1: 9700 -8220.341 0.080 0.060
Chain 1: 9800 -8523.773 0.082 0.060
Chain 1: 9900 -8737.608 0.066 0.055
Chain 1: 10000 -8826.981 0.042 0.042
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001469 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.69 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -50821.989 1.000 1.000
Chain 1: 200 -15951.624 1.593 2.186
Chain 1: 300 -8533.555 1.352 1.000
Chain 1: 400 -8190.951 1.024 1.000
Chain 1: 500 -8026.031 0.824 0.869
Chain 1: 600 -8483.629 0.695 0.869
Chain 1: 700 -7622.010 0.612 0.113
Chain 1: 800 -8088.740 0.543 0.113
Chain 1: 900 -7838.272 0.486 0.058
Chain 1: 1000 -7712.526 0.439 0.058
Chain 1: 1100 -7523.383 0.342 0.054
Chain 1: 1200 -7494.728 0.123 0.042
Chain 1: 1300 -7639.444 0.038 0.032
Chain 1: 1400 -7727.441 0.035 0.025
Chain 1: 1500 -7519.172 0.036 0.028
Chain 1: 1600 -7660.738 0.032 0.025
Chain 1: 1700 -7408.426 0.025 0.025
Chain 1: 1800 -7495.674 0.020 0.019
Chain 1: 1900 -7515.926 0.017 0.018
Chain 1: 2000 -7543.372 0.016 0.018
Chain 1: 2100 -7500.675 0.014 0.012
Chain 1: 2200 -7582.065 0.014 0.012
Chain 1: 2300 -7499.704 0.014 0.011
Chain 1: 2400 -7526.498 0.013 0.011
Chain 1: 2500 -7404.744 0.012 0.011
Chain 1: 2600 -7443.226 0.010 0.011
Chain 1: 2700 -7485.679 0.008 0.006 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002948 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.48 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85681.618 1.000 1.000
Chain 1: 200 -13167.056 3.254 5.507
Chain 1: 300 -9614.960 2.292 1.000
Chain 1: 400 -10414.573 1.738 1.000
Chain 1: 500 -8514.507 1.435 0.369
Chain 1: 600 -8160.371 1.203 0.369
Chain 1: 700 -8080.261 1.033 0.223
Chain 1: 800 -8697.550 0.913 0.223
Chain 1: 900 -8431.554 0.815 0.077
Chain 1: 1000 -8239.182 0.736 0.077
Chain 1: 1100 -8516.808 0.639 0.071 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8023.964 0.094 0.061
Chain 1: 1300 -8343.147 0.061 0.043
Chain 1: 1400 -8333.567 0.054 0.038
Chain 1: 1500 -8210.639 0.033 0.033
Chain 1: 1600 -8315.071 0.030 0.032
Chain 1: 1700 -8400.491 0.030 0.032
Chain 1: 1800 -8007.004 0.028 0.032
Chain 1: 1900 -8108.700 0.026 0.023
Chain 1: 2000 -8079.070 0.024 0.015
Chain 1: 2100 -8203.205 0.022 0.015
Chain 1: 2200 -7986.926 0.018 0.015
Chain 1: 2300 -8137.369 0.016 0.015
Chain 1: 2400 -8151.995 0.017 0.015
Chain 1: 2500 -8120.161 0.015 0.013
Chain 1: 2600 -8122.445 0.014 0.013
Chain 1: 2700 -8028.931 0.014 0.013
Chain 1: 2800 -8000.919 0.010 0.012 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002925 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.25 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8399993.433 1.000 1.000
Chain 1: 200 -1583975.834 2.652 4.303
Chain 1: 300 -890574.588 2.027 1.000
Chain 1: 400 -457381.891 1.757 1.000
Chain 1: 500 -357778.655 1.461 0.947
Chain 1: 600 -232756.132 1.307 0.947
Chain 1: 700 -118922.977 1.257 0.947
Chain 1: 800 -86127.124 1.148 0.947
Chain 1: 900 -66456.289 1.053 0.779
Chain 1: 1000 -51243.349 0.978 0.779
Chain 1: 1100 -38714.975 0.910 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37885.345 0.482 0.381
Chain 1: 1300 -25844.783 0.450 0.381
Chain 1: 1400 -25561.427 0.357 0.324
Chain 1: 1500 -22150.215 0.344 0.324
Chain 1: 1600 -21366.282 0.294 0.297
Chain 1: 1700 -20241.012 0.204 0.296
Chain 1: 1800 -20185.033 0.166 0.154
Chain 1: 1900 -20510.764 0.138 0.056
Chain 1: 2000 -19023.178 0.117 0.056
Chain 1: 2100 -19261.404 0.085 0.037
Chain 1: 2200 -19487.568 0.084 0.037
Chain 1: 2300 -19105.138 0.040 0.020
Chain 1: 2400 -18877.398 0.040 0.020
Chain 1: 2500 -18679.395 0.026 0.016
Chain 1: 2600 -18310.039 0.024 0.016
Chain 1: 2700 -18267.103 0.019 0.012
Chain 1: 2800 -17984.180 0.020 0.016
Chain 1: 2900 -18265.211 0.020 0.015
Chain 1: 3000 -18251.426 0.012 0.012
Chain 1: 3100 -18336.351 0.011 0.012
Chain 1: 3200 -18027.302 0.012 0.015
Chain 1: 3300 -18231.796 0.011 0.012
Chain 1: 3400 -17707.222 0.013 0.015
Chain 1: 3500 -18318.355 0.015 0.016
Chain 1: 3600 -17626.006 0.017 0.016
Chain 1: 3700 -18012.096 0.019 0.017
Chain 1: 3800 -16973.305 0.023 0.021
Chain 1: 3900 -16969.482 0.022 0.021
Chain 1: 4000 -17086.780 0.022 0.021
Chain 1: 4100 -17000.640 0.023 0.021
Chain 1: 4200 -16817.188 0.022 0.021
Chain 1: 4300 -16955.367 0.022 0.021
Chain 1: 4400 -16912.459 0.019 0.011
Chain 1: 4500 -16815.048 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001445 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.45 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12540.350 1.000 1.000
Chain 1: 200 -9358.415 0.670 1.000
Chain 1: 300 -8045.762 0.501 0.340
Chain 1: 400 -8271.399 0.383 0.340
Chain 1: 500 -7945.542 0.314 0.163
Chain 1: 600 -8020.276 0.263 0.163
Chain 1: 700 -7942.149 0.227 0.041
Chain 1: 800 -7957.597 0.199 0.041
Chain 1: 900 -7880.201 0.178 0.027
Chain 1: 1000 -8001.549 0.162 0.027
Chain 1: 1100 -7985.692 0.062 0.015
Chain 1: 1200 -7965.985 0.028 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001465 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.65 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -51951.180 1.000 1.000
Chain 1: 200 -16510.226 1.573 2.147
Chain 1: 300 -8797.955 1.341 1.000
Chain 1: 400 -8006.614 1.031 1.000
Chain 1: 500 -8113.594 0.827 0.877
Chain 1: 600 -8144.910 0.690 0.877
Chain 1: 700 -7942.339 0.595 0.099
Chain 1: 800 -8289.712 0.526 0.099
Chain 1: 900 -7795.781 0.474 0.063
Chain 1: 1000 -7994.707 0.429 0.063
Chain 1: 1100 -7721.898 0.333 0.042
Chain 1: 1200 -7623.192 0.120 0.035
Chain 1: 1300 -7832.601 0.035 0.027
Chain 1: 1400 -7734.178 0.026 0.026
Chain 1: 1500 -7586.855 0.027 0.026
Chain 1: 1600 -7776.636 0.029 0.026
Chain 1: 1700 -7508.712 0.030 0.027
Chain 1: 1800 -7645.230 0.027 0.025
Chain 1: 1900 -7572.743 0.022 0.024
Chain 1: 2000 -7608.901 0.020 0.019
Chain 1: 2100 -7602.498 0.016 0.018
Chain 1: 2200 -7717.145 0.017 0.018
Chain 1: 2300 -7604.284 0.015 0.015
Chain 1: 2400 -7654.139 0.015 0.015
Chain 1: 2500 -7570.036 0.014 0.015
Chain 1: 2600 -7530.587 0.012 0.011
Chain 1: 2700 -7523.320 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002953 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.53 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86446.278 1.000 1.000
Chain 1: 200 -13636.598 3.170 5.339
Chain 1: 300 -9974.069 2.235 1.000
Chain 1: 400 -10988.017 1.700 1.000
Chain 1: 500 -8953.106 1.405 0.367
Chain 1: 600 -8913.212 1.172 0.367
Chain 1: 700 -8466.861 1.012 0.227
Chain 1: 800 -8724.190 0.889 0.227
Chain 1: 900 -8699.827 0.791 0.092
Chain 1: 1000 -8620.367 0.712 0.092
Chain 1: 1100 -8795.493 0.614 0.053 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8404.085 0.085 0.047
Chain 1: 1300 -8644.835 0.051 0.029
Chain 1: 1400 -8668.565 0.042 0.028
Chain 1: 1500 -8511.106 0.021 0.020
Chain 1: 1600 -8625.818 0.022 0.020
Chain 1: 1700 -8700.022 0.018 0.019
Chain 1: 1800 -8274.960 0.020 0.019
Chain 1: 1900 -8376.934 0.021 0.019
Chain 1: 2000 -8351.563 0.020 0.019
Chain 1: 2100 -8477.854 0.020 0.015
Chain 1: 2200 -8278.532 0.018 0.015
Chain 1: 2300 -8371.895 0.016 0.013
Chain 1: 2400 -8440.334 0.017 0.013
Chain 1: 2500 -8386.567 0.015 0.012
Chain 1: 2600 -8388.439 0.014 0.011
Chain 1: 2700 -8304.935 0.014 0.011
Chain 1: 2800 -8264.138 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003098 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.98 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8398299.503 1.000 1.000
Chain 1: 200 -1585919.674 2.648 4.296
Chain 1: 300 -892030.931 2.024 1.000
Chain 1: 400 -458108.736 1.755 1.000
Chain 1: 500 -358290.082 1.460 0.947
Chain 1: 600 -233168.636 1.306 0.947
Chain 1: 700 -119391.973 1.256 0.947
Chain 1: 800 -86583.679 1.146 0.947
Chain 1: 900 -66929.732 1.051 0.778
Chain 1: 1000 -51732.987 0.976 0.778
Chain 1: 1100 -39208.131 0.907 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38387.399 0.480 0.379
Chain 1: 1300 -26338.835 0.448 0.379
Chain 1: 1400 -26058.149 0.354 0.319
Chain 1: 1500 -22643.503 0.342 0.319
Chain 1: 1600 -21859.670 0.291 0.294
Chain 1: 1700 -20732.726 0.202 0.294
Chain 1: 1800 -20676.889 0.164 0.151
Chain 1: 1900 -21003.134 0.136 0.054
Chain 1: 2000 -19513.698 0.114 0.054
Chain 1: 2100 -19752.127 0.084 0.036
Chain 1: 2200 -19978.668 0.083 0.036
Chain 1: 2300 -19595.791 0.039 0.020
Chain 1: 2400 -19367.823 0.039 0.020
Chain 1: 2500 -19169.803 0.025 0.016
Chain 1: 2600 -18799.842 0.023 0.016
Chain 1: 2700 -18756.854 0.018 0.012
Chain 1: 2800 -18473.564 0.019 0.015
Chain 1: 2900 -18754.929 0.019 0.015
Chain 1: 3000 -18741.143 0.012 0.012
Chain 1: 3100 -18826.109 0.011 0.012
Chain 1: 3200 -18516.717 0.012 0.015
Chain 1: 3300 -18721.542 0.011 0.012
Chain 1: 3400 -18196.256 0.012 0.015
Chain 1: 3500 -18808.406 0.015 0.015
Chain 1: 3600 -18114.822 0.017 0.015
Chain 1: 3700 -18501.746 0.018 0.017
Chain 1: 3800 -17461.007 0.023 0.021
Chain 1: 3900 -17457.165 0.021 0.021
Chain 1: 4000 -17574.463 0.022 0.021
Chain 1: 4100 -17488.129 0.022 0.021
Chain 1: 4200 -17304.364 0.021 0.021
Chain 1: 4300 -17442.794 0.021 0.021
Chain 1: 4400 -17399.525 0.018 0.011
Chain 1: 4500 -17302.065 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001427 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.27 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12858.626 1.000 1.000
Chain 1: 200 -9465.530 0.679 1.000
Chain 1: 300 -8260.147 0.501 0.358
Chain 1: 400 -8429.043 0.381 0.358
Chain 1: 500 -8072.588 0.314 0.146
Chain 1: 600 -8136.373 0.263 0.146
Chain 1: 700 -8044.052 0.227 0.044
Chain 1: 800 -8066.903 0.199 0.044
Chain 1: 900 -8092.548 0.177 0.020
Chain 1: 1000 -8106.299 0.160 0.020
Chain 1: 1100 -8186.922 0.061 0.011
Chain 1: 1200 -8061.982 0.026 0.011
Chain 1: 1300 -8006.584 0.012 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001409 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.09 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -55313.331 1.000 1.000
Chain 1: 200 -17678.594 1.564 2.129
Chain 1: 300 -8952.143 1.368 1.000
Chain 1: 400 -8533.366 1.038 1.000
Chain 1: 500 -8844.419 0.838 0.975
Chain 1: 600 -8661.256 0.702 0.975
Chain 1: 700 -8504.646 0.604 0.049
Chain 1: 800 -8200.122 0.533 0.049
Chain 1: 900 -8401.253 0.477 0.037
Chain 1: 1000 -7808.336 0.436 0.049
Chain 1: 1100 -7746.879 0.337 0.037
Chain 1: 1200 -7935.809 0.127 0.035
Chain 1: 1300 -7885.590 0.030 0.024
Chain 1: 1400 -7924.166 0.025 0.024
Chain 1: 1500 -7563.370 0.027 0.024
Chain 1: 1600 -7844.712 0.028 0.024
Chain 1: 1700 -7573.412 0.030 0.036
Chain 1: 1800 -7614.680 0.027 0.024
Chain 1: 1900 -7590.326 0.025 0.024
Chain 1: 2000 -7667.495 0.018 0.010
Chain 1: 2100 -7589.190 0.018 0.010
Chain 1: 2200 -7791.057 0.019 0.010
Chain 1: 2300 -7562.572 0.021 0.026
Chain 1: 2400 -7548.585 0.021 0.026
Chain 1: 2500 -7592.750 0.016 0.010
Chain 1: 2600 -7557.751 0.013 0.010
Chain 1: 2700 -7474.925 0.011 0.010
Chain 1: 2800 -7654.148 0.013 0.010
Chain 1: 2900 -7397.856 0.016 0.011
Chain 1: 3000 -7561.383 0.017 0.022
Chain 1: 3100 -7547.452 0.016 0.022
Chain 1: 3200 -7767.371 0.016 0.022
Chain 1: 3300 -7469.376 0.017 0.022
Chain 1: 3400 -7725.543 0.020 0.023
Chain 1: 3500 -7466.327 0.023 0.028
Chain 1: 3600 -7524.417 0.024 0.028
Chain 1: 3700 -7481.301 0.023 0.028
Chain 1: 3800 -7489.258 0.021 0.028
Chain 1: 3900 -7453.712 0.018 0.022
Chain 1: 4000 -7425.290 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003024 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.24 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86467.375 1.000 1.000
Chain 1: 200 -14002.085 3.088 5.175
Chain 1: 300 -10228.668 2.181 1.000
Chain 1: 400 -11877.434 1.671 1.000
Chain 1: 500 -8954.375 1.402 0.369
Chain 1: 600 -9302.263 1.174 0.369
Chain 1: 700 -8749.942 1.016 0.326
Chain 1: 800 -8420.480 0.894 0.326
Chain 1: 900 -8442.472 0.795 0.139
Chain 1: 1000 -8610.285 0.717 0.139
Chain 1: 1100 -8933.054 0.621 0.063 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8517.181 0.108 0.049
Chain 1: 1300 -8789.209 0.074 0.039
Chain 1: 1400 -8788.707 0.060 0.037
Chain 1: 1500 -8693.307 0.029 0.036
Chain 1: 1600 -8745.524 0.026 0.031
Chain 1: 1700 -8838.408 0.020 0.019
Chain 1: 1800 -8399.296 0.022 0.019
Chain 1: 1900 -8495.995 0.023 0.019
Chain 1: 2000 -8506.259 0.021 0.011
Chain 1: 2100 -8601.895 0.018 0.011
Chain 1: 2200 -8386.072 0.016 0.011
Chain 1: 2300 -8547.533 0.015 0.011
Chain 1: 2400 -8392.389 0.017 0.011
Chain 1: 2500 -8467.000 0.016 0.011
Chain 1: 2600 -8377.773 0.017 0.011
Chain 1: 2700 -8412.056 0.016 0.011
Chain 1: 2800 -8363.288 0.012 0.011
Chain 1: 2900 -8477.399 0.012 0.011
Chain 1: 3000 -8394.971 0.013 0.011
Chain 1: 3100 -8355.562 0.012 0.011
Chain 1: 3200 -8327.973 0.010 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003094 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.94 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8393620.824 1.000 1.000
Chain 1: 200 -1585766.479 2.647 4.293
Chain 1: 300 -891882.415 2.024 1.000
Chain 1: 400 -458682.716 1.754 1.000
Chain 1: 500 -358901.082 1.459 0.944
Chain 1: 600 -233823.019 1.305 0.944
Chain 1: 700 -119891.859 1.254 0.944
Chain 1: 800 -87081.052 1.144 0.944
Chain 1: 900 -67402.659 1.050 0.778
Chain 1: 1000 -52194.167 0.974 0.778
Chain 1: 1100 -39658.814 0.905 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38841.237 0.478 0.377
Chain 1: 1300 -26769.576 0.446 0.377
Chain 1: 1400 -26490.809 0.352 0.316
Chain 1: 1500 -23069.662 0.339 0.316
Chain 1: 1600 -22285.062 0.289 0.292
Chain 1: 1700 -21154.409 0.200 0.291
Chain 1: 1800 -21098.220 0.162 0.148
Chain 1: 1900 -21425.080 0.134 0.053
Chain 1: 2000 -19932.541 0.113 0.053
Chain 1: 2100 -20171.288 0.082 0.035
Chain 1: 2200 -20398.532 0.081 0.035
Chain 1: 2300 -20014.803 0.038 0.019
Chain 1: 2400 -19786.510 0.038 0.019
Chain 1: 2500 -19588.646 0.025 0.015
Chain 1: 2600 -19217.937 0.023 0.015
Chain 1: 2700 -19174.678 0.018 0.012
Chain 1: 2800 -18891.171 0.019 0.015
Chain 1: 2900 -19172.835 0.019 0.015
Chain 1: 3000 -19159.006 0.012 0.012
Chain 1: 3100 -19244.089 0.011 0.012
Chain 1: 3200 -18934.246 0.011 0.015
Chain 1: 3300 -19139.391 0.011 0.012
Chain 1: 3400 -18613.364 0.012 0.015
Chain 1: 3500 -19226.720 0.014 0.015
Chain 1: 3600 -18531.477 0.016 0.015
Chain 1: 3700 -18919.651 0.018 0.016
Chain 1: 3800 -17876.450 0.022 0.021
Chain 1: 3900 -17872.518 0.021 0.021
Chain 1: 4000 -17989.824 0.021 0.021
Chain 1: 4100 -17903.401 0.022 0.021
Chain 1: 4200 -17719.036 0.021 0.021
Chain 1: 4300 -17857.869 0.021 0.021
Chain 1: 4400 -17814.142 0.018 0.010
Chain 1: 4500 -17716.582 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001302 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.02 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13471.172 1.000 1.000
Chain 1: 200 -10288.413 0.655 1.000
Chain 1: 300 -8815.331 0.492 0.309
Chain 1: 400 -9025.919 0.375 0.309
Chain 1: 500 -8568.033 0.311 0.167
Chain 1: 600 -8709.751 0.262 0.167
Chain 1: 700 -8701.251 0.224 0.053
Chain 1: 800 -8667.344 0.197 0.053
Chain 1: 900 -8697.207 0.175 0.023
Chain 1: 1000 -8710.783 0.158 0.023
Chain 1: 1100 -8711.565 0.058 0.016
Chain 1: 1200 -8629.223 0.028 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001585 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.85 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -47286.080 1.000 1.000
Chain 1: 200 -16601.612 1.424 1.848
Chain 1: 300 -9271.511 1.213 1.000
Chain 1: 400 -8304.873 0.939 1.000
Chain 1: 500 -8849.983 0.763 0.791
Chain 1: 600 -9057.915 0.640 0.791
Chain 1: 700 -7912.180 0.569 0.145
Chain 1: 800 -8466.828 0.506 0.145
Chain 1: 900 -8333.014 0.452 0.116
Chain 1: 1000 -7641.524 0.416 0.116
Chain 1: 1100 -7885.211 0.319 0.090
Chain 1: 1200 -8104.492 0.137 0.066
Chain 1: 1300 -7833.076 0.061 0.062
Chain 1: 1400 -7572.933 0.053 0.035
Chain 1: 1500 -7725.677 0.049 0.034
Chain 1: 1600 -7764.671 0.047 0.034
Chain 1: 1700 -7424.519 0.037 0.034
Chain 1: 1800 -7659.722 0.033 0.031
Chain 1: 1900 -7550.489 0.033 0.031
Chain 1: 2000 -8003.684 0.030 0.031
Chain 1: 2100 -7687.886 0.031 0.034
Chain 1: 2200 -7606.553 0.029 0.034
Chain 1: 2300 -7710.993 0.027 0.031
Chain 1: 2400 -7700.188 0.024 0.020
Chain 1: 2500 -7636.952 0.023 0.014
Chain 1: 2600 -7593.038 0.023 0.014
Chain 1: 2700 -7597.643 0.018 0.014
Chain 1: 2800 -7575.840 0.016 0.011
Chain 1: 2900 -7418.425 0.016 0.011
Chain 1: 3000 -7581.379 0.013 0.011
Chain 1: 3100 -7563.173 0.009 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003065 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.65 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87809.522 1.000 1.000
Chain 1: 200 -14628.051 3.001 5.003
Chain 1: 300 -10816.148 2.118 1.000
Chain 1: 400 -12780.980 1.627 1.000
Chain 1: 500 -9229.270 1.379 0.385
Chain 1: 600 -9155.235 1.150 0.385
Chain 1: 700 -9329.040 0.989 0.352
Chain 1: 800 -9380.616 0.866 0.352
Chain 1: 900 -9480.139 0.771 0.154
Chain 1: 1000 -9370.963 0.695 0.154
Chain 1: 1100 -9533.065 0.597 0.019 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -9104.433 0.101 0.019
Chain 1: 1300 -9451.742 0.069 0.019
Chain 1: 1400 -9353.813 0.055 0.017
Chain 1: 1500 -9297.597 0.017 0.012
Chain 1: 1600 -9332.880 0.017 0.012
Chain 1: 1700 -9413.777 0.016 0.010
Chain 1: 1800 -8970.368 0.020 0.012
Chain 1: 1900 -9070.072 0.020 0.012
Chain 1: 2000 -9089.467 0.019 0.011
Chain 1: 2100 -9175.276 0.018 0.010
Chain 1: 2200 -8955.657 0.016 0.010
Chain 1: 2300 -9153.217 0.015 0.010
Chain 1: 2400 -8966.573 0.016 0.011
Chain 1: 2500 -9040.659 0.016 0.011
Chain 1: 2600 -8951.319 0.017 0.011
Chain 1: 2700 -8985.012 0.016 0.011
Chain 1: 2800 -8935.845 0.012 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003053 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.53 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8399489.241 1.000 1.000
Chain 1: 200 -1585560.024 2.649 4.297
Chain 1: 300 -892512.495 2.025 1.000
Chain 1: 400 -459116.886 1.754 1.000
Chain 1: 500 -359572.595 1.459 0.944
Chain 1: 600 -234474.441 1.305 0.944
Chain 1: 700 -120544.124 1.253 0.944
Chain 1: 800 -87714.120 1.143 0.944
Chain 1: 900 -68038.291 1.049 0.777
Chain 1: 1000 -52824.713 0.972 0.777
Chain 1: 1100 -40280.270 0.904 0.534 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39463.625 0.476 0.374
Chain 1: 1300 -27383.652 0.442 0.374
Chain 1: 1400 -27103.601 0.349 0.311
Chain 1: 1500 -23680.706 0.336 0.311
Chain 1: 1600 -22895.164 0.286 0.289
Chain 1: 1700 -21763.855 0.197 0.288
Chain 1: 1800 -21707.330 0.159 0.145
Chain 1: 1900 -22034.358 0.132 0.052
Chain 1: 2000 -20541.164 0.110 0.052
Chain 1: 2100 -20779.895 0.080 0.034
Chain 1: 2200 -21007.314 0.079 0.034
Chain 1: 2300 -20623.438 0.037 0.019
Chain 1: 2400 -20395.160 0.037 0.019
Chain 1: 2500 -20197.228 0.024 0.015
Chain 1: 2600 -19826.466 0.022 0.015
Chain 1: 2700 -19783.148 0.017 0.011
Chain 1: 2800 -19499.649 0.018 0.015
Chain 1: 2900 -19781.369 0.018 0.014
Chain 1: 3000 -19767.491 0.011 0.011
Chain 1: 3100 -19852.595 0.011 0.011
Chain 1: 3200 -19542.671 0.011 0.014
Chain 1: 3300 -19747.875 0.010 0.011
Chain 1: 3400 -19221.728 0.012 0.014
Chain 1: 3500 -19835.218 0.014 0.015
Chain 1: 3600 -19139.835 0.016 0.015
Chain 1: 3700 -19528.190 0.017 0.016
Chain 1: 3800 -18484.647 0.022 0.020
Chain 1: 3900 -18480.709 0.020 0.020
Chain 1: 4000 -18598.029 0.021 0.020
Chain 1: 4100 -18511.609 0.021 0.020
Chain 1: 4200 -18327.147 0.020 0.020
Chain 1: 4300 -18466.040 0.020 0.020
Chain 1: 4400 -18422.282 0.017 0.010
Chain 1: 4500 -18324.712 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001157 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 11.57 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12295.611 1.000 1.000
Chain 1: 200 -9209.372 0.668 1.000
Chain 1: 300 -8056.654 0.493 0.335
Chain 1: 400 -8202.166 0.374 0.335
Chain 1: 500 -8097.383 0.302 0.143
Chain 1: 600 -7979.816 0.254 0.143
Chain 1: 700 -7908.844 0.219 0.018
Chain 1: 800 -7919.726 0.192 0.018
Chain 1: 900 -7820.503 0.172 0.015
Chain 1: 1000 -8012.067 0.157 0.018
Chain 1: 1100 -7966.286 0.058 0.015
Chain 1: 1200 -7952.966 0.024 0.013
Chain 1: 1300 -7880.616 0.011 0.013
Chain 1: 1400 -7904.422 0.009 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001475 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.75 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -51069.337 1.000 1.000
Chain 1: 200 -16128.647 1.583 2.166
Chain 1: 300 -8625.474 1.345 1.000
Chain 1: 400 -8238.302 1.021 1.000
Chain 1: 500 -8196.926 0.818 0.870
Chain 1: 600 -8437.278 0.686 0.870
Chain 1: 700 -7738.284 0.601 0.090
Chain 1: 800 -8072.798 0.531 0.090
Chain 1: 900 -8005.925 0.473 0.047
Chain 1: 1000 -7630.716 0.431 0.049
Chain 1: 1100 -7659.410 0.331 0.047
Chain 1: 1200 -7689.162 0.115 0.041
Chain 1: 1300 -7588.958 0.029 0.028
Chain 1: 1400 -7731.081 0.026 0.018
Chain 1: 1500 -7541.270 0.028 0.025
Chain 1: 1600 -7688.750 0.027 0.019
Chain 1: 1700 -7426.887 0.022 0.019
Chain 1: 1800 -7506.670 0.019 0.018
Chain 1: 1900 -7490.686 0.018 0.018
Chain 1: 2000 -7521.471 0.014 0.013
Chain 1: 2100 -7507.691 0.013 0.013
Chain 1: 2200 -7615.611 0.014 0.014
Chain 1: 2300 -7524.329 0.014 0.014
Chain 1: 2400 -7560.316 0.013 0.012
Chain 1: 2500 -7556.624 0.010 0.011
Chain 1: 2600 -7465.328 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003052 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.52 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86054.973 1.000 1.000
Chain 1: 200 -13357.747 3.221 5.442
Chain 1: 300 -9799.831 2.268 1.000
Chain 1: 400 -10531.358 1.719 1.000
Chain 1: 500 -8698.466 1.417 0.363
Chain 1: 600 -8400.797 1.187 0.363
Chain 1: 700 -8450.904 1.018 0.211
Chain 1: 800 -8619.243 0.893 0.211
Chain 1: 900 -8674.811 0.795 0.069
Chain 1: 1000 -8391.480 0.719 0.069
Chain 1: 1100 -8707.873 0.622 0.036 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8428.117 0.081 0.035
Chain 1: 1300 -8554.649 0.047 0.034
Chain 1: 1400 -8460.105 0.041 0.033
Chain 1: 1500 -8407.978 0.020 0.020
Chain 1: 1600 -8395.479 0.017 0.015
Chain 1: 1700 -8322.364 0.017 0.015
Chain 1: 1800 -8209.434 0.017 0.014
Chain 1: 1900 -8326.878 0.017 0.014
Chain 1: 2000 -8287.255 0.014 0.014
Chain 1: 2100 -8415.967 0.012 0.014
Chain 1: 2200 -8206.845 0.012 0.014
Chain 1: 2300 -8349.496 0.012 0.014
Chain 1: 2400 -8363.630 0.011 0.014
Chain 1: 2500 -8331.528 0.011 0.014
Chain 1: 2600 -8330.761 0.010 0.014
Chain 1: 2700 -8238.786 0.011 0.014
Chain 1: 2800 -8213.710 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003262 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.62 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8414958.887 1.000 1.000
Chain 1: 200 -1586071.962 2.653 4.306
Chain 1: 300 -891105.333 2.028 1.000
Chain 1: 400 -457586.979 1.758 1.000
Chain 1: 500 -357645.481 1.462 0.947
Chain 1: 600 -232750.253 1.308 0.947
Chain 1: 700 -118993.588 1.258 0.947
Chain 1: 800 -86224.257 1.148 0.947
Chain 1: 900 -66574.533 1.053 0.780
Chain 1: 1000 -51377.044 0.978 0.780
Chain 1: 1100 -38864.746 0.910 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38036.666 0.481 0.380
Chain 1: 1300 -26011.161 0.450 0.380
Chain 1: 1400 -25730.045 0.356 0.322
Chain 1: 1500 -22322.536 0.343 0.322
Chain 1: 1600 -21539.728 0.293 0.296
Chain 1: 1700 -20416.185 0.203 0.295
Chain 1: 1800 -20360.581 0.165 0.153
Chain 1: 1900 -20686.231 0.138 0.055
Chain 1: 2000 -19199.594 0.116 0.055
Chain 1: 2100 -19437.833 0.085 0.036
Chain 1: 2200 -19663.814 0.084 0.036
Chain 1: 2300 -19281.553 0.039 0.020
Chain 1: 2400 -19053.829 0.040 0.020
Chain 1: 2500 -18855.812 0.025 0.016
Chain 1: 2600 -18486.610 0.024 0.016
Chain 1: 2700 -18443.681 0.018 0.012
Chain 1: 2800 -18160.805 0.020 0.016
Chain 1: 2900 -18441.743 0.020 0.015
Chain 1: 3000 -18428.018 0.012 0.012
Chain 1: 3100 -18512.944 0.011 0.012
Chain 1: 3200 -18203.949 0.012 0.015
Chain 1: 3300 -18408.366 0.011 0.012
Chain 1: 3400 -17883.929 0.013 0.015
Chain 1: 3500 -18494.858 0.015 0.016
Chain 1: 3600 -17802.706 0.017 0.016
Chain 1: 3700 -18188.670 0.019 0.017
Chain 1: 3800 -17150.209 0.023 0.021
Chain 1: 3900 -17146.361 0.022 0.021
Chain 1: 4000 -17263.686 0.022 0.021
Chain 1: 4100 -17177.581 0.022 0.021
Chain 1: 4200 -16994.163 0.022 0.021
Chain 1: 4300 -17132.320 0.021 0.021
Chain 1: 4400 -17089.467 0.019 0.011
Chain 1: 4500 -16992.035 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001285 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.85 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49423.680 1.000 1.000
Chain 1: 200 -25069.587 0.986 1.000
Chain 1: 300 -35042.348 0.752 0.971
Chain 1: 400 -20149.693 0.749 0.971
Chain 1: 500 -13991.331 0.687 0.739
Chain 1: 600 -26092.591 0.650 0.739
Chain 1: 700 -11960.788 0.726 0.739
Chain 1: 800 -13207.655 0.647 0.739
Chain 1: 900 -13372.198 0.576 0.464
Chain 1: 1000 -13618.507 0.521 0.464
Chain 1: 1100 -10920.790 0.445 0.440
Chain 1: 1200 -10658.035 0.351 0.285
Chain 1: 1300 -10710.976 0.323 0.247
Chain 1: 1400 -10509.112 0.251 0.094
Chain 1: 1500 -10539.857 0.207 0.025
Chain 1: 1600 -12196.160 0.174 0.025
Chain 1: 1700 -12775.381 0.060 0.025
Chain 1: 1800 -10605.380 0.071 0.025
Chain 1: 1900 -11196.061 0.076 0.045
Chain 1: 2000 -12252.608 0.082 0.053
Chain 1: 2100 -9494.265 0.087 0.053
Chain 1: 2200 -9721.190 0.087 0.053
Chain 1: 2300 -9270.672 0.091 0.053
Chain 1: 2400 -20296.833 0.143 0.086
Chain 1: 2500 -10087.329 0.244 0.136
Chain 1: 2600 -9411.787 0.238 0.086
Chain 1: 2700 -10641.928 0.245 0.116
Chain 1: 2800 -10951.343 0.227 0.086
Chain 1: 2900 -10391.799 0.227 0.086
Chain 1: 3000 -9166.444 0.232 0.116
Chain 1: 3100 -9834.995 0.210 0.072
Chain 1: 3200 -13135.048 0.233 0.116
Chain 1: 3300 -10582.902 0.252 0.134
Chain 1: 3400 -17388.710 0.237 0.134
Chain 1: 3500 -9632.127 0.216 0.134
Chain 1: 3600 -10508.905 0.217 0.134
Chain 1: 3700 -9185.653 0.220 0.144
Chain 1: 3800 -15213.655 0.257 0.241
Chain 1: 3900 -15137.880 0.252 0.241
Chain 1: 4000 -10292.462 0.286 0.251
Chain 1: 4100 -10065.814 0.281 0.251
Chain 1: 4200 -10146.148 0.257 0.241
Chain 1: 4300 -11394.370 0.244 0.144
Chain 1: 4400 -10172.238 0.216 0.120
Chain 1: 4500 -9777.719 0.140 0.110
Chain 1: 4600 -9152.330 0.138 0.110
Chain 1: 4700 -13791.367 0.158 0.110
Chain 1: 4800 -12452.585 0.129 0.108
Chain 1: 4900 -8846.275 0.169 0.110
Chain 1: 5000 -11787.895 0.147 0.110
Chain 1: 5100 -8650.208 0.181 0.120
Chain 1: 5200 -8876.299 0.183 0.120
Chain 1: 5300 -8553.092 0.176 0.120
Chain 1: 5400 -13526.934 0.200 0.250
Chain 1: 5500 -9423.920 0.240 0.336
Chain 1: 5600 -8732.624 0.241 0.336
Chain 1: 5700 -18036.006 0.259 0.363
Chain 1: 5800 -9035.014 0.348 0.368
Chain 1: 5900 -8645.328 0.311 0.363
Chain 1: 6000 -9396.187 0.295 0.363
Chain 1: 6100 -10566.243 0.269 0.111
Chain 1: 6200 -11483.113 0.275 0.111
Chain 1: 6300 -8918.244 0.300 0.288
Chain 1: 6400 -8253.858 0.271 0.111
Chain 1: 6500 -9200.779 0.238 0.103
Chain 1: 6600 -8797.875 0.234 0.103
Chain 1: 6700 -9608.097 0.191 0.084
Chain 1: 6800 -13033.641 0.118 0.084
Chain 1: 6900 -10289.258 0.140 0.103
Chain 1: 7000 -10570.308 0.135 0.103
Chain 1: 7100 -9232.485 0.138 0.103
Chain 1: 7200 -8390.173 0.140 0.103
Chain 1: 7300 -8612.362 0.114 0.100
Chain 1: 7400 -8709.322 0.107 0.100
Chain 1: 7500 -10971.688 0.117 0.100
Chain 1: 7600 -8265.976 0.146 0.145
Chain 1: 7700 -8561.743 0.141 0.145
Chain 1: 7800 -8758.174 0.117 0.100
Chain 1: 7900 -8812.865 0.091 0.035
Chain 1: 8000 -8627.405 0.090 0.035
Chain 1: 8100 -12938.015 0.109 0.035
Chain 1: 8200 -8654.789 0.148 0.035
Chain 1: 8300 -8280.015 0.150 0.045
Chain 1: 8400 -8536.861 0.152 0.045
Chain 1: 8500 -8279.579 0.135 0.035
Chain 1: 8600 -10446.551 0.123 0.035
Chain 1: 8700 -10697.576 0.122 0.031
Chain 1: 8800 -11021.894 0.122 0.031
Chain 1: 8900 -8459.750 0.152 0.045
Chain 1: 9000 -9742.198 0.163 0.132
Chain 1: 9100 -8585.066 0.143 0.132
Chain 1: 9200 -13016.237 0.128 0.132
Chain 1: 9300 -10480.203 0.147 0.135
Chain 1: 9400 -12815.030 0.163 0.182
Chain 1: 9500 -8247.569 0.215 0.207
Chain 1: 9600 -9740.309 0.209 0.182
Chain 1: 9700 -8115.055 0.227 0.200
Chain 1: 9800 -8236.554 0.226 0.200
Chain 1: 9900 -9265.052 0.206 0.182
Chain 1: 10000 -11877.010 0.215 0.200
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001383 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.83 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -63825.820 1.000 1.000
Chain 1: 200 -18544.887 1.721 2.442
Chain 1: 300 -8930.901 1.506 1.076
Chain 1: 400 -8049.391 1.157 1.076
Chain 1: 500 -9024.190 0.947 1.000
Chain 1: 600 -8654.272 0.796 1.000
Chain 1: 700 -8544.766 0.684 0.110
Chain 1: 800 -8468.727 0.600 0.110
Chain 1: 900 -7976.280 0.540 0.108
Chain 1: 1000 -7893.800 0.487 0.108
Chain 1: 1100 -7658.938 0.390 0.062
Chain 1: 1200 -7678.908 0.146 0.043
Chain 1: 1300 -7571.499 0.040 0.031
Chain 1: 1400 -7741.545 0.031 0.022
Chain 1: 1500 -7592.715 0.023 0.020
Chain 1: 1600 -7770.245 0.021 0.020
Chain 1: 1700 -7584.352 0.022 0.022
Chain 1: 1800 -7705.402 0.022 0.022
Chain 1: 1900 -7686.421 0.017 0.020
Chain 1: 2000 -7644.007 0.016 0.020
Chain 1: 2100 -7631.149 0.013 0.016
Chain 1: 2200 -7769.398 0.015 0.018
Chain 1: 2300 -7613.396 0.015 0.020
Chain 1: 2400 -7708.053 0.014 0.018
Chain 1: 2500 -7524.270 0.015 0.018
Chain 1: 2600 -7567.673 0.013 0.016
Chain 1: 2700 -7589.872 0.011 0.012
Chain 1: 2800 -7655.314 0.010 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002991 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.91 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85437.641 1.000 1.000
Chain 1: 200 -13721.459 3.113 5.227
Chain 1: 300 -9996.371 2.200 1.000
Chain 1: 400 -11539.191 1.683 1.000
Chain 1: 500 -8836.351 1.408 0.373
Chain 1: 600 -8374.065 1.182 0.373
Chain 1: 700 -8275.231 1.015 0.306
Chain 1: 800 -8763.420 0.895 0.306
Chain 1: 900 -8709.908 0.796 0.134
Chain 1: 1000 -8395.269 0.721 0.134
Chain 1: 1100 -8794.143 0.625 0.056 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8332.031 0.108 0.055
Chain 1: 1300 -8661.681 0.074 0.055
Chain 1: 1400 -8438.485 0.064 0.045
Chain 1: 1500 -8479.110 0.034 0.038
Chain 1: 1600 -8583.807 0.029 0.037
Chain 1: 1700 -8647.148 0.029 0.037
Chain 1: 1800 -8199.838 0.029 0.037
Chain 1: 1900 -8308.067 0.029 0.037
Chain 1: 2000 -8292.570 0.026 0.026
Chain 1: 2100 -8429.202 0.023 0.016
Chain 1: 2200 -8203.917 0.020 0.016
Chain 1: 2300 -8319.023 0.018 0.014
Chain 1: 2400 -8376.141 0.016 0.013
Chain 1: 2500 -8316.886 0.016 0.013
Chain 1: 2600 -8330.035 0.015 0.013
Chain 1: 2700 -8237.869 0.015 0.013
Chain 1: 2800 -8184.599 0.011 0.011
Chain 1: 2900 -8288.014 0.011 0.011
Chain 1: 3000 -8126.365 0.012 0.012
Chain 1: 3100 -8269.136 0.012 0.012
Chain 1: 3200 -8138.533 0.011 0.012
Chain 1: 3300 -8354.458 0.012 0.012
Chain 1: 3400 -8400.176 0.012 0.012
Chain 1: 3500 -8223.933 0.014 0.016
Chain 1: 3600 -8097.136 0.015 0.016
Chain 1: 3700 -8240.370 0.016 0.017
Chain 1: 3800 -8254.043 0.015 0.017
Chain 1: 3900 -8030.551 0.017 0.017
Chain 1: 4000 -8195.383 0.017 0.017
Chain 1: 4100 -8105.153 0.016 0.017
Chain 1: 4200 -8091.299 0.015 0.017
Chain 1: 4300 -8124.900 0.013 0.016
Chain 1: 4400 -8078.795 0.013 0.016
Chain 1: 4500 -8179.400 0.012 0.012
Chain 1: 4600 -8071.495 0.012 0.012
Chain 1: 4700 -8275.928 0.012 0.012
Chain 1: 4800 -8160.253 0.014 0.013
Chain 1: 4900 -8169.772 0.011 0.012
Chain 1: 5000 -8103.411 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003041 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.41 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8408563.116 1.000 1.000
Chain 1: 200 -1585809.926 2.651 4.302
Chain 1: 300 -891176.750 2.027 1.000
Chain 1: 400 -458309.390 1.757 1.000
Chain 1: 500 -358688.310 1.461 0.944
Chain 1: 600 -233551.371 1.307 0.944
Chain 1: 700 -119641.959 1.256 0.944
Chain 1: 800 -86814.688 1.146 0.944
Chain 1: 900 -67126.993 1.051 0.779
Chain 1: 1000 -51915.556 0.976 0.779
Chain 1: 1100 -39376.279 0.907 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38553.766 0.479 0.378
Chain 1: 1300 -26484.061 0.447 0.378
Chain 1: 1400 -26202.942 0.354 0.318
Chain 1: 1500 -22783.306 0.341 0.318
Chain 1: 1600 -21998.692 0.291 0.293
Chain 1: 1700 -20868.755 0.201 0.293
Chain 1: 1800 -20812.508 0.164 0.150
Chain 1: 1900 -21139.244 0.136 0.054
Chain 1: 2000 -19647.412 0.114 0.054
Chain 1: 2100 -19885.871 0.083 0.036
Chain 1: 2200 -20113.132 0.082 0.036
Chain 1: 2300 -19729.520 0.039 0.019
Chain 1: 2400 -19501.366 0.039 0.019
Chain 1: 2500 -19303.511 0.025 0.015
Chain 1: 2600 -18932.960 0.023 0.015
Chain 1: 2700 -18889.704 0.018 0.012
Chain 1: 2800 -18606.333 0.019 0.015
Chain 1: 2900 -18887.889 0.019 0.015
Chain 1: 3000 -18874.017 0.012 0.012
Chain 1: 3100 -18959.120 0.011 0.012
Chain 1: 3200 -18649.380 0.012 0.015
Chain 1: 3300 -18854.415 0.011 0.012
Chain 1: 3400 -18328.599 0.012 0.015
Chain 1: 3500 -18941.645 0.015 0.015
Chain 1: 3600 -18246.815 0.016 0.015
Chain 1: 3700 -18634.754 0.018 0.017
Chain 1: 3800 -17592.161 0.023 0.021
Chain 1: 3900 -17588.269 0.021 0.021
Chain 1: 4000 -17705.563 0.022 0.021
Chain 1: 4100 -17619.231 0.022 0.021
Chain 1: 4200 -17434.968 0.021 0.021
Chain 1: 4300 -17573.707 0.021 0.021
Chain 1: 4400 -17530.106 0.018 0.011
Chain 1: 4500 -17432.610 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001187 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 11.87 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13172.959 1.000 1.000
Chain 1: 200 -9826.028 0.670 1.000
Chain 1: 300 -8332.882 0.507 0.341
Chain 1: 400 -8351.968 0.381 0.341
Chain 1: 500 -8287.538 0.306 0.179
Chain 1: 600 -8185.574 0.257 0.179
Chain 1: 700 -8093.722 0.222 0.012
Chain 1: 800 -8127.881 0.195 0.012
Chain 1: 900 -7998.690 0.175 0.012
Chain 1: 1000 -8149.234 0.159 0.016
Chain 1: 1100 -8228.969 0.060 0.012
Chain 1: 1200 -8102.718 0.028 0.012
Chain 1: 1300 -8105.125 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002288 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 22.88 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58124.163 1.000 1.000
Chain 1: 200 -17768.244 1.636 2.271
Chain 1: 300 -8731.058 1.435 1.035
Chain 1: 400 -8102.768 1.096 1.035
Chain 1: 500 -7995.072 0.879 1.000
Chain 1: 600 -7856.828 0.736 1.000
Chain 1: 700 -7872.807 0.631 0.078
Chain 1: 800 -7682.959 0.555 0.078
Chain 1: 900 -8470.618 0.504 0.078
Chain 1: 1000 -7627.582 0.465 0.093
Chain 1: 1100 -7669.174 0.365 0.078
Chain 1: 1200 -7654.924 0.138 0.025
Chain 1: 1300 -7645.149 0.035 0.018
Chain 1: 1400 -7769.705 0.029 0.016
Chain 1: 1500 -7559.454 0.030 0.018
Chain 1: 1600 -7763.064 0.031 0.025
Chain 1: 1700 -7522.197 0.034 0.026
Chain 1: 1800 -7578.200 0.032 0.026
Chain 1: 1900 -7587.644 0.023 0.016
Chain 1: 2000 -7583.754 0.012 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003113 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.13 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86309.150 1.000 1.000
Chain 1: 200 -13656.216 3.160 5.320
Chain 1: 300 -10048.767 2.226 1.000
Chain 1: 400 -10757.282 1.686 1.000
Chain 1: 500 -8986.325 1.388 0.359
Chain 1: 600 -8563.317 1.165 0.359
Chain 1: 700 -8584.472 0.999 0.197
Chain 1: 800 -8859.229 0.878 0.197
Chain 1: 900 -8765.321 0.782 0.066
Chain 1: 1000 -8604.755 0.705 0.066
Chain 1: 1100 -8911.140 0.609 0.049 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8507.503 0.082 0.047
Chain 1: 1300 -8738.071 0.048 0.034
Chain 1: 1400 -8723.731 0.042 0.031
Chain 1: 1500 -8629.441 0.023 0.026
Chain 1: 1600 -8735.924 0.020 0.019
Chain 1: 1700 -8824.590 0.020 0.019
Chain 1: 1800 -8417.707 0.022 0.019
Chain 1: 1900 -8514.703 0.022 0.019
Chain 1: 2000 -8486.786 0.021 0.012
Chain 1: 2100 -8607.257 0.019 0.012
Chain 1: 2200 -8416.814 0.016 0.012
Chain 1: 2300 -8554.188 0.015 0.012
Chain 1: 2400 -8561.378 0.015 0.012
Chain 1: 2500 -8528.026 0.014 0.012
Chain 1: 2600 -8526.076 0.013 0.011
Chain 1: 2700 -8439.944 0.013 0.011
Chain 1: 2800 -8405.261 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003004 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.04 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8410780.205 1.000 1.000
Chain 1: 200 -1585230.203 2.653 4.306
Chain 1: 300 -890638.178 2.029 1.000
Chain 1: 400 -457546.466 1.758 1.000
Chain 1: 500 -357674.315 1.462 0.947
Chain 1: 600 -232756.589 1.308 0.947
Chain 1: 700 -119205.198 1.257 0.947
Chain 1: 800 -86443.210 1.147 0.947
Chain 1: 900 -66825.394 1.053 0.780
Chain 1: 1000 -51649.210 0.977 0.780
Chain 1: 1100 -39153.786 0.909 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38335.143 0.480 0.379
Chain 1: 1300 -26321.749 0.448 0.379
Chain 1: 1400 -26043.626 0.354 0.319
Chain 1: 1500 -22638.010 0.341 0.319
Chain 1: 1600 -21856.858 0.291 0.294
Chain 1: 1700 -20734.180 0.201 0.294
Chain 1: 1800 -20679.252 0.164 0.150
Chain 1: 1900 -21005.264 0.136 0.054
Chain 1: 2000 -19518.675 0.114 0.054
Chain 1: 2100 -19756.948 0.084 0.036
Chain 1: 2200 -19982.905 0.083 0.036
Chain 1: 2300 -19600.593 0.039 0.020
Chain 1: 2400 -19372.775 0.039 0.020
Chain 1: 2500 -19174.648 0.025 0.016
Chain 1: 2600 -18805.111 0.023 0.016
Chain 1: 2700 -18762.262 0.018 0.012
Chain 1: 2800 -18479.046 0.019 0.015
Chain 1: 2900 -18760.242 0.019 0.015
Chain 1: 3000 -18746.499 0.012 0.012
Chain 1: 3100 -18831.410 0.011 0.012
Chain 1: 3200 -18522.262 0.012 0.015
Chain 1: 3300 -18726.908 0.011 0.012
Chain 1: 3400 -18202.001 0.012 0.015
Chain 1: 3500 -18813.513 0.015 0.015
Chain 1: 3600 -18120.730 0.017 0.015
Chain 1: 3700 -18507.060 0.018 0.017
Chain 1: 3800 -17467.519 0.023 0.021
Chain 1: 3900 -17463.685 0.021 0.021
Chain 1: 4000 -17581.009 0.022 0.021
Chain 1: 4100 -17494.719 0.022 0.021
Chain 1: 4200 -17311.221 0.021 0.021
Chain 1: 4300 -17449.472 0.021 0.021
Chain 1: 4400 -17406.444 0.018 0.011
Chain 1: 4500 -17308.991 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001508 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.08 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49081.641 1.000 1.000
Chain 1: 200 -12899.491 1.902 2.805
Chain 1: 300 -16361.841 1.339 1.000
Chain 1: 400 -19250.466 1.042 1.000
Chain 1: 500 -15124.313 0.888 0.273
Chain 1: 600 -28154.324 0.817 0.463
Chain 1: 700 -16477.859 0.802 0.463
Chain 1: 800 -11427.438 0.757 0.463
Chain 1: 900 -11256.612 0.674 0.442
Chain 1: 1000 -11955.501 0.613 0.442
Chain 1: 1100 -10161.695 0.530 0.273 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -11618.917 0.262 0.212
Chain 1: 1300 -11087.508 0.246 0.177
Chain 1: 1400 -9912.877 0.243 0.177
Chain 1: 1500 -10850.972 0.224 0.125
Chain 1: 1600 -11031.262 0.180 0.118
Chain 1: 1700 -13128.188 0.125 0.118
Chain 1: 1800 -13217.252 0.081 0.086
Chain 1: 1900 -13694.212 0.083 0.086
Chain 1: 2000 -20802.380 0.111 0.118
Chain 1: 2100 -9738.997 0.207 0.118
Chain 1: 2200 -17549.663 0.239 0.118
Chain 1: 2300 -22405.062 0.256 0.160
Chain 1: 2400 -9134.959 0.390 0.217
Chain 1: 2500 -15762.693 0.423 0.342
Chain 1: 2600 -15490.853 0.423 0.342
Chain 1: 2700 -9802.557 0.465 0.420
Chain 1: 2800 -10640.526 0.472 0.420
Chain 1: 2900 -15040.681 0.498 0.420
Chain 1: 3000 -8939.549 0.532 0.445 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 3100 -10619.096 0.434 0.420
Chain 1: 3200 -13468.576 0.411 0.293
Chain 1: 3300 -9448.321 0.432 0.420
Chain 1: 3400 -17408.565 0.332 0.420
Chain 1: 3500 -15752.929 0.301 0.293
Chain 1: 3600 -8919.427 0.376 0.425
Chain 1: 3700 -9169.914 0.320 0.293
Chain 1: 3800 -11390.494 0.332 0.293
Chain 1: 3900 -10164.706 0.315 0.212
Chain 1: 4000 -10161.141 0.247 0.195
Chain 1: 4100 -9262.972 0.241 0.195
Chain 1: 4200 -9043.252 0.222 0.121
Chain 1: 4300 -14404.333 0.217 0.121
Chain 1: 4400 -12483.151 0.186 0.121
Chain 1: 4500 -11392.092 0.185 0.121
Chain 1: 4600 -8539.338 0.142 0.121
Chain 1: 4700 -8633.548 0.140 0.121
Chain 1: 4800 -8813.557 0.123 0.097
Chain 1: 4900 -8646.638 0.113 0.096
Chain 1: 5000 -13276.588 0.148 0.097
Chain 1: 5100 -8643.764 0.192 0.154
Chain 1: 5200 -8925.459 0.192 0.154
Chain 1: 5300 -10257.020 0.168 0.130
Chain 1: 5400 -8608.815 0.172 0.130
Chain 1: 5500 -12943.186 0.196 0.191
Chain 1: 5600 -10445.647 0.186 0.191
Chain 1: 5700 -9417.271 0.196 0.191
Chain 1: 5800 -9091.624 0.198 0.191
Chain 1: 5900 -11921.297 0.219 0.237
Chain 1: 6000 -8563.675 0.224 0.237
Chain 1: 6100 -12293.756 0.200 0.237
Chain 1: 6200 -8345.385 0.245 0.239
Chain 1: 6300 -8954.269 0.238 0.239
Chain 1: 6400 -10667.478 0.235 0.239
Chain 1: 6500 -8794.769 0.223 0.237
Chain 1: 6600 -8635.648 0.201 0.213
Chain 1: 6700 -8249.453 0.195 0.213
Chain 1: 6800 -8936.432 0.199 0.213
Chain 1: 6900 -11539.309 0.198 0.213
Chain 1: 7000 -10716.827 0.166 0.161
Chain 1: 7100 -9515.844 0.149 0.126
Chain 1: 7200 -8612.653 0.112 0.105
Chain 1: 7300 -8745.641 0.106 0.105
Chain 1: 7400 -11596.622 0.115 0.105
Chain 1: 7500 -8155.833 0.136 0.105
Chain 1: 7600 -11441.786 0.163 0.126
Chain 1: 7700 -8382.112 0.195 0.226
Chain 1: 7800 -8812.465 0.192 0.226
Chain 1: 7900 -8650.110 0.171 0.126
Chain 1: 8000 -8502.010 0.165 0.126
Chain 1: 8100 -11524.550 0.179 0.246
Chain 1: 8200 -8293.264 0.207 0.262
Chain 1: 8300 -8414.477 0.207 0.262
Chain 1: 8400 -8693.175 0.186 0.262
Chain 1: 8500 -8160.241 0.150 0.065
Chain 1: 8600 -8330.004 0.123 0.049
Chain 1: 8700 -9825.908 0.102 0.049
Chain 1: 8800 -11286.322 0.110 0.065
Chain 1: 8900 -12655.067 0.119 0.108
Chain 1: 9000 -8323.900 0.169 0.129
Chain 1: 9100 -8761.697 0.148 0.108
Chain 1: 9200 -8021.333 0.118 0.092
Chain 1: 9300 -8777.915 0.126 0.092
Chain 1: 9400 -8284.477 0.128 0.092
Chain 1: 9500 -8889.037 0.129 0.092
Chain 1: 9600 -8408.793 0.132 0.092
Chain 1: 9700 -8763.186 0.121 0.086
Chain 1: 9800 -10675.513 0.126 0.086
Chain 1: 9900 -10185.845 0.120 0.068
Chain 1: 10000 -8133.746 0.093 0.068
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001387 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.87 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57947.939 1.000 1.000
Chain 1: 200 -17741.901 1.633 2.266
Chain 1: 300 -8703.983 1.435 1.038
Chain 1: 400 -8115.991 1.094 1.038
Chain 1: 500 -8605.046 0.887 1.000
Chain 1: 600 -7992.027 0.752 1.000
Chain 1: 700 -7757.182 0.649 0.077
Chain 1: 800 -8100.883 0.573 0.077
Chain 1: 900 -7990.947 0.511 0.072
Chain 1: 1000 -7858.327 0.461 0.072
Chain 1: 1100 -7838.239 0.362 0.057
Chain 1: 1200 -7848.342 0.135 0.042
Chain 1: 1300 -7788.814 0.032 0.030
Chain 1: 1400 -7681.065 0.026 0.017
Chain 1: 1500 -7535.879 0.022 0.017
Chain 1: 1600 -7569.860 0.015 0.014
Chain 1: 1700 -7554.148 0.012 0.014
Chain 1: 1800 -7628.986 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002565 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.65 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86029.635 1.000 1.000
Chain 1: 200 -13575.644 3.169 5.337
Chain 1: 300 -9898.182 2.236 1.000
Chain 1: 400 -10847.092 1.699 1.000
Chain 1: 500 -8888.158 1.403 0.372
Chain 1: 600 -8370.826 1.180 0.372
Chain 1: 700 -8757.547 1.017 0.220
Chain 1: 800 -8881.652 0.892 0.220
Chain 1: 900 -8719.785 0.795 0.087
Chain 1: 1000 -8505.774 0.718 0.087
Chain 1: 1100 -8692.083 0.620 0.062 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8305.121 0.091 0.047
Chain 1: 1300 -8580.611 0.057 0.044
Chain 1: 1400 -8591.634 0.049 0.032
Chain 1: 1500 -8435.931 0.028 0.025
Chain 1: 1600 -8550.155 0.024 0.021
Chain 1: 1700 -8621.959 0.020 0.019
Chain 1: 1800 -8193.298 0.024 0.021
Chain 1: 1900 -8296.734 0.023 0.021
Chain 1: 2000 -8271.705 0.021 0.018
Chain 1: 2100 -8400.535 0.020 0.015
Chain 1: 2200 -8197.934 0.018 0.015
Chain 1: 2300 -8293.128 0.016 0.013
Chain 1: 2400 -8359.775 0.017 0.013
Chain 1: 2500 -8305.699 0.016 0.012
Chain 1: 2600 -8308.926 0.014 0.011
Chain 1: 2700 -8224.741 0.014 0.011
Chain 1: 2800 -8182.465 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002625 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 26.25 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8410719.658 1.000 1.000
Chain 1: 200 -1587851.341 2.648 4.297
Chain 1: 300 -890756.970 2.027 1.000
Chain 1: 400 -457399.135 1.757 1.000
Chain 1: 500 -357497.870 1.461 0.947
Chain 1: 600 -232657.024 1.307 0.947
Chain 1: 700 -119115.519 1.257 0.947
Chain 1: 800 -86360.474 1.147 0.947
Chain 1: 900 -66752.585 1.052 0.783
Chain 1: 1000 -51585.990 0.976 0.783
Chain 1: 1100 -39093.272 0.908 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38277.776 0.481 0.379
Chain 1: 1300 -26263.641 0.448 0.379
Chain 1: 1400 -25987.310 0.355 0.320
Chain 1: 1500 -22580.930 0.342 0.320
Chain 1: 1600 -21799.673 0.292 0.294
Chain 1: 1700 -20676.725 0.202 0.294
Chain 1: 1800 -20621.853 0.164 0.151
Chain 1: 1900 -20948.230 0.136 0.054
Chain 1: 2000 -19460.517 0.114 0.054
Chain 1: 2100 -19699.056 0.084 0.036
Chain 1: 2200 -19925.253 0.083 0.036
Chain 1: 2300 -19542.589 0.039 0.020
Chain 1: 2400 -19314.603 0.039 0.020
Chain 1: 2500 -19116.420 0.025 0.016
Chain 1: 2600 -18746.677 0.023 0.016
Chain 1: 2700 -18703.659 0.018 0.012
Chain 1: 2800 -18420.268 0.019 0.015
Chain 1: 2900 -18701.603 0.019 0.015
Chain 1: 3000 -18687.901 0.012 0.012
Chain 1: 3100 -18772.875 0.011 0.012
Chain 1: 3200 -18463.490 0.012 0.015
Chain 1: 3300 -18668.269 0.011 0.012
Chain 1: 3400 -18142.949 0.012 0.015
Chain 1: 3500 -18755.103 0.015 0.015
Chain 1: 3600 -18061.426 0.017 0.015
Chain 1: 3700 -18448.450 0.018 0.017
Chain 1: 3800 -17407.524 0.023 0.021
Chain 1: 3900 -17403.602 0.021 0.021
Chain 1: 4000 -17520.967 0.022 0.021
Chain 1: 4100 -17434.642 0.022 0.021
Chain 1: 4200 -17250.762 0.021 0.021
Chain 1: 4300 -17389.286 0.021 0.021
Chain 1: 4400 -17346.015 0.018 0.011
Chain 1: 4500 -17248.476 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001382 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.82 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12718.030 1.000 1.000
Chain 1: 200 -9538.106 0.667 1.000
Chain 1: 300 -8195.145 0.499 0.333
Chain 1: 400 -8297.788 0.377 0.333
Chain 1: 500 -8289.326 0.302 0.164
Chain 1: 600 -8123.422 0.255 0.164
Chain 1: 700 -8041.012 0.220 0.020
Chain 1: 800 -8046.361 0.193 0.020
Chain 1: 900 -7943.562 0.173 0.013
Chain 1: 1000 -8105.687 0.157 0.020
Chain 1: 1100 -8074.407 0.058 0.013
Chain 1: 1200 -8065.849 0.025 0.012
Chain 1: 1300 -8010.498 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001434 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.34 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49472.827 1.000 1.000
Chain 1: 200 -16128.388 1.534 2.067
Chain 1: 300 -8781.812 1.301 1.000
Chain 1: 400 -8501.917 0.984 1.000
Chain 1: 500 -8249.772 0.793 0.837
Chain 1: 600 -8232.872 0.662 0.837
Chain 1: 700 -8134.364 0.569 0.033
Chain 1: 800 -8076.364 0.499 0.033
Chain 1: 900 -7737.652 0.448 0.033
Chain 1: 1000 -7849.753 0.405 0.033
Chain 1: 1100 -7797.023 0.305 0.031
Chain 1: 1200 -7671.204 0.100 0.016
Chain 1: 1300 -7816.903 0.018 0.016
Chain 1: 1400 -7706.316 0.017 0.014
Chain 1: 1500 -7565.713 0.015 0.014
Chain 1: 1600 -7833.842 0.019 0.016
Chain 1: 1700 -7427.899 0.023 0.019
Chain 1: 1800 -7645.019 0.025 0.019
Chain 1: 1900 -7508.290 0.022 0.019
Chain 1: 2000 -7631.934 0.023 0.019
Chain 1: 2100 -7618.590 0.022 0.019
Chain 1: 2200 -7702.694 0.022 0.019
Chain 1: 2300 -7593.831 0.021 0.018
Chain 1: 2400 -7634.172 0.020 0.018
Chain 1: 2500 -7611.298 0.019 0.016
Chain 1: 2600 -7525.425 0.016 0.014
Chain 1: 2700 -7583.570 0.012 0.011
Chain 1: 2800 -7499.927 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002952 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.52 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86766.708 1.000 1.000
Chain 1: 200 -13649.959 3.178 5.357
Chain 1: 300 -10015.571 2.240 1.000
Chain 1: 400 -10785.942 1.698 1.000
Chain 1: 500 -8985.822 1.398 0.363
Chain 1: 600 -8495.771 1.175 0.363
Chain 1: 700 -8719.479 1.011 0.200
Chain 1: 800 -9333.818 0.893 0.200
Chain 1: 900 -8821.477 0.800 0.071
Chain 1: 1000 -8664.129 0.722 0.071
Chain 1: 1100 -8832.981 0.624 0.066 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8526.908 0.092 0.058
Chain 1: 1300 -8670.127 0.057 0.058
Chain 1: 1400 -8707.595 0.050 0.036
Chain 1: 1500 -8581.932 0.032 0.026
Chain 1: 1600 -8690.189 0.027 0.019
Chain 1: 1700 -8780.184 0.026 0.018
Chain 1: 1800 -8368.361 0.024 0.018
Chain 1: 1900 -8464.327 0.019 0.017
Chain 1: 2000 -8437.326 0.018 0.015
Chain 1: 2100 -8559.430 0.017 0.014
Chain 1: 2200 -8378.175 0.016 0.014
Chain 1: 2300 -8459.805 0.015 0.012
Chain 1: 2400 -8529.087 0.015 0.012
Chain 1: 2500 -8474.393 0.015 0.011
Chain 1: 2600 -8473.429 0.013 0.010
Chain 1: 2700 -8390.712 0.013 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003126 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.26 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8406311.594 1.000 1.000
Chain 1: 200 -1585388.105 2.651 4.302
Chain 1: 300 -891732.505 2.027 1.000
Chain 1: 400 -457803.342 1.757 1.000
Chain 1: 500 -357942.168 1.461 0.948
Chain 1: 600 -232914.818 1.307 0.948
Chain 1: 700 -119301.475 1.257 0.948
Chain 1: 800 -86503.229 1.147 0.948
Chain 1: 900 -66866.600 1.052 0.778
Chain 1: 1000 -51676.350 0.976 0.778
Chain 1: 1100 -39165.918 0.908 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38345.865 0.480 0.379
Chain 1: 1300 -26321.590 0.448 0.379
Chain 1: 1400 -26042.041 0.354 0.319
Chain 1: 1500 -22633.258 0.341 0.319
Chain 1: 1600 -21850.841 0.291 0.294
Chain 1: 1700 -20727.260 0.202 0.294
Chain 1: 1800 -20672.091 0.164 0.151
Chain 1: 1900 -20998.088 0.136 0.054
Chain 1: 2000 -19510.632 0.114 0.054
Chain 1: 2100 -19749.139 0.084 0.036
Chain 1: 2200 -19975.095 0.083 0.036
Chain 1: 2300 -19592.755 0.039 0.020
Chain 1: 2400 -19364.892 0.039 0.020
Chain 1: 2500 -19166.655 0.025 0.016
Chain 1: 2600 -18797.152 0.023 0.016
Chain 1: 2700 -18754.283 0.018 0.012
Chain 1: 2800 -18470.990 0.019 0.015
Chain 1: 2900 -18752.224 0.019 0.015
Chain 1: 3000 -18738.477 0.012 0.012
Chain 1: 3100 -18823.419 0.011 0.012
Chain 1: 3200 -18514.211 0.012 0.015
Chain 1: 3300 -18718.886 0.011 0.012
Chain 1: 3400 -18193.874 0.012 0.015
Chain 1: 3500 -18805.527 0.015 0.015
Chain 1: 3600 -18112.545 0.017 0.015
Chain 1: 3700 -18499.042 0.018 0.017
Chain 1: 3800 -17459.169 0.023 0.021
Chain 1: 3900 -17455.300 0.021 0.021
Chain 1: 4000 -17572.640 0.022 0.021
Chain 1: 4100 -17486.322 0.022 0.021
Chain 1: 4200 -17302.744 0.021 0.021
Chain 1: 4300 -17441.071 0.021 0.021
Chain 1: 4400 -17397.993 0.018 0.011
Chain 1: 4500 -17300.492 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001284 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.84 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13324.672 1.000 1.000
Chain 1: 200 -10237.443 0.651 1.000
Chain 1: 300 -8753.922 0.490 0.302
Chain 1: 400 -8978.015 0.374 0.302
Chain 1: 500 -8621.095 0.307 0.169
Chain 1: 600 -8673.701 0.257 0.169
Chain 1: 700 -8568.860 0.222 0.041
Chain 1: 800 -8490.122 0.196 0.041
Chain 1: 900 -8608.004 0.175 0.025
Chain 1: 1000 -8628.191 0.158 0.025
Chain 1: 1100 -8716.925 0.059 0.014
Chain 1: 1200 -8590.311 0.030 0.014
Chain 1: 1300 -8523.611 0.014 0.012
Chain 1: 1400 -8540.405 0.012 0.010
Chain 1: 1500 -8661.855 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001378 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.78 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -63316.678 1.000 1.000
Chain 1: 200 -19015.534 1.665 2.330
Chain 1: 300 -9496.626 1.444 1.002
Chain 1: 400 -8951.925 1.098 1.002
Chain 1: 500 -9239.104 0.885 1.000
Chain 1: 600 -9473.893 0.741 1.000
Chain 1: 700 -8173.857 0.658 0.159
Chain 1: 800 -8482.843 0.581 0.159
Chain 1: 900 -8141.550 0.521 0.061
Chain 1: 1000 -7816.503 0.473 0.061
Chain 1: 1100 -8164.792 0.377 0.043
Chain 1: 1200 -7913.250 0.147 0.042
Chain 1: 1300 -8190.913 0.050 0.042
Chain 1: 1400 -7952.979 0.047 0.036
Chain 1: 1500 -7649.592 0.048 0.040
Chain 1: 1600 -7897.893 0.049 0.040
Chain 1: 1700 -7901.567 0.033 0.036
Chain 1: 1800 -7930.516 0.030 0.034
Chain 1: 1900 -7771.538 0.028 0.032
Chain 1: 2000 -7955.503 0.026 0.031
Chain 1: 2100 -7778.409 0.024 0.030
Chain 1: 2200 -7976.642 0.023 0.025
Chain 1: 2300 -7738.508 0.023 0.025
Chain 1: 2400 -7860.617 0.021 0.023
Chain 1: 2500 -7753.711 0.019 0.023
Chain 1: 2600 -7674.550 0.017 0.020
Chain 1: 2700 -7666.249 0.017 0.020
Chain 1: 2800 -7698.610 0.017 0.020
Chain 1: 2900 -7522.068 0.017 0.023
Chain 1: 3000 -7683.800 0.017 0.021
Chain 1: 3100 -7648.521 0.015 0.016
Chain 1: 3200 -7754.597 0.014 0.014
Chain 1: 3300 -7590.732 0.013 0.014
Chain 1: 3400 -7823.482 0.014 0.014
Chain 1: 3500 -7603.923 0.016 0.021
Chain 1: 3600 -7630.394 0.015 0.021
Chain 1: 3700 -7547.337 0.016 0.021
Chain 1: 3800 -7636.549 0.017 0.021
Chain 1: 3900 -7537.167 0.016 0.014
Chain 1: 4000 -7521.139 0.014 0.013
Chain 1: 4100 -7530.369 0.014 0.013
Chain 1: 4200 -7670.273 0.014 0.013
Chain 1: 4300 -7510.465 0.014 0.013
Chain 1: 4400 -7562.119 0.012 0.012
Chain 1: 4500 -7713.760 0.011 0.012
Chain 1: 4600 -7591.066 0.012 0.013
Chain 1: 4700 -7584.229 0.011 0.013
Chain 1: 4800 -7539.963 0.011 0.013
Chain 1: 4900 -7660.488 0.011 0.016
Chain 1: 5000 -7720.182 0.011 0.016
Chain 1: 5100 -7617.214 0.013 0.016
Chain 1: 5200 -7637.028 0.011 0.014
Chain 1: 5300 -7609.568 0.009 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002614 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 26.14 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87023.580 1.000 1.000
Chain 1: 200 -14569.007 2.987 4.973
Chain 1: 300 -10778.060 2.108 1.000
Chain 1: 400 -12630.945 1.618 1.000
Chain 1: 500 -9166.328 1.370 0.378
Chain 1: 600 -9020.656 1.144 0.378
Chain 1: 700 -9397.667 0.987 0.352
Chain 1: 800 -9307.187 0.864 0.352
Chain 1: 900 -9609.160 0.772 0.147
Chain 1: 1000 -9012.119 0.701 0.147
Chain 1: 1100 -9543.307 0.607 0.066 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -9010.275 0.115 0.059
Chain 1: 1300 -9353.374 0.084 0.056
Chain 1: 1400 -9231.656 0.071 0.040
Chain 1: 1500 -9226.591 0.033 0.037
Chain 1: 1600 -9303.735 0.032 0.037
Chain 1: 1700 -9356.357 0.029 0.031
Chain 1: 1800 -8903.203 0.033 0.037
Chain 1: 1900 -9011.497 0.031 0.037
Chain 1: 2000 -9032.846 0.024 0.013
Chain 1: 2100 -9120.997 0.020 0.012
Chain 1: 2200 -8896.327 0.016 0.012
Chain 1: 2300 -9101.223 0.015 0.012
Chain 1: 2400 -8909.920 0.016 0.012
Chain 1: 2500 -8982.246 0.017 0.012
Chain 1: 2600 -8891.757 0.017 0.012
Chain 1: 2700 -8925.641 0.017 0.012
Chain 1: 2800 -8876.811 0.012 0.010
Chain 1: 2900 -8991.687 0.012 0.010
Chain 1: 3000 -8900.944 0.013 0.010
Chain 1: 3100 -8868.181 0.012 0.010
Chain 1: 3200 -8839.178 0.010 0.010
Chain 1: 3300 -9103.049 0.011 0.010
Chain 1: 3400 -9149.928 0.009 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002898 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 28.98 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8390340.605 1.000 1.000
Chain 1: 200 -1580053.726 2.655 4.310
Chain 1: 300 -890310.063 2.028 1.000
Chain 1: 400 -458273.124 1.757 1.000
Chain 1: 500 -359337.190 1.461 0.943
Chain 1: 600 -234367.602 1.306 0.943
Chain 1: 700 -120500.889 1.254 0.943
Chain 1: 800 -87727.085 1.144 0.943
Chain 1: 900 -68032.991 1.049 0.775
Chain 1: 1000 -52815.538 0.973 0.775
Chain 1: 1100 -40266.749 0.904 0.533 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39447.506 0.475 0.374
Chain 1: 1300 -27353.460 0.442 0.374
Chain 1: 1400 -27072.173 0.349 0.312
Chain 1: 1500 -23647.074 0.336 0.312
Chain 1: 1600 -22861.693 0.286 0.289
Chain 1: 1700 -21727.795 0.197 0.288
Chain 1: 1800 -21671.022 0.160 0.145
Chain 1: 1900 -21998.132 0.132 0.052
Chain 1: 2000 -20503.974 0.111 0.052
Chain 1: 2100 -20742.432 0.081 0.034
Chain 1: 2200 -20970.394 0.080 0.034
Chain 1: 2300 -20586.098 0.037 0.019
Chain 1: 2400 -20357.793 0.037 0.019
Chain 1: 2500 -20160.189 0.024 0.015
Chain 1: 2600 -19789.045 0.022 0.015
Chain 1: 2700 -19745.614 0.017 0.011
Chain 1: 2800 -19462.252 0.019 0.015
Chain 1: 2900 -19744.022 0.018 0.014
Chain 1: 3000 -19729.951 0.011 0.011
Chain 1: 3100 -19815.150 0.011 0.011
Chain 1: 3200 -19505.117 0.011 0.014
Chain 1: 3300 -19710.418 0.010 0.011
Chain 1: 3400 -19184.234 0.012 0.014
Chain 1: 3500 -19797.906 0.014 0.015
Chain 1: 3600 -19102.246 0.016 0.015
Chain 1: 3700 -19490.866 0.017 0.016
Chain 1: 3800 -18447.042 0.022 0.020
Chain 1: 3900 -18443.161 0.020 0.020
Chain 1: 4000 -18560.400 0.021 0.020
Chain 1: 4100 -18474.053 0.021 0.020
Chain 1: 4200 -18289.497 0.020 0.020
Chain 1: 4300 -18428.400 0.020 0.020
Chain 1: 4400 -18384.574 0.018 0.010
Chain 1: 4500 -18287.052 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001176 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 11.76 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49341.377 1.000 1.000
Chain 1: 200 -107731.999 0.771 1.000
Chain 1: 300 -22124.714 1.804 1.000
Chain 1: 400 -12735.012 1.537 1.000
Chain 1: 500 -16246.573 1.273 0.737
Chain 1: 600 -17756.585 1.075 0.737
Chain 1: 700 -14835.583 0.950 0.542
Chain 1: 800 -28785.445 0.891 0.542
Chain 1: 900 -11517.784 0.959 0.542
Chain 1: 1000 -10975.550 0.868 0.542
Chain 1: 1100 -10446.769 0.773 0.485 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -12360.233 0.734 0.216 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1300 -12384.499 0.348 0.197
Chain 1: 1400 -19667.836 0.311 0.197
Chain 1: 1500 -11779.191 0.356 0.197
Chain 1: 1600 -12972.415 0.357 0.197
Chain 1: 1700 -13991.143 0.345 0.155
Chain 1: 1800 -18132.766 0.319 0.155
Chain 1: 1900 -16760.477 0.177 0.092
Chain 1: 2000 -11192.044 0.222 0.155
Chain 1: 2100 -10876.928 0.220 0.155
Chain 1: 2200 -9928.831 0.214 0.095
Chain 1: 2300 -14401.095 0.245 0.228
Chain 1: 2400 -9968.190 0.252 0.228
Chain 1: 2500 -10682.297 0.192 0.095
Chain 1: 2600 -12039.823 0.194 0.113
Chain 1: 2700 -10519.283 0.201 0.145
Chain 1: 2800 -12288.350 0.193 0.144
Chain 1: 2900 -17251.918 0.213 0.145
Chain 1: 3000 -9243.178 0.250 0.145
Chain 1: 3100 -9054.894 0.249 0.145
Chain 1: 3200 -15196.349 0.280 0.288
Chain 1: 3300 -14639.525 0.253 0.145
Chain 1: 3400 -13404.632 0.218 0.144
Chain 1: 3500 -9593.813 0.251 0.145
Chain 1: 3600 -9567.574 0.240 0.145
Chain 1: 3700 -8944.988 0.232 0.144
Chain 1: 3800 -9615.574 0.225 0.092
Chain 1: 3900 -9636.865 0.196 0.070
Chain 1: 4000 -12502.489 0.133 0.070
Chain 1: 4100 -15007.557 0.147 0.092
Chain 1: 4200 -9647.366 0.162 0.092
Chain 1: 4300 -8952.557 0.166 0.092
Chain 1: 4400 -12383.116 0.185 0.167
Chain 1: 4500 -9561.454 0.175 0.167
Chain 1: 4600 -9392.220 0.176 0.167
Chain 1: 4700 -9254.006 0.171 0.167
Chain 1: 4800 -9108.107 0.165 0.167
Chain 1: 4900 -8903.565 0.167 0.167
Chain 1: 5000 -8963.611 0.145 0.078
Chain 1: 5100 -9356.853 0.133 0.042
Chain 1: 5200 -9049.725 0.080 0.034
Chain 1: 5300 -14496.276 0.110 0.034
Chain 1: 5400 -8861.936 0.146 0.034
Chain 1: 5500 -11523.124 0.140 0.034
Chain 1: 5600 -8614.900 0.172 0.042
Chain 1: 5700 -9635.786 0.181 0.106
Chain 1: 5800 -12960.894 0.205 0.231
Chain 1: 5900 -8833.529 0.249 0.257
Chain 1: 6000 -9240.494 0.253 0.257
Chain 1: 6100 -9229.581 0.249 0.257
Chain 1: 6200 -8617.964 0.253 0.257
Chain 1: 6300 -11987.956 0.243 0.257
Chain 1: 6400 -14664.899 0.198 0.231
Chain 1: 6500 -9664.649 0.226 0.257
Chain 1: 6600 -9272.927 0.197 0.183
Chain 1: 6700 -10808.170 0.201 0.183
Chain 1: 6800 -8729.711 0.199 0.183
Chain 1: 6900 -13520.727 0.187 0.183
Chain 1: 7000 -8494.245 0.242 0.238
Chain 1: 7100 -8515.489 0.242 0.238
Chain 1: 7200 -8806.848 0.239 0.238
Chain 1: 7300 -8606.897 0.213 0.183
Chain 1: 7400 -8479.234 0.196 0.142
Chain 1: 7500 -8501.354 0.144 0.042
Chain 1: 7600 -8833.684 0.144 0.038
Chain 1: 7700 -8802.394 0.130 0.033
Chain 1: 7800 -8501.982 0.110 0.033
Chain 1: 7900 -9071.083 0.081 0.033
Chain 1: 8000 -8597.489 0.027 0.033
Chain 1: 8100 -9101.091 0.032 0.035
Chain 1: 8200 -12333.576 0.055 0.038
Chain 1: 8300 -13021.014 0.058 0.053
Chain 1: 8400 -11154.493 0.073 0.055
Chain 1: 8500 -8460.656 0.105 0.055
Chain 1: 8600 -9503.648 0.112 0.063
Chain 1: 8700 -9597.683 0.113 0.063
Chain 1: 8800 -9811.872 0.112 0.063
Chain 1: 8900 -12478.760 0.127 0.110
Chain 1: 9000 -11372.744 0.131 0.110
Chain 1: 9100 -8787.277 0.155 0.167
Chain 1: 9200 -9179.904 0.133 0.110
Chain 1: 9300 -8637.798 0.134 0.110
Chain 1: 9400 -8625.746 0.117 0.097
Chain 1: 9500 -9713.234 0.097 0.097
Chain 1: 9600 -12194.046 0.106 0.097
Chain 1: 9700 -10456.043 0.122 0.112
Chain 1: 9800 -8236.096 0.146 0.166
Chain 1: 9900 -8973.334 0.133 0.112
Chain 1: 10000 -10629.182 0.139 0.156
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001412 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.12 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58758.476 1.000 1.000
Chain 1: 200 -18260.045 1.609 2.218
Chain 1: 300 -8950.893 1.419 1.040
Chain 1: 400 -8175.249 1.088 1.040
Chain 1: 500 -8246.333 0.872 1.000
Chain 1: 600 -8128.693 0.729 1.000
Chain 1: 700 -7847.916 0.630 0.095
Chain 1: 800 -8094.249 0.555 0.095
Chain 1: 900 -7972.068 0.495 0.036
Chain 1: 1000 -7770.337 0.448 0.036
Chain 1: 1100 -7812.376 0.349 0.030
Chain 1: 1200 -7685.346 0.129 0.026
Chain 1: 1300 -7595.992 0.026 0.017
Chain 1: 1400 -8069.442 0.022 0.017
Chain 1: 1500 -7585.494 0.028 0.026
Chain 1: 1600 -7994.611 0.031 0.030
Chain 1: 1700 -7479.752 0.035 0.030
Chain 1: 1800 -7662.585 0.034 0.026
Chain 1: 1900 -7730.318 0.033 0.026
Chain 1: 2000 -7715.876 0.031 0.024
Chain 1: 2100 -7617.746 0.032 0.024
Chain 1: 2200 -7795.270 0.032 0.024
Chain 1: 2300 -7672.606 0.033 0.024
Chain 1: 2400 -7590.851 0.028 0.023
Chain 1: 2500 -7629.612 0.022 0.016
Chain 1: 2600 -7585.076 0.018 0.013
Chain 1: 2700 -7572.802 0.011 0.011
Chain 1: 2800 -7715.324 0.010 0.011
Chain 1: 2900 -7428.282 0.013 0.013
Chain 1: 3000 -7582.480 0.015 0.016
Chain 1: 3100 -7576.412 0.014 0.016
Chain 1: 3200 -7798.406 0.015 0.016
Chain 1: 3300 -7507.141 0.017 0.018
Chain 1: 3400 -7750.799 0.019 0.020
Chain 1: 3500 -7489.872 0.022 0.028
Chain 1: 3600 -7552.730 0.022 0.028
Chain 1: 3700 -7505.449 0.023 0.028
Chain 1: 3800 -7499.000 0.021 0.028
Chain 1: 3900 -7460.443 0.018 0.020
Chain 1: 4000 -7453.995 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002563 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.63 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86717.337 1.000 1.000
Chain 1: 200 -13995.946 3.098 5.196
Chain 1: 300 -10212.143 2.189 1.000
Chain 1: 400 -11831.957 1.676 1.000
Chain 1: 500 -8952.994 1.405 0.371
Chain 1: 600 -9934.234 1.187 0.371
Chain 1: 700 -9081.169 1.031 0.322
Chain 1: 800 -8381.889 0.913 0.322
Chain 1: 900 -8443.166 0.812 0.137
Chain 1: 1000 -8719.320 0.734 0.137
Chain 1: 1100 -8925.494 0.636 0.099 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8452.938 0.122 0.094
Chain 1: 1300 -8777.304 0.089 0.083
Chain 1: 1400 -8732.731 0.076 0.056
Chain 1: 1500 -8658.239 0.044 0.037
Chain 1: 1600 -8725.550 0.035 0.032
Chain 1: 1700 -8810.832 0.027 0.023
Chain 1: 1800 -8366.064 0.024 0.023
Chain 1: 1900 -8466.302 0.024 0.023
Chain 1: 2000 -8482.850 0.021 0.012
Chain 1: 2100 -8569.923 0.020 0.010
Chain 1: 2200 -8353.928 0.017 0.010
Chain 1: 2300 -8514.842 0.015 0.010
Chain 1: 2400 -8362.909 0.017 0.012
Chain 1: 2500 -8436.490 0.017 0.012
Chain 1: 2600 -8347.129 0.017 0.012
Chain 1: 2700 -8381.294 0.016 0.012
Chain 1: 2800 -8332.503 0.012 0.011
Chain 1: 2900 -8447.233 0.012 0.011
Chain 1: 3000 -8360.248 0.013 0.011
Chain 1: 3100 -8324.688 0.012 0.011
Chain 1: 3200 -8296.534 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002863 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 28.63 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8401921.544 1.000 1.000
Chain 1: 200 -1584318.652 2.652 4.303
Chain 1: 300 -890100.257 2.028 1.000
Chain 1: 400 -457538.625 1.757 1.000
Chain 1: 500 -358186.883 1.461 0.945
Chain 1: 600 -233359.289 1.307 0.945
Chain 1: 700 -119714.849 1.256 0.945
Chain 1: 800 -86948.546 1.146 0.945
Chain 1: 900 -67314.511 1.051 0.780
Chain 1: 1000 -52132.237 0.975 0.780
Chain 1: 1100 -39615.377 0.907 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38800.957 0.478 0.377
Chain 1: 1300 -26748.985 0.445 0.377
Chain 1: 1400 -26471.314 0.352 0.316
Chain 1: 1500 -23055.371 0.339 0.316
Chain 1: 1600 -22272.109 0.289 0.292
Chain 1: 1700 -21143.899 0.199 0.291
Chain 1: 1800 -21088.255 0.162 0.148
Chain 1: 1900 -21415.261 0.134 0.053
Chain 1: 2000 -19923.649 0.113 0.053
Chain 1: 2100 -20162.371 0.082 0.035
Chain 1: 2200 -20389.511 0.081 0.035
Chain 1: 2300 -20005.865 0.038 0.019
Chain 1: 2400 -19777.618 0.038 0.019
Chain 1: 2500 -19579.602 0.025 0.015
Chain 1: 2600 -19209.030 0.023 0.015
Chain 1: 2700 -19165.725 0.018 0.012
Chain 1: 2800 -18882.179 0.019 0.015
Chain 1: 2900 -19163.846 0.019 0.015
Chain 1: 3000 -19149.982 0.012 0.012
Chain 1: 3100 -19235.115 0.011 0.012
Chain 1: 3200 -18925.249 0.011 0.015
Chain 1: 3300 -19130.391 0.011 0.012
Chain 1: 3400 -18604.319 0.012 0.015
Chain 1: 3500 -19217.715 0.014 0.015
Chain 1: 3600 -18522.383 0.016 0.015
Chain 1: 3700 -18910.697 0.018 0.016
Chain 1: 3800 -17867.299 0.022 0.021
Chain 1: 3900 -17863.341 0.021 0.021
Chain 1: 4000 -17980.664 0.021 0.021
Chain 1: 4100 -17894.273 0.022 0.021
Chain 1: 4200 -17709.817 0.021 0.021
Chain 1: 4300 -17848.728 0.021 0.021
Chain 1: 4400 -17805.001 0.018 0.010
Chain 1: 4500 -17707.402 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001286 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.86 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49456.226 1.000 1.000
Chain 1: 200 -16886.018 1.464 1.929
Chain 1: 300 -23797.136 1.073 1.000
Chain 1: 400 -17968.976 0.886 1.000
Chain 1: 500 -12326.036 0.800 0.458
Chain 1: 600 -15996.496 0.705 0.458
Chain 1: 700 -14242.533 0.622 0.324
Chain 1: 800 -12565.521 0.561 0.324
Chain 1: 900 -22307.035 0.547 0.324
Chain 1: 1000 -11605.882 0.585 0.437
Chain 1: 1100 -15295.908 0.509 0.324 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -12158.429 0.342 0.290
Chain 1: 1300 -13397.378 0.322 0.258
Chain 1: 1400 -10887.521 0.312 0.241
Chain 1: 1500 -10877.751 0.267 0.231
Chain 1: 1600 -13482.944 0.263 0.231
Chain 1: 1700 -10739.480 0.276 0.241
Chain 1: 1800 -13016.461 0.281 0.241
Chain 1: 1900 -10523.915 0.261 0.237
Chain 1: 2000 -11130.860 0.174 0.231
Chain 1: 2100 -9723.919 0.164 0.193
Chain 1: 2200 -10569.479 0.146 0.175
Chain 1: 2300 -15546.019 0.169 0.193
Chain 1: 2400 -10472.899 0.195 0.193
Chain 1: 2500 -12405.954 0.210 0.193
Chain 1: 2600 -10238.880 0.212 0.212
Chain 1: 2700 -17352.118 0.227 0.212
Chain 1: 2800 -13820.727 0.235 0.237
Chain 1: 2900 -9483.282 0.257 0.256
Chain 1: 3000 -9320.492 0.254 0.256
Chain 1: 3100 -9448.495 0.241 0.256
Chain 1: 3200 -9271.305 0.234 0.256
Chain 1: 3300 -15762.412 0.244 0.256
Chain 1: 3400 -10236.744 0.249 0.256
Chain 1: 3500 -14923.407 0.265 0.314
Chain 1: 3600 -9625.850 0.299 0.410
Chain 1: 3700 -9332.616 0.261 0.314
Chain 1: 3800 -16116.504 0.278 0.412
Chain 1: 3900 -10792.310 0.281 0.412
Chain 1: 4000 -11811.102 0.288 0.412
Chain 1: 4100 -9556.674 0.310 0.412
Chain 1: 4200 -14087.989 0.341 0.412
Chain 1: 4300 -17042.842 0.317 0.322
Chain 1: 4400 -9630.074 0.340 0.322
Chain 1: 4500 -9948.769 0.311 0.322
Chain 1: 4600 -13164.655 0.281 0.244
Chain 1: 4700 -10319.001 0.305 0.276
Chain 1: 4800 -8920.398 0.279 0.244
Chain 1: 4900 -9267.989 0.233 0.236
Chain 1: 5000 -10419.419 0.236 0.236
Chain 1: 5100 -9714.308 0.219 0.173
Chain 1: 5200 -10900.828 0.198 0.157
Chain 1: 5300 -12449.989 0.193 0.124
Chain 1: 5400 -9206.015 0.152 0.124
Chain 1: 5500 -14759.006 0.186 0.157
Chain 1: 5600 -13460.074 0.171 0.124
Chain 1: 5700 -9717.299 0.182 0.124
Chain 1: 5800 -9300.423 0.171 0.111
Chain 1: 5900 -10456.045 0.178 0.111
Chain 1: 6000 -8881.670 0.185 0.124
Chain 1: 6100 -9661.856 0.186 0.124
Chain 1: 6200 -8601.605 0.187 0.124
Chain 1: 6300 -9033.273 0.179 0.123
Chain 1: 6400 -9110.823 0.145 0.111
Chain 1: 6500 -9139.545 0.108 0.097
Chain 1: 6600 -8949.267 0.100 0.081
Chain 1: 6700 -8776.623 0.064 0.048
Chain 1: 6800 -8667.790 0.060 0.048
Chain 1: 6900 -12998.832 0.083 0.048
Chain 1: 7000 -9020.544 0.109 0.048
Chain 1: 7100 -10351.579 0.114 0.048
Chain 1: 7200 -9065.758 0.116 0.048
Chain 1: 7300 -11911.043 0.135 0.129
Chain 1: 7400 -13635.156 0.147 0.129
Chain 1: 7500 -8573.014 0.205 0.142
Chain 1: 7600 -9029.205 0.208 0.142
Chain 1: 7700 -9189.267 0.208 0.142
Chain 1: 7800 -9281.248 0.208 0.142
Chain 1: 7900 -8614.515 0.182 0.129
Chain 1: 8000 -11121.372 0.161 0.129
Chain 1: 8100 -9191.800 0.169 0.142
Chain 1: 8200 -9254.393 0.155 0.126
Chain 1: 8300 -8807.869 0.136 0.077
Chain 1: 8400 -11300.103 0.146 0.077
Chain 1: 8500 -10668.163 0.093 0.059
Chain 1: 8600 -10378.889 0.091 0.059
Chain 1: 8700 -8789.218 0.107 0.077
Chain 1: 8800 -8693.192 0.107 0.077
Chain 1: 8900 -12853.010 0.132 0.181
Chain 1: 9000 -9892.024 0.139 0.181
Chain 1: 9100 -8528.928 0.134 0.160
Chain 1: 9200 -8778.094 0.136 0.160
Chain 1: 9300 -8484.947 0.135 0.160
Chain 1: 9400 -8742.881 0.115 0.059
Chain 1: 9500 -9838.092 0.121 0.111
Chain 1: 9600 -9275.738 0.124 0.111
Chain 1: 9700 -9059.307 0.108 0.061
Chain 1: 9800 -8738.262 0.111 0.061
Chain 1: 9900 -9633.968 0.088 0.061
Chain 1: 10000 -8378.651 0.073 0.061
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001422 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.22 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -46614.724 1.000 1.000
Chain 1: 200 -15945.315 1.462 1.923
Chain 1: 300 -8910.428 1.238 1.000
Chain 1: 400 -8100.394 0.953 1.000
Chain 1: 500 -8777.180 0.778 0.790
Chain 1: 600 -8788.142 0.649 0.790
Chain 1: 700 -7927.196 0.571 0.109
Chain 1: 800 -8172.745 0.504 0.109
Chain 1: 900 -7687.769 0.455 0.100
Chain 1: 1000 -7972.994 0.413 0.100
Chain 1: 1100 -7777.410 0.315 0.077
Chain 1: 1200 -7780.482 0.123 0.063
Chain 1: 1300 -7726.454 0.045 0.036
Chain 1: 1400 -7928.307 0.037 0.030
Chain 1: 1500 -7546.017 0.035 0.030
Chain 1: 1600 -7656.775 0.036 0.030
Chain 1: 1700 -7564.090 0.026 0.025
Chain 1: 1800 -7544.871 0.024 0.025
Chain 1: 1900 -7515.965 0.018 0.014
Chain 1: 2000 -7672.372 0.016 0.014
Chain 1: 2100 -7442.907 0.017 0.014
Chain 1: 2200 -7866.174 0.022 0.020
Chain 1: 2300 -7526.418 0.026 0.025
Chain 1: 2400 -7497.128 0.024 0.020
Chain 1: 2500 -7590.769 0.020 0.014
Chain 1: 2600 -7497.170 0.020 0.012
Chain 1: 2700 -7480.528 0.019 0.012
Chain 1: 2800 -7495.312 0.019 0.012
Chain 1: 2900 -7342.940 0.020 0.020
Chain 1: 3000 -7493.635 0.020 0.020
Chain 1: 3100 -7500.568 0.017 0.012
Chain 1: 3200 -7715.363 0.015 0.012
Chain 1: 3300 -7411.603 0.014 0.012
Chain 1: 3400 -7672.665 0.017 0.020
Chain 1: 3500 -7410.311 0.020 0.021
Chain 1: 3600 -7469.294 0.019 0.021
Chain 1: 3700 -7424.937 0.020 0.021
Chain 1: 3800 -7424.926 0.019 0.021
Chain 1: 3900 -7376.704 0.018 0.020
Chain 1: 4000 -7370.540 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003016 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.16 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87123.427 1.000 1.000
Chain 1: 200 -14008.330 3.110 5.219
Chain 1: 300 -10279.182 2.194 1.000
Chain 1: 400 -11351.156 1.669 1.000
Chain 1: 500 -9220.913 1.382 0.363
Chain 1: 600 -9520.948 1.157 0.363
Chain 1: 700 -9248.868 0.996 0.231
Chain 1: 800 -8607.219 0.880 0.231
Chain 1: 900 -8634.964 0.783 0.094
Chain 1: 1000 -8757.826 0.706 0.094
Chain 1: 1100 -9074.540 0.610 0.075 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8629.396 0.093 0.052
Chain 1: 1300 -8949.621 0.060 0.036
Chain 1: 1400 -8940.345 0.051 0.035
Chain 1: 1500 -8784.469 0.029 0.032
Chain 1: 1600 -8893.298 0.027 0.029
Chain 1: 1700 -8955.522 0.025 0.018
Chain 1: 1800 -8517.327 0.023 0.018
Chain 1: 1900 -8621.767 0.024 0.018
Chain 1: 2000 -8601.191 0.023 0.018
Chain 1: 2100 -8739.807 0.021 0.016
Chain 1: 2200 -8522.570 0.018 0.016
Chain 1: 2300 -8688.709 0.016 0.016
Chain 1: 2400 -8523.672 0.018 0.018
Chain 1: 2500 -8594.477 0.017 0.016
Chain 1: 2600 -8506.697 0.017 0.016
Chain 1: 2700 -8540.817 0.017 0.016
Chain 1: 2800 -8499.350 0.012 0.012
Chain 1: 2900 -8595.660 0.012 0.011
Chain 1: 3000 -8434.618 0.014 0.016
Chain 1: 3100 -8583.537 0.014 0.017
Chain 1: 3200 -8454.565 0.013 0.015
Chain 1: 3300 -8466.797 0.011 0.011
Chain 1: 3400 -8642.981 0.011 0.011
Chain 1: 3500 -8646.133 0.010 0.011
Chain 1: 3600 -8410.559 0.012 0.015
Chain 1: 3700 -8559.006 0.014 0.017
Chain 1: 3800 -8416.352 0.015 0.017
Chain 1: 3900 -8350.044 0.014 0.017
Chain 1: 4000 -8433.244 0.013 0.017
Chain 1: 4100 -8422.168 0.012 0.015
Chain 1: 4200 -8407.436 0.011 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002963 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.63 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8396509.384 1.000 1.000
Chain 1: 200 -1587257.180 2.645 4.290
Chain 1: 300 -891930.785 2.023 1.000
Chain 1: 400 -458198.995 1.754 1.000
Chain 1: 500 -358404.129 1.459 0.947
Chain 1: 600 -233318.058 1.305 0.947
Chain 1: 700 -119646.312 1.254 0.947
Chain 1: 800 -86873.742 1.145 0.947
Chain 1: 900 -67247.215 1.050 0.780
Chain 1: 1000 -52072.584 0.974 0.780
Chain 1: 1100 -39566.208 0.906 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38750.402 0.479 0.377
Chain 1: 1300 -26717.457 0.446 0.377
Chain 1: 1400 -26439.955 0.352 0.316
Chain 1: 1500 -23028.800 0.339 0.316
Chain 1: 1600 -22246.300 0.289 0.292
Chain 1: 1700 -21121.123 0.200 0.291
Chain 1: 1800 -21065.805 0.162 0.148
Chain 1: 1900 -21392.422 0.134 0.053
Chain 1: 2000 -19902.986 0.113 0.053
Chain 1: 2100 -20141.622 0.082 0.035
Chain 1: 2200 -20368.185 0.081 0.035
Chain 1: 2300 -19985.130 0.038 0.019
Chain 1: 2400 -19757.035 0.038 0.019
Chain 1: 2500 -19558.876 0.024 0.015
Chain 1: 2600 -19188.805 0.023 0.015
Chain 1: 2700 -19145.702 0.018 0.012
Chain 1: 2800 -18862.253 0.019 0.015
Chain 1: 2900 -19143.714 0.019 0.015
Chain 1: 3000 -19129.941 0.012 0.012
Chain 1: 3100 -19214.969 0.011 0.012
Chain 1: 3200 -18905.404 0.011 0.015
Chain 1: 3300 -19110.322 0.011 0.012
Chain 1: 3400 -18584.688 0.012 0.015
Chain 1: 3500 -19197.363 0.014 0.015
Chain 1: 3600 -18503.014 0.016 0.015
Chain 1: 3700 -18890.541 0.018 0.016
Chain 1: 3800 -17848.604 0.022 0.021
Chain 1: 3900 -17844.669 0.021 0.021
Chain 1: 4000 -17962.018 0.021 0.021
Chain 1: 4100 -17875.657 0.022 0.021
Chain 1: 4200 -17691.554 0.021 0.021
Chain 1: 4300 -17830.233 0.021 0.021
Chain 1: 4400 -17786.763 0.018 0.010
Chain 1: 4500 -17689.199 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00129 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.9 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49023.417 1.000 1.000
Chain 1: 200 -20074.813 1.221 1.442
Chain 1: 300 -20947.815 0.828 1.000
Chain 1: 400 -17195.229 0.675 1.000
Chain 1: 500 -22122.248 0.585 0.223
Chain 1: 600 -13878.875 0.586 0.594
Chain 1: 700 -11644.765 0.530 0.223
Chain 1: 800 -10860.683 0.473 0.223
Chain 1: 900 -13144.705 0.440 0.218
Chain 1: 1000 -13495.691 0.398 0.218
Chain 1: 1100 -10402.703 0.328 0.218
Chain 1: 1200 -18291.503 0.227 0.218
Chain 1: 1300 -11107.040 0.287 0.223
Chain 1: 1400 -16678.395 0.299 0.297
Chain 1: 1500 -11083.702 0.327 0.334
Chain 1: 1600 -9606.481 0.283 0.297
Chain 1: 1700 -20768.797 0.318 0.334
Chain 1: 1800 -9178.502 0.437 0.431
Chain 1: 1900 -9618.068 0.424 0.431
Chain 1: 2000 -10388.185 0.429 0.431
Chain 1: 2100 -20027.635 0.447 0.481
Chain 1: 2200 -9751.218 0.509 0.505 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2300 -13018.555 0.470 0.481
Chain 1: 2400 -8889.546 0.483 0.481
Chain 1: 2500 -9926.167 0.443 0.464
Chain 1: 2600 -9184.693 0.436 0.464
Chain 1: 2700 -8920.313 0.385 0.251
Chain 1: 2800 -10582.693 0.274 0.157
Chain 1: 2900 -11728.678 0.279 0.157
Chain 1: 3000 -9882.993 0.291 0.187
Chain 1: 3100 -9806.343 0.243 0.157
Chain 1: 3200 -8649.260 0.151 0.134
Chain 1: 3300 -15217.960 0.169 0.134
Chain 1: 3400 -10325.151 0.170 0.134
Chain 1: 3500 -9391.032 0.170 0.134
Chain 1: 3600 -8794.559 0.169 0.134
Chain 1: 3700 -8608.621 0.168 0.134
Chain 1: 3800 -8796.353 0.154 0.099
Chain 1: 3900 -9539.575 0.152 0.099
Chain 1: 4000 -10007.573 0.138 0.078
Chain 1: 4100 -8692.349 0.153 0.099
Chain 1: 4200 -11765.762 0.165 0.099
Chain 1: 4300 -12455.309 0.128 0.078
Chain 1: 4400 -16617.703 0.105 0.078
Chain 1: 4500 -18510.601 0.106 0.078
Chain 1: 4600 -13320.069 0.138 0.102
Chain 1: 4700 -12543.555 0.142 0.102
Chain 1: 4800 -8534.713 0.187 0.151
Chain 1: 4900 -8535.938 0.179 0.151
Chain 1: 5000 -12122.717 0.204 0.250
Chain 1: 5100 -8528.820 0.231 0.261
Chain 1: 5200 -8902.352 0.209 0.250
Chain 1: 5300 -9178.564 0.206 0.250
Chain 1: 5400 -8565.899 0.188 0.102
Chain 1: 5500 -8284.754 0.182 0.072
Chain 1: 5600 -8233.368 0.143 0.062
Chain 1: 5700 -12765.924 0.173 0.072
Chain 1: 5800 -9242.966 0.164 0.072
Chain 1: 5900 -9038.704 0.166 0.072
Chain 1: 6000 -8506.671 0.143 0.063
Chain 1: 6100 -9782.942 0.114 0.063
Chain 1: 6200 -8106.877 0.130 0.072
Chain 1: 6300 -8402.848 0.131 0.072
Chain 1: 6400 -10690.997 0.145 0.130
Chain 1: 6500 -8414.716 0.168 0.207
Chain 1: 6600 -8565.561 0.170 0.207
Chain 1: 6700 -8138.659 0.139 0.130
Chain 1: 6800 -8740.929 0.108 0.069
Chain 1: 6900 -8943.732 0.108 0.069
Chain 1: 7000 -10530.637 0.117 0.130
Chain 1: 7100 -8174.689 0.133 0.151
Chain 1: 7200 -8453.987 0.115 0.069
Chain 1: 7300 -8518.973 0.113 0.069
Chain 1: 7400 -8317.369 0.094 0.052
Chain 1: 7500 -11909.706 0.097 0.052
Chain 1: 7600 -9509.561 0.120 0.069
Chain 1: 7700 -8182.572 0.131 0.151
Chain 1: 7800 -14281.913 0.167 0.162
Chain 1: 7900 -8290.352 0.237 0.252
Chain 1: 8000 -8755.728 0.227 0.252
Chain 1: 8100 -8066.377 0.207 0.162
Chain 1: 8200 -9919.383 0.222 0.187
Chain 1: 8300 -9395.366 0.227 0.187
Chain 1: 8400 -8455.371 0.236 0.187
Chain 1: 8500 -10542.553 0.225 0.187
Chain 1: 8600 -11924.375 0.212 0.162
Chain 1: 8700 -10095.021 0.214 0.181
Chain 1: 8800 -8305.943 0.193 0.181
Chain 1: 8900 -8958.416 0.128 0.116
Chain 1: 9000 -8844.979 0.124 0.116
Chain 1: 9100 -8115.604 0.124 0.116
Chain 1: 9200 -8320.143 0.108 0.111
Chain 1: 9300 -8684.192 0.106 0.111
Chain 1: 9400 -9181.999 0.101 0.090
Chain 1: 9500 -8159.629 0.093 0.090
Chain 1: 9600 -8163.323 0.082 0.073
Chain 1: 9700 -8857.051 0.072 0.073
Chain 1: 9800 -8583.061 0.053 0.054
Chain 1: 9900 -9923.875 0.059 0.054
Chain 1: 10000 -7904.366 0.084 0.078
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001368 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.68 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57394.483 1.000 1.000
Chain 1: 200 -17128.719 1.675 2.351
Chain 1: 300 -8672.174 1.442 1.000
Chain 1: 400 -8519.334 1.086 1.000
Chain 1: 500 -8380.517 0.872 0.975
Chain 1: 600 -8443.727 0.728 0.975
Chain 1: 700 -7777.762 0.636 0.086
Chain 1: 800 -8301.436 0.565 0.086
Chain 1: 900 -7991.529 0.506 0.063
Chain 1: 1000 -7840.458 0.457 0.063
Chain 1: 1100 -7759.179 0.359 0.039
Chain 1: 1200 -7659.263 0.125 0.019
Chain 1: 1300 -7683.878 0.028 0.018
Chain 1: 1400 -7956.707 0.029 0.019
Chain 1: 1500 -7641.268 0.032 0.034
Chain 1: 1600 -7650.753 0.031 0.034
Chain 1: 1700 -7531.274 0.024 0.019
Chain 1: 1800 -7605.680 0.019 0.016
Chain 1: 1900 -7578.882 0.015 0.013
Chain 1: 2000 -7661.270 0.014 0.011
Chain 1: 2100 -7628.303 0.014 0.011
Chain 1: 2200 -7709.591 0.013 0.011
Chain 1: 2300 -7624.738 0.014 0.011
Chain 1: 2400 -7663.230 0.011 0.011
Chain 1: 2500 -7590.912 0.008 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002993 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.93 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86532.557 1.000 1.000
Chain 1: 200 -13348.374 3.241 5.483
Chain 1: 300 -9709.795 2.286 1.000
Chain 1: 400 -10808.867 1.740 1.000
Chain 1: 500 -8504.189 1.446 0.375
Chain 1: 600 -8244.897 1.210 0.375
Chain 1: 700 -8381.325 1.040 0.271
Chain 1: 800 -8489.975 0.911 0.271
Chain 1: 900 -8484.936 0.810 0.102
Chain 1: 1000 -8427.667 0.730 0.102
Chain 1: 1100 -8554.530 0.631 0.031 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8161.824 0.088 0.031
Chain 1: 1300 -8388.388 0.053 0.027
Chain 1: 1400 -8406.815 0.043 0.016
Chain 1: 1500 -8254.327 0.018 0.016
Chain 1: 1600 -8368.456 0.016 0.015
Chain 1: 1700 -8447.527 0.015 0.014
Chain 1: 1800 -8028.936 0.019 0.015
Chain 1: 1900 -8127.659 0.020 0.015
Chain 1: 2000 -8101.535 0.020 0.015
Chain 1: 2100 -8225.662 0.020 0.015
Chain 1: 2200 -8038.477 0.018 0.015
Chain 1: 2300 -8122.230 0.016 0.014
Chain 1: 2400 -8191.696 0.017 0.014
Chain 1: 2500 -8137.617 0.015 0.012
Chain 1: 2600 -8137.995 0.014 0.010
Chain 1: 2700 -8055.168 0.014 0.010
Chain 1: 2800 -8016.628 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003076 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.76 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8453774.629 1.000 1.000
Chain 1: 200 -1595846.594 2.649 4.297
Chain 1: 300 -892486.803 2.028 1.000
Chain 1: 400 -457504.107 1.759 1.000
Chain 1: 500 -356840.964 1.464 0.951
Chain 1: 600 -231556.701 1.310 0.951
Chain 1: 700 -118340.162 1.259 0.951
Chain 1: 800 -85695.722 1.150 0.951
Chain 1: 900 -66176.514 1.055 0.788
Chain 1: 1000 -51108.308 0.979 0.788
Chain 1: 1100 -38703.107 0.911 0.541 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37895.477 0.483 0.381
Chain 1: 1300 -25969.435 0.450 0.381
Chain 1: 1400 -25699.928 0.356 0.321
Chain 1: 1500 -22317.644 0.343 0.321
Chain 1: 1600 -21543.300 0.293 0.295
Chain 1: 1700 -20431.218 0.202 0.295
Chain 1: 1800 -20378.688 0.165 0.152
Chain 1: 1900 -20704.902 0.137 0.054
Chain 1: 2000 -19223.254 0.115 0.054
Chain 1: 2100 -19461.248 0.084 0.036
Chain 1: 2200 -19686.562 0.083 0.036
Chain 1: 2300 -19304.780 0.039 0.020
Chain 1: 2400 -19077.020 0.039 0.020
Chain 1: 2500 -18878.480 0.025 0.016
Chain 1: 2600 -18509.161 0.024 0.016
Chain 1: 2700 -18466.338 0.018 0.012
Chain 1: 2800 -18182.933 0.020 0.016
Chain 1: 2900 -18464.020 0.020 0.015
Chain 1: 3000 -18450.370 0.012 0.012
Chain 1: 3100 -18535.327 0.011 0.012
Chain 1: 3200 -18226.110 0.012 0.015
Chain 1: 3300 -18430.767 0.011 0.012
Chain 1: 3400 -17905.633 0.013 0.015
Chain 1: 3500 -18517.367 0.015 0.016
Chain 1: 3600 -17824.233 0.017 0.016
Chain 1: 3700 -18210.760 0.019 0.017
Chain 1: 3800 -17170.625 0.023 0.021
Chain 1: 3900 -17166.708 0.022 0.021
Chain 1: 4000 -17284.093 0.022 0.021
Chain 1: 4100 -17197.811 0.022 0.021
Chain 1: 4200 -17014.134 0.022 0.021
Chain 1: 4300 -17152.539 0.021 0.021
Chain 1: 4400 -17109.381 0.019 0.011
Chain 1: 4500 -17011.873 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001139 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 11.39 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12721.449 1.000 1.000
Chain 1: 200 -9465.951 0.672 1.000
Chain 1: 300 -7992.849 0.509 0.344
Chain 1: 400 -8206.058 0.389 0.344
Chain 1: 500 -8209.951 0.311 0.184
Chain 1: 600 -7889.433 0.266 0.184
Chain 1: 700 -7817.160 0.229 0.041
Chain 1: 800 -7790.771 0.201 0.041
Chain 1: 900 -7893.732 0.180 0.026
Chain 1: 1000 -7867.970 0.162 0.026
Chain 1: 1100 -7857.680 0.063 0.013
Chain 1: 1200 -7803.440 0.029 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001519 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.19 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57679.007 1.000 1.000
Chain 1: 200 -17863.392 1.614 2.229
Chain 1: 300 -8929.314 1.410 1.001
Chain 1: 400 -8292.059 1.077 1.001
Chain 1: 500 -9221.800 0.881 1.000
Chain 1: 600 -8605.711 0.746 1.000
Chain 1: 700 -8483.092 0.642 0.101
Chain 1: 800 -8363.081 0.563 0.101
Chain 1: 900 -7989.936 0.506 0.077
Chain 1: 1000 -7972.214 0.456 0.077
Chain 1: 1100 -7581.570 0.361 0.072
Chain 1: 1200 -7667.208 0.139 0.052
Chain 1: 1300 -7619.506 0.040 0.047
Chain 1: 1400 -7924.489 0.036 0.038
Chain 1: 1500 -7557.808 0.031 0.038
Chain 1: 1600 -7738.743 0.026 0.023
Chain 1: 1700 -7671.933 0.025 0.023
Chain 1: 1800 -7719.218 0.024 0.023
Chain 1: 1900 -7466.717 0.023 0.023
Chain 1: 2000 -7592.312 0.024 0.023
Chain 1: 2100 -7541.976 0.020 0.017
Chain 1: 2200 -7704.159 0.021 0.021
Chain 1: 2300 -7564.227 0.022 0.021
Chain 1: 2400 -7511.765 0.019 0.018
Chain 1: 2500 -7556.044 0.015 0.017
Chain 1: 2600 -7511.518 0.013 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003089 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.89 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86608.657 1.000 1.000
Chain 1: 200 -13831.924 3.131 5.262
Chain 1: 300 -10020.240 2.214 1.000
Chain 1: 400 -12139.649 1.704 1.000
Chain 1: 500 -8412.931 1.452 0.443
Chain 1: 600 -8905.641 1.219 0.443
Chain 1: 700 -8603.006 1.050 0.380
Chain 1: 800 -8162.235 0.925 0.380
Chain 1: 900 -8195.171 0.823 0.175
Chain 1: 1000 -8827.783 0.748 0.175
Chain 1: 1100 -8512.140 0.652 0.072 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8936.134 0.130 0.055
Chain 1: 1300 -8301.624 0.100 0.055
Chain 1: 1400 -8357.599 0.083 0.054
Chain 1: 1500 -8389.886 0.039 0.047
Chain 1: 1600 -8349.847 0.034 0.037
Chain 1: 1700 -8214.875 0.032 0.037
Chain 1: 1800 -8267.905 0.027 0.016
Chain 1: 1900 -8307.287 0.028 0.016
Chain 1: 2000 -8453.002 0.022 0.016
Chain 1: 2100 -8184.299 0.022 0.016
Chain 1: 2200 -8205.840 0.017 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002911 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.11 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8427211.566 1.000 1.000
Chain 1: 200 -1585345.201 2.658 4.316
Chain 1: 300 -891551.309 2.031 1.000
Chain 1: 400 -458198.900 1.760 1.000
Chain 1: 500 -358576.180 1.463 0.946
Chain 1: 600 -233333.188 1.309 0.946
Chain 1: 700 -119558.553 1.258 0.946
Chain 1: 800 -86817.044 1.148 0.946
Chain 1: 900 -67157.276 1.053 0.778
Chain 1: 1000 -51974.162 0.977 0.778
Chain 1: 1100 -39464.137 0.908 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38649.090 0.479 0.377
Chain 1: 1300 -26592.936 0.447 0.377
Chain 1: 1400 -26315.468 0.353 0.317
Chain 1: 1500 -22899.671 0.340 0.317
Chain 1: 1600 -22117.421 0.290 0.293
Chain 1: 1700 -20987.936 0.200 0.292
Chain 1: 1800 -20932.184 0.163 0.149
Chain 1: 1900 -21259.308 0.135 0.054
Chain 1: 2000 -19767.461 0.113 0.054
Chain 1: 2100 -20005.804 0.083 0.035
Chain 1: 2200 -20233.378 0.082 0.035
Chain 1: 2300 -19849.382 0.038 0.019
Chain 1: 2400 -19621.115 0.039 0.019
Chain 1: 2500 -19423.320 0.025 0.015
Chain 1: 2600 -19052.294 0.023 0.015
Chain 1: 2700 -19008.929 0.018 0.012
Chain 1: 2800 -18725.475 0.019 0.015
Chain 1: 2900 -19007.140 0.019 0.015
Chain 1: 3000 -18993.192 0.012 0.012
Chain 1: 3100 -19078.383 0.011 0.012
Chain 1: 3200 -18768.338 0.011 0.015
Chain 1: 3300 -18973.639 0.011 0.012
Chain 1: 3400 -18447.402 0.012 0.015
Chain 1: 3500 -19061.094 0.014 0.015
Chain 1: 3600 -18365.361 0.016 0.015
Chain 1: 3700 -18753.959 0.018 0.017
Chain 1: 3800 -17710.045 0.023 0.021
Chain 1: 3900 -17706.128 0.021 0.021
Chain 1: 4000 -17823.393 0.022 0.021
Chain 1: 4100 -17737.028 0.022 0.021
Chain 1: 4200 -17552.476 0.021 0.021
Chain 1: 4300 -17691.408 0.021 0.021
Chain 1: 4400 -17647.555 0.018 0.011
Chain 1: 4500 -17550.016 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001187 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 11.87 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48628.922 1.000 1.000
Chain 1: 200 -19169.419 1.268 1.537
Chain 1: 300 -15624.094 0.921 1.000
Chain 1: 400 -17518.365 0.718 1.000
Chain 1: 500 -12551.329 0.654 0.396
Chain 1: 600 -13982.871 0.562 0.396
Chain 1: 700 -15132.819 0.492 0.227
Chain 1: 800 -12908.990 0.452 0.227
Chain 1: 900 -13733.442 0.409 0.172
Chain 1: 1000 -12286.740 0.380 0.172
Chain 1: 1100 -13459.699 0.288 0.118
Chain 1: 1200 -12650.187 0.141 0.108
Chain 1: 1300 -9809.550 0.147 0.108
Chain 1: 1400 -11092.257 0.148 0.116
Chain 1: 1500 -10002.754 0.119 0.109
Chain 1: 1600 -10182.117 0.111 0.109
Chain 1: 1700 -11184.544 0.112 0.109
Chain 1: 1800 -11669.258 0.099 0.090
Chain 1: 1900 -10611.177 0.103 0.100
Chain 1: 2000 -10416.238 0.093 0.090
Chain 1: 2100 -10310.651 0.086 0.090
Chain 1: 2200 -9677.177 0.086 0.090
Chain 1: 2300 -9100.852 0.063 0.065
Chain 1: 2400 -17860.161 0.101 0.065
Chain 1: 2500 -10413.092 0.161 0.065
Chain 1: 2600 -11653.430 0.170 0.090
Chain 1: 2700 -8946.791 0.191 0.100
Chain 1: 2800 -10751.039 0.204 0.106
Chain 1: 2900 -9643.793 0.205 0.115
Chain 1: 3000 -9015.274 0.211 0.115
Chain 1: 3100 -10343.339 0.222 0.128
Chain 1: 3200 -12044.214 0.230 0.141
Chain 1: 3300 -9430.781 0.251 0.168
Chain 1: 3400 -9450.160 0.203 0.141
Chain 1: 3500 -9314.540 0.132 0.128
Chain 1: 3600 -9248.233 0.123 0.128
Chain 1: 3700 -8749.943 0.098 0.115
Chain 1: 3800 -12139.258 0.109 0.115
Chain 1: 3900 -10371.840 0.115 0.128
Chain 1: 4000 -8729.181 0.127 0.141
Chain 1: 4100 -8493.541 0.116 0.141
Chain 1: 4200 -12207.828 0.133 0.170
Chain 1: 4300 -9348.008 0.136 0.170
Chain 1: 4400 -13613.931 0.167 0.188
Chain 1: 4500 -10056.493 0.201 0.279
Chain 1: 4600 -9102.265 0.210 0.279
Chain 1: 4700 -8856.937 0.208 0.279
Chain 1: 4800 -8892.505 0.180 0.188
Chain 1: 4900 -8724.545 0.165 0.188
Chain 1: 5000 -9692.885 0.156 0.105
Chain 1: 5100 -14247.721 0.185 0.304
Chain 1: 5200 -8891.372 0.215 0.306
Chain 1: 5300 -12726.129 0.215 0.301
Chain 1: 5400 -8311.114 0.236 0.301
Chain 1: 5500 -11630.546 0.230 0.285
Chain 1: 5600 -8294.207 0.259 0.301
Chain 1: 5700 -8536.324 0.259 0.301
Chain 1: 5800 -8422.938 0.260 0.301
Chain 1: 5900 -13653.671 0.297 0.320
Chain 1: 6000 -8486.024 0.348 0.383
Chain 1: 6100 -12639.861 0.349 0.383
Chain 1: 6200 -8099.891 0.344 0.383
Chain 1: 6300 -8141.753 0.315 0.383
Chain 1: 6400 -13396.181 0.301 0.383
Chain 1: 6500 -9017.737 0.321 0.392
Chain 1: 6600 -10248.131 0.293 0.383
Chain 1: 6700 -8921.423 0.305 0.383
Chain 1: 6800 -10053.872 0.315 0.383
Chain 1: 6900 -8774.449 0.291 0.329
Chain 1: 7000 -9361.048 0.236 0.149
Chain 1: 7100 -13120.482 0.232 0.149
Chain 1: 7200 -13155.776 0.176 0.146
Chain 1: 7300 -10606.644 0.200 0.149
Chain 1: 7400 -8125.204 0.191 0.149
Chain 1: 7500 -9924.152 0.161 0.149
Chain 1: 7600 -8439.319 0.166 0.176
Chain 1: 7700 -8441.400 0.151 0.176
Chain 1: 7800 -8424.950 0.140 0.176
Chain 1: 7900 -11121.540 0.150 0.181
Chain 1: 8000 -8173.636 0.180 0.240
Chain 1: 8100 -8833.134 0.159 0.181
Chain 1: 8200 -8471.766 0.163 0.181
Chain 1: 8300 -11792.568 0.167 0.181
Chain 1: 8400 -8159.065 0.181 0.181
Chain 1: 8500 -8111.998 0.163 0.176
Chain 1: 8600 -8156.736 0.146 0.075
Chain 1: 8700 -10090.193 0.165 0.192
Chain 1: 8800 -8336.114 0.186 0.210
Chain 1: 8900 -9389.386 0.173 0.192
Chain 1: 9000 -9065.588 0.141 0.112
Chain 1: 9100 -9168.393 0.134 0.112
Chain 1: 9200 -8299.996 0.140 0.112
Chain 1: 9300 -8031.019 0.116 0.105
Chain 1: 9400 -7912.256 0.073 0.036
Chain 1: 9500 -8026.417 0.073 0.036
Chain 1: 9600 -8353.600 0.077 0.039
Chain 1: 9700 -8428.342 0.058 0.036
Chain 1: 9800 -9242.944 0.046 0.036
Chain 1: 9900 -8505.199 0.044 0.036
Chain 1: 10000 -8211.036 0.044 0.036
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001375 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.75 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61314.102 1.000 1.000
Chain 1: 200 -17524.115 1.749 2.499
Chain 1: 300 -8651.273 1.508 1.026
Chain 1: 400 -8904.034 1.138 1.026
Chain 1: 500 -7725.708 0.941 1.000
Chain 1: 600 -8708.729 0.803 1.000
Chain 1: 700 -8156.775 0.698 0.153
Chain 1: 800 -8030.319 0.613 0.153
Chain 1: 900 -7879.465 0.547 0.113
Chain 1: 1000 -7782.176 0.493 0.113
Chain 1: 1100 -7700.447 0.394 0.068
Chain 1: 1200 -7672.661 0.145 0.028
Chain 1: 1300 -7611.187 0.043 0.019
Chain 1: 1400 -7591.255 0.041 0.016
Chain 1: 1500 -7598.123 0.025 0.013
Chain 1: 1600 -7491.850 0.016 0.013
Chain 1: 1700 -7476.524 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00299 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.9 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85235.107 1.000 1.000
Chain 1: 200 -13126.590 3.247 5.493
Chain 1: 300 -9628.058 2.286 1.000
Chain 1: 400 -10451.697 1.734 1.000
Chain 1: 500 -8495.252 1.433 0.363
Chain 1: 600 -8224.983 1.200 0.363
Chain 1: 700 -8589.378 1.034 0.230
Chain 1: 800 -8793.527 0.908 0.230
Chain 1: 900 -8495.637 0.811 0.079
Chain 1: 1000 -8216.145 0.733 0.079
Chain 1: 1100 -8466.715 0.636 0.042 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8225.116 0.090 0.035
Chain 1: 1300 -8393.231 0.056 0.034
Chain 1: 1400 -8326.586 0.048 0.033
Chain 1: 1500 -8271.871 0.026 0.030
Chain 1: 1600 -8268.575 0.023 0.029
Chain 1: 1700 -8207.290 0.019 0.023
Chain 1: 1800 -8087.664 0.019 0.020
Chain 1: 1900 -8200.929 0.016 0.015
Chain 1: 2000 -8162.725 0.013 0.014
Chain 1: 2100 -8304.728 0.012 0.014
Chain 1: 2200 -8088.403 0.012 0.014
Chain 1: 2300 -8229.652 0.012 0.014
Chain 1: 2400 -8116.481 0.012 0.014
Chain 1: 2500 -8172.159 0.012 0.014
Chain 1: 2600 -8185.397 0.012 0.014
Chain 1: 2700 -8106.908 0.013 0.014
Chain 1: 2800 -8091.158 0.011 0.014
Chain 1: 2900 -8088.463 0.010 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003042 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.42 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8400981.222 1.000 1.000
Chain 1: 200 -1582620.111 2.654 4.308
Chain 1: 300 -890362.720 2.029 1.000
Chain 1: 400 -457125.442 1.758 1.000
Chain 1: 500 -357472.179 1.462 0.948
Chain 1: 600 -232628.729 1.308 0.948
Chain 1: 700 -118854.711 1.258 0.948
Chain 1: 800 -86057.544 1.148 0.948
Chain 1: 900 -66393.688 1.054 0.778
Chain 1: 1000 -51177.434 0.978 0.778
Chain 1: 1100 -38648.042 0.910 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37817.257 0.482 0.381
Chain 1: 1300 -25778.470 0.451 0.381
Chain 1: 1400 -25494.770 0.357 0.324
Chain 1: 1500 -22084.042 0.345 0.324
Chain 1: 1600 -21300.052 0.295 0.297
Chain 1: 1700 -20175.151 0.205 0.296
Chain 1: 1800 -20119.257 0.167 0.154
Chain 1: 1900 -20444.583 0.139 0.056
Chain 1: 2000 -18958.075 0.117 0.056
Chain 1: 2100 -19196.156 0.086 0.037
Chain 1: 2200 -19422.003 0.085 0.037
Chain 1: 2300 -19040.009 0.040 0.020
Chain 1: 2400 -18812.419 0.040 0.020
Chain 1: 2500 -18614.471 0.026 0.016
Chain 1: 2600 -18245.383 0.024 0.016
Chain 1: 2700 -18202.639 0.019 0.012
Chain 1: 2800 -17919.832 0.020 0.016
Chain 1: 2900 -18200.733 0.020 0.015
Chain 1: 3000 -18186.994 0.012 0.012
Chain 1: 3100 -18271.818 0.011 0.012
Chain 1: 3200 -17963.025 0.012 0.015
Chain 1: 3300 -18167.365 0.011 0.012
Chain 1: 3400 -17643.210 0.013 0.015
Chain 1: 3500 -18253.662 0.015 0.016
Chain 1: 3600 -17562.287 0.017 0.016
Chain 1: 3700 -17947.609 0.019 0.017
Chain 1: 3800 -16910.282 0.023 0.021
Chain 1: 3900 -16906.536 0.022 0.021
Chain 1: 4000 -17023.819 0.023 0.021
Chain 1: 4100 -16937.681 0.023 0.021
Chain 1: 4200 -16754.646 0.022 0.021
Chain 1: 4300 -16892.532 0.022 0.021
Chain 1: 4400 -16849.873 0.019 0.011
Chain 1: 4500 -16752.552 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001316 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.16 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48813.279 1.000 1.000
Chain 1: 200 -23035.144 1.060 1.119
Chain 1: 300 -16868.869 0.828 1.000
Chain 1: 400 -12996.729 0.696 1.000
Chain 1: 500 -12875.987 0.558 0.366
Chain 1: 600 -15226.535 0.491 0.366
Chain 1: 700 -27162.134 0.484 0.366
Chain 1: 800 -14987.667 0.525 0.439
Chain 1: 900 -14189.653 0.473 0.366
Chain 1: 1000 -12903.413 0.435 0.366
Chain 1: 1100 -9644.912 0.369 0.338
Chain 1: 1200 -10843.011 0.268 0.298
Chain 1: 1300 -13093.141 0.249 0.172
Chain 1: 1400 -10349.312 0.246 0.172
Chain 1: 1500 -11910.669 0.258 0.172
Chain 1: 1600 -19681.635 0.282 0.265
Chain 1: 1700 -16532.675 0.257 0.190
Chain 1: 1800 -10866.205 0.228 0.190
Chain 1: 1900 -10093.339 0.230 0.190
Chain 1: 2000 -9739.582 0.224 0.190
Chain 1: 2100 -9662.202 0.191 0.172
Chain 1: 2200 -9733.563 0.180 0.172
Chain 1: 2300 -8932.272 0.172 0.131
Chain 1: 2400 -9755.657 0.154 0.090
Chain 1: 2500 -10109.915 0.144 0.084
Chain 1: 2600 -10250.719 0.106 0.077
Chain 1: 2700 -13206.628 0.110 0.077
Chain 1: 2800 -11077.837 0.077 0.077
Chain 1: 2900 -9732.473 0.083 0.084
Chain 1: 3000 -13164.366 0.105 0.090
Chain 1: 3100 -9380.587 0.145 0.138
Chain 1: 3200 -8776.952 0.151 0.138
Chain 1: 3300 -9429.296 0.149 0.138
Chain 1: 3400 -9283.006 0.142 0.138
Chain 1: 3500 -11830.415 0.160 0.192
Chain 1: 3600 -9587.632 0.182 0.215
Chain 1: 3700 -8880.908 0.168 0.192
Chain 1: 3800 -8582.660 0.152 0.138
Chain 1: 3900 -13240.653 0.173 0.215
Chain 1: 4000 -9915.298 0.181 0.215
Chain 1: 4100 -9984.577 0.141 0.080
Chain 1: 4200 -8702.807 0.149 0.147
Chain 1: 4300 -8453.207 0.145 0.147
Chain 1: 4400 -8625.615 0.145 0.147
Chain 1: 4500 -9904.011 0.137 0.129
Chain 1: 4600 -8490.732 0.130 0.129
Chain 1: 4700 -10189.105 0.139 0.147
Chain 1: 4800 -8890.150 0.150 0.147
Chain 1: 4900 -8794.231 0.116 0.146
Chain 1: 5000 -10365.678 0.097 0.146
Chain 1: 5100 -8343.227 0.121 0.147
Chain 1: 5200 -8631.500 0.110 0.146
Chain 1: 5300 -11614.956 0.132 0.152
Chain 1: 5400 -8244.705 0.171 0.166
Chain 1: 5500 -12571.503 0.193 0.167
Chain 1: 5600 -8311.657 0.227 0.242
Chain 1: 5700 -11206.110 0.237 0.257
Chain 1: 5800 -8829.553 0.249 0.258
Chain 1: 5900 -8523.188 0.251 0.258
Chain 1: 6000 -11307.406 0.261 0.258
Chain 1: 6100 -11563.417 0.239 0.258
Chain 1: 6200 -9910.946 0.252 0.258
Chain 1: 6300 -8443.359 0.244 0.258
Chain 1: 6400 -13565.994 0.241 0.258
Chain 1: 6500 -8520.561 0.265 0.258
Chain 1: 6600 -10684.555 0.234 0.246
Chain 1: 6700 -8283.526 0.238 0.246
Chain 1: 6800 -13225.305 0.248 0.246
Chain 1: 6900 -11257.236 0.262 0.246
Chain 1: 7000 -8684.565 0.267 0.290
Chain 1: 7100 -11964.773 0.292 0.290
Chain 1: 7200 -9360.609 0.303 0.290
Chain 1: 7300 -8367.519 0.298 0.290
Chain 1: 7400 -13727.142 0.299 0.290
Chain 1: 7500 -10007.598 0.277 0.290
Chain 1: 7600 -8255.313 0.278 0.290
Chain 1: 7700 -10245.132 0.268 0.278
Chain 1: 7800 -8477.931 0.252 0.274
Chain 1: 7900 -8101.114 0.239 0.274
Chain 1: 8000 -12231.270 0.243 0.274
Chain 1: 8100 -7956.481 0.270 0.278
Chain 1: 8200 -8098.824 0.243 0.212
Chain 1: 8300 -8225.909 0.233 0.212
Chain 1: 8400 -8184.713 0.195 0.208
Chain 1: 8500 -8202.741 0.158 0.194
Chain 1: 8600 -7799.404 0.142 0.052
Chain 1: 8700 -8359.701 0.129 0.052
Chain 1: 8800 -9577.817 0.121 0.052
Chain 1: 8900 -8322.863 0.131 0.067
Chain 1: 9000 -10119.141 0.115 0.067
Chain 1: 9100 -9954.766 0.063 0.052
Chain 1: 9200 -9070.315 0.071 0.067
Chain 1: 9300 -8010.127 0.083 0.098
Chain 1: 9400 -8078.343 0.083 0.098
Chain 1: 9500 -8312.443 0.086 0.098
Chain 1: 9600 -8098.352 0.083 0.098
Chain 1: 9700 -9940.337 0.095 0.127
Chain 1: 9800 -10928.187 0.091 0.098
Chain 1: 9900 -7941.278 0.114 0.098
Chain 1: 10000 -10631.081 0.121 0.098
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001405 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.05 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61201.806 1.000 1.000
Chain 1: 200 -17487.483 1.750 2.500
Chain 1: 300 -8701.506 1.503 1.010
Chain 1: 400 -8177.802 1.143 1.010
Chain 1: 500 -8267.566 0.917 1.000
Chain 1: 600 -8113.406 0.767 1.000
Chain 1: 700 -7742.034 0.664 0.064
Chain 1: 800 -8135.965 0.587 0.064
Chain 1: 900 -7811.367 0.527 0.048
Chain 1: 1000 -7753.467 0.475 0.048
Chain 1: 1100 -7672.020 0.376 0.048
Chain 1: 1200 -7676.088 0.126 0.042
Chain 1: 1300 -7726.945 0.026 0.019
Chain 1: 1400 -7834.770 0.021 0.014
Chain 1: 1500 -7571.969 0.023 0.019
Chain 1: 1600 -7515.943 0.022 0.014
Chain 1: 1700 -7473.629 0.018 0.011
Chain 1: 1800 -7513.598 0.013 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002907 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.07 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86228.932 1.000 1.000
Chain 1: 200 -13155.579 3.277 5.555
Chain 1: 300 -9583.456 2.309 1.000
Chain 1: 400 -10398.116 1.751 1.000
Chain 1: 500 -8516.963 1.445 0.373
Chain 1: 600 -8121.833 1.213 0.373
Chain 1: 700 -8219.288 1.041 0.221
Chain 1: 800 -8819.595 0.919 0.221
Chain 1: 900 -8349.322 0.823 0.078
Chain 1: 1000 -8190.899 0.743 0.078
Chain 1: 1100 -8442.091 0.646 0.068 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8145.405 0.094 0.056
Chain 1: 1300 -8286.102 0.059 0.049
Chain 1: 1400 -8305.520 0.051 0.036
Chain 1: 1500 -8190.097 0.030 0.030
Chain 1: 1600 -8292.123 0.027 0.019
Chain 1: 1700 -8379.824 0.027 0.019
Chain 1: 1800 -7984.385 0.025 0.019
Chain 1: 1900 -8086.250 0.020 0.017
Chain 1: 2000 -8056.563 0.019 0.014
Chain 1: 2100 -8180.047 0.017 0.014
Chain 1: 2200 -7962.867 0.016 0.014
Chain 1: 2300 -8114.816 0.017 0.014
Chain 1: 2400 -8129.070 0.017 0.014
Chain 1: 2500 -8097.823 0.016 0.013
Chain 1: 2600 -8100.247 0.014 0.013
Chain 1: 2700 -8006.614 0.014 0.013
Chain 1: 2800 -7978.265 0.010 0.012 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003153 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.53 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8421396.676 1.000 1.000
Chain 1: 200 -1586700.852 2.654 4.307
Chain 1: 300 -891597.465 2.029 1.000
Chain 1: 400 -457922.702 1.759 1.000
Chain 1: 500 -358014.782 1.463 0.947
Chain 1: 600 -232733.173 1.309 0.947
Chain 1: 700 -118872.835 1.258 0.947
Chain 1: 800 -86081.383 1.149 0.947
Chain 1: 900 -66410.377 1.054 0.780
Chain 1: 1000 -51201.198 0.978 0.780
Chain 1: 1100 -38682.726 0.911 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37854.270 0.482 0.381
Chain 1: 1300 -25820.959 0.451 0.381
Chain 1: 1400 -25539.027 0.357 0.324
Chain 1: 1500 -22129.976 0.345 0.324
Chain 1: 1600 -21347.228 0.295 0.297
Chain 1: 1700 -20222.479 0.204 0.296
Chain 1: 1800 -20166.778 0.166 0.154
Chain 1: 1900 -20492.517 0.138 0.056
Chain 1: 2000 -19005.460 0.117 0.056
Chain 1: 2100 -19243.540 0.085 0.037
Chain 1: 2200 -19469.723 0.084 0.037
Chain 1: 2300 -19087.306 0.040 0.020
Chain 1: 2400 -18859.561 0.040 0.020
Chain 1: 2500 -18661.631 0.026 0.016
Chain 1: 2600 -18292.132 0.024 0.016
Chain 1: 2700 -18249.226 0.019 0.012
Chain 1: 2800 -17966.291 0.020 0.016
Chain 1: 2900 -18247.333 0.020 0.015
Chain 1: 3000 -18233.518 0.012 0.012
Chain 1: 3100 -18318.457 0.011 0.012
Chain 1: 3200 -18009.389 0.012 0.015
Chain 1: 3300 -18213.926 0.011 0.012
Chain 1: 3400 -17689.316 0.013 0.015
Chain 1: 3500 -18300.495 0.015 0.016
Chain 1: 3600 -17608.092 0.017 0.016
Chain 1: 3700 -17994.187 0.019 0.017
Chain 1: 3800 -16955.351 0.023 0.021
Chain 1: 3900 -16951.549 0.022 0.021
Chain 1: 4000 -17068.830 0.023 0.021
Chain 1: 4100 -16982.675 0.023 0.021
Chain 1: 4200 -16799.241 0.022 0.021
Chain 1: 4300 -16937.393 0.022 0.021
Chain 1: 4400 -16894.466 0.019 0.011
Chain 1: 4500 -16797.070 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00137 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.7 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48772.355 1.000 1.000
Chain 1: 200 -19022.324 1.282 1.564
Chain 1: 300 -14580.263 0.956 1.000
Chain 1: 400 -13339.060 0.740 1.000
Chain 1: 500 -12898.106 0.599 0.305
Chain 1: 600 -19761.849 0.557 0.347
Chain 1: 700 -12102.806 0.568 0.347
Chain 1: 800 -11071.550 0.509 0.347
Chain 1: 900 -14285.474 0.477 0.305
Chain 1: 1000 -12701.040 0.442 0.305
Chain 1: 1100 -29752.485 0.399 0.305
Chain 1: 1200 -13715.529 0.360 0.305
Chain 1: 1300 -12301.862 0.341 0.225
Chain 1: 1400 -10371.990 0.350 0.225
Chain 1: 1500 -12356.609 0.363 0.225
Chain 1: 1600 -9878.067 0.353 0.225
Chain 1: 1700 -14672.547 0.322 0.225
Chain 1: 1800 -9777.835 0.363 0.251
Chain 1: 1900 -16575.410 0.382 0.327
Chain 1: 2000 -10608.375 0.425 0.410
Chain 1: 2100 -15333.056 0.399 0.327
Chain 1: 2200 -10499.617 0.328 0.327
Chain 1: 2300 -9508.288 0.327 0.327
Chain 1: 2400 -9785.295 0.311 0.327
Chain 1: 2500 -14639.731 0.328 0.332
Chain 1: 2600 -19510.091 0.328 0.332
Chain 1: 2700 -9578.862 0.399 0.410
Chain 1: 2800 -10393.750 0.357 0.332
Chain 1: 2900 -10743.928 0.319 0.308
Chain 1: 3000 -9029.177 0.282 0.250
Chain 1: 3100 -9380.213 0.255 0.190
Chain 1: 3200 -10311.796 0.218 0.104
Chain 1: 3300 -13913.700 0.233 0.190
Chain 1: 3400 -12766.630 0.240 0.190
Chain 1: 3500 -9176.823 0.245 0.190
Chain 1: 3600 -9543.749 0.224 0.090
Chain 1: 3700 -9548.556 0.121 0.090
Chain 1: 3800 -8663.037 0.123 0.090
Chain 1: 3900 -10065.696 0.134 0.102
Chain 1: 4000 -18098.508 0.159 0.102
Chain 1: 4100 -10106.604 0.235 0.139
Chain 1: 4200 -13095.096 0.248 0.228
Chain 1: 4300 -15729.480 0.239 0.167
Chain 1: 4400 -11018.656 0.273 0.228
Chain 1: 4500 -8792.427 0.259 0.228
Chain 1: 4600 -12169.729 0.283 0.253
Chain 1: 4700 -13350.264 0.292 0.253
Chain 1: 4800 -8601.557 0.337 0.278
Chain 1: 4900 -8590.032 0.323 0.278
Chain 1: 5000 -14675.257 0.320 0.278
Chain 1: 5100 -8874.212 0.306 0.278
Chain 1: 5200 -9868.675 0.294 0.278
Chain 1: 5300 -13377.026 0.303 0.278
Chain 1: 5400 -9629.303 0.299 0.278
Chain 1: 5500 -8442.088 0.288 0.278
Chain 1: 5600 -9100.526 0.268 0.262
Chain 1: 5700 -13645.245 0.292 0.333
Chain 1: 5800 -8870.661 0.291 0.333
Chain 1: 5900 -9243.435 0.295 0.333
Chain 1: 6000 -8639.594 0.260 0.262
Chain 1: 6100 -8640.102 0.195 0.141
Chain 1: 6200 -9939.589 0.198 0.141
Chain 1: 6300 -8775.491 0.185 0.133
Chain 1: 6400 -9677.559 0.155 0.131
Chain 1: 6500 -9031.885 0.148 0.093
Chain 1: 6600 -9554.115 0.146 0.093
Chain 1: 6700 -13022.538 0.140 0.093
Chain 1: 6800 -8328.738 0.142 0.093
Chain 1: 6900 -8569.009 0.141 0.093
Chain 1: 7000 -13533.836 0.171 0.131
Chain 1: 7100 -8246.482 0.235 0.133
Chain 1: 7200 -8461.730 0.224 0.133
Chain 1: 7300 -9020.159 0.217 0.093
Chain 1: 7400 -8530.488 0.214 0.071
Chain 1: 7500 -9234.082 0.214 0.076
Chain 1: 7600 -8628.597 0.216 0.076
Chain 1: 7700 -11558.543 0.214 0.076
Chain 1: 7800 -8327.645 0.197 0.076
Chain 1: 7900 -8927.634 0.201 0.076
Chain 1: 8000 -8193.123 0.173 0.076
Chain 1: 8100 -12761.350 0.145 0.076
Chain 1: 8200 -8379.367 0.194 0.090
Chain 1: 8300 -8805.502 0.193 0.090
Chain 1: 8400 -8907.788 0.189 0.090
Chain 1: 8500 -10217.547 0.194 0.128
Chain 1: 8600 -8889.290 0.202 0.149
Chain 1: 8700 -9483.333 0.183 0.128
Chain 1: 8800 -8067.494 0.161 0.128
Chain 1: 8900 -11054.986 0.182 0.149
Chain 1: 9000 -8819.266 0.198 0.175
Chain 1: 9100 -9359.801 0.168 0.149
Chain 1: 9200 -8307.521 0.128 0.128
Chain 1: 9300 -10505.292 0.144 0.149
Chain 1: 9400 -8989.378 0.160 0.169
Chain 1: 9500 -8256.788 0.156 0.169
Chain 1: 9600 -8800.271 0.147 0.169
Chain 1: 9700 -8235.691 0.148 0.169
Chain 1: 9800 -8302.658 0.131 0.127
Chain 1: 9900 -10293.235 0.124 0.127
Chain 1: 10000 -8522.937 0.119 0.127
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00143 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.3 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -46310.179 1.000 1.000
Chain 1: 200 -15450.225 1.499 1.997
Chain 1: 300 -8704.841 1.257 1.000
Chain 1: 400 -8609.318 0.946 1.000
Chain 1: 500 -8398.205 0.762 0.775
Chain 1: 600 -7805.300 0.647 0.775
Chain 1: 700 -7818.700 0.555 0.076
Chain 1: 800 -8228.726 0.492 0.076
Chain 1: 900 -8115.469 0.439 0.050
Chain 1: 1000 -8113.589 0.395 0.050
Chain 1: 1100 -7752.818 0.300 0.047
Chain 1: 1200 -7621.098 0.102 0.025
Chain 1: 1300 -7861.827 0.027 0.025
Chain 1: 1400 -7691.270 0.028 0.025
Chain 1: 1500 -7623.597 0.027 0.022
Chain 1: 1600 -7824.686 0.022 0.022
Chain 1: 1700 -7559.285 0.025 0.026
Chain 1: 1800 -7656.846 0.021 0.022
Chain 1: 1900 -7661.757 0.020 0.022
Chain 1: 2000 -7618.719 0.021 0.022
Chain 1: 2100 -7647.361 0.016 0.017
Chain 1: 2200 -7752.114 0.016 0.014
Chain 1: 2300 -7642.678 0.014 0.014
Chain 1: 2400 -7692.039 0.013 0.013
Chain 1: 2500 -7626.433 0.013 0.013
Chain 1: 2600 -7587.305 0.011 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002993 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.93 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86254.762 1.000 1.000
Chain 1: 200 -13531.208 3.187 5.375
Chain 1: 300 -9863.827 2.249 1.000
Chain 1: 400 -10697.912 1.706 1.000
Chain 1: 500 -8651.506 1.412 0.372
Chain 1: 600 -8279.736 1.184 0.372
Chain 1: 700 -8498.231 1.019 0.237
Chain 1: 800 -8670.810 0.894 0.237
Chain 1: 900 -8647.150 0.795 0.078
Chain 1: 1000 -8298.278 0.720 0.078
Chain 1: 1100 -8693.493 0.624 0.045 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8272.575 0.092 0.045
Chain 1: 1300 -8471.006 0.057 0.045
Chain 1: 1400 -8528.682 0.050 0.042
Chain 1: 1500 -8386.874 0.028 0.026
Chain 1: 1600 -8495.240 0.025 0.023
Chain 1: 1700 -8577.527 0.023 0.020
Chain 1: 1800 -8149.835 0.026 0.023
Chain 1: 1900 -8253.006 0.027 0.023
Chain 1: 2000 -8227.832 0.023 0.017
Chain 1: 2100 -8355.647 0.020 0.015
Chain 1: 2200 -8153.761 0.018 0.015
Chain 1: 2300 -8248.662 0.017 0.013
Chain 1: 2400 -8316.057 0.017 0.013
Chain 1: 2500 -8262.150 0.016 0.013
Chain 1: 2600 -8264.939 0.014 0.012
Chain 1: 2700 -8180.988 0.014 0.012
Chain 1: 2800 -8139.154 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003117 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.17 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8388875.145 1.000 1.000
Chain 1: 200 -1581080.468 2.653 4.306
Chain 1: 300 -889976.203 2.027 1.000
Chain 1: 400 -457546.280 1.757 1.000
Chain 1: 500 -358109.853 1.461 0.945
Chain 1: 600 -233109.578 1.307 0.945
Chain 1: 700 -119295.406 1.256 0.945
Chain 1: 800 -86514.741 1.147 0.945
Chain 1: 900 -66847.865 1.052 0.777
Chain 1: 1000 -51644.611 0.976 0.777
Chain 1: 1100 -39117.237 0.908 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38294.499 0.480 0.379
Chain 1: 1300 -26238.720 0.448 0.379
Chain 1: 1400 -25957.652 0.355 0.320
Chain 1: 1500 -22542.036 0.342 0.320
Chain 1: 1600 -21758.118 0.292 0.294
Chain 1: 1700 -20629.936 0.202 0.294
Chain 1: 1800 -20573.819 0.165 0.152
Chain 1: 1900 -20900.197 0.137 0.055
Chain 1: 2000 -19410.119 0.115 0.055
Chain 1: 2100 -19648.500 0.084 0.036
Chain 1: 2200 -19875.311 0.083 0.036
Chain 1: 2300 -19492.150 0.039 0.020
Chain 1: 2400 -19264.150 0.039 0.020
Chain 1: 2500 -19066.261 0.025 0.016
Chain 1: 2600 -18696.131 0.024 0.016
Chain 1: 2700 -18653.040 0.018 0.012
Chain 1: 2800 -18369.874 0.020 0.015
Chain 1: 2900 -18651.217 0.019 0.015
Chain 1: 3000 -18637.369 0.012 0.012
Chain 1: 3100 -18722.385 0.011 0.012
Chain 1: 3200 -18412.916 0.012 0.015
Chain 1: 3300 -18617.772 0.011 0.012
Chain 1: 3400 -18092.471 0.013 0.015
Chain 1: 3500 -18704.712 0.015 0.015
Chain 1: 3600 -18010.929 0.017 0.015
Chain 1: 3700 -18398.084 0.018 0.017
Chain 1: 3800 -17357.080 0.023 0.021
Chain 1: 3900 -17353.225 0.021 0.021
Chain 1: 4000 -17470.516 0.022 0.021
Chain 1: 4100 -17384.246 0.022 0.021
Chain 1: 4200 -17200.347 0.021 0.021
Chain 1: 4300 -17338.836 0.021 0.021
Chain 1: 4400 -17295.525 0.019 0.011
Chain 1: 4500 -17198.049 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001251 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.51 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13058.002 1.000 1.000
Chain 1: 200 -9881.541 0.661 1.000
Chain 1: 300 -8417.026 0.498 0.321
Chain 1: 400 -8594.714 0.379 0.321
Chain 1: 500 -8205.491 0.313 0.174
Chain 1: 600 -8312.697 0.263 0.174
Chain 1: 700 -8274.326 0.226 0.047
Chain 1: 800 -8264.024 0.198 0.047
Chain 1: 900 -8302.777 0.176 0.021
Chain 1: 1000 -8350.997 0.159 0.021
Chain 1: 1100 -8336.549 0.059 0.013
Chain 1: 1200 -8259.354 0.028 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001415 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.15 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58831.419 1.000 1.000
Chain 1: 200 -18456.891 1.594 2.188
Chain 1: 300 -9093.949 1.406 1.030
Chain 1: 400 -8115.829 1.084 1.030
Chain 1: 500 -8580.733 0.878 1.000
Chain 1: 600 -8504.571 0.733 1.000
Chain 1: 700 -8223.795 0.634 0.121
Chain 1: 800 -8645.467 0.560 0.121
Chain 1: 900 -8277.304 0.503 0.054
Chain 1: 1000 -8205.441 0.454 0.054
Chain 1: 1100 -7872.326 0.358 0.049
Chain 1: 1200 -7786.049 0.140 0.044
Chain 1: 1300 -7797.849 0.037 0.042
Chain 1: 1400 -7764.553 0.026 0.034
Chain 1: 1500 -7650.121 0.022 0.015
Chain 1: 1600 -7882.501 0.024 0.029
Chain 1: 1700 -7901.605 0.021 0.015
Chain 1: 1800 -7706.216 0.018 0.015
Chain 1: 1900 -7770.075 0.015 0.011
Chain 1: 2000 -7830.304 0.015 0.011
Chain 1: 2100 -7732.124 0.012 0.011
Chain 1: 2200 -8140.177 0.016 0.013
Chain 1: 2300 -7695.402 0.021 0.015
Chain 1: 2400 -7779.910 0.022 0.015
Chain 1: 2500 -7727.109 0.021 0.013
Chain 1: 2600 -7622.169 0.020 0.013
Chain 1: 2700 -7591.059 0.020 0.013
Chain 1: 2800 -7621.429 0.018 0.011
Chain 1: 2900 -7462.042 0.019 0.013
Chain 1: 3000 -7628.487 0.020 0.014
Chain 1: 3100 -7615.661 0.019 0.014
Chain 1: 3200 -7835.978 0.017 0.014
Chain 1: 3300 -7526.553 0.015 0.014
Chain 1: 3400 -7750.167 0.017 0.021
Chain 1: 3500 -7533.759 0.019 0.022
Chain 1: 3600 -7597.951 0.019 0.022
Chain 1: 3700 -7549.990 0.019 0.022
Chain 1: 3800 -7522.326 0.019 0.022
Chain 1: 3900 -7497.197 0.017 0.022
Chain 1: 4000 -7492.530 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003461 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.61 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87148.930 1.000 1.000
Chain 1: 200 -14229.027 3.062 5.125
Chain 1: 300 -10421.980 2.163 1.000
Chain 1: 400 -12450.844 1.663 1.000
Chain 1: 500 -8944.344 1.409 0.392
Chain 1: 600 -9647.884 1.186 0.392
Chain 1: 700 -8801.754 1.031 0.365
Chain 1: 800 -8966.501 0.904 0.365
Chain 1: 900 -9005.516 0.804 0.163
Chain 1: 1000 -9199.088 0.726 0.163
Chain 1: 1100 -9150.518 0.626 0.096 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8639.182 0.120 0.073
Chain 1: 1300 -9033.919 0.088 0.059
Chain 1: 1400 -8860.716 0.073 0.044
Chain 1: 1500 -8876.928 0.034 0.021
Chain 1: 1600 -8971.691 0.028 0.020
Chain 1: 1700 -9017.627 0.019 0.018
Chain 1: 1800 -8545.062 0.023 0.020
Chain 1: 1900 -8669.584 0.024 0.020
Chain 1: 2000 -8688.822 0.022 0.014
Chain 1: 2100 -8773.531 0.022 0.014
Chain 1: 2200 -8551.882 0.019 0.014
Chain 1: 2300 -8775.234 0.017 0.014
Chain 1: 2400 -8560.034 0.018 0.014
Chain 1: 2500 -8637.132 0.018 0.014
Chain 1: 2600 -8546.523 0.018 0.014
Chain 1: 2700 -8582.489 0.018 0.014
Chain 1: 2800 -8534.035 0.013 0.011
Chain 1: 2900 -8648.482 0.013 0.011
Chain 1: 3000 -8557.923 0.014 0.011
Chain 1: 3100 -8525.115 0.013 0.011
Chain 1: 3200 -8496.152 0.011 0.011
Chain 1: 3300 -8759.422 0.012 0.011
Chain 1: 3400 -8805.680 0.010 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003086 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.86 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8441191.115 1.000 1.000
Chain 1: 200 -1591276.706 2.652 4.305
Chain 1: 300 -892288.678 2.029 1.000
Chain 1: 400 -459201.779 1.758 1.000
Chain 1: 500 -358957.495 1.462 0.943
Chain 1: 600 -233694.820 1.308 0.943
Chain 1: 700 -119903.402 1.256 0.943
Chain 1: 800 -87130.749 1.146 0.943
Chain 1: 900 -67488.973 1.051 0.783
Chain 1: 1000 -52314.347 0.975 0.783
Chain 1: 1100 -39814.943 0.907 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39000.443 0.478 0.376
Chain 1: 1300 -26967.124 0.445 0.376
Chain 1: 1400 -26691.264 0.351 0.314
Chain 1: 1500 -23280.604 0.338 0.314
Chain 1: 1600 -22499.152 0.288 0.291
Chain 1: 1700 -21373.275 0.198 0.290
Chain 1: 1800 -21318.078 0.161 0.147
Chain 1: 1900 -21645.064 0.133 0.053
Chain 1: 2000 -20154.968 0.112 0.053
Chain 1: 2100 -20393.401 0.081 0.035
Chain 1: 2200 -20620.427 0.080 0.035
Chain 1: 2300 -20236.911 0.038 0.019
Chain 1: 2400 -20008.704 0.038 0.019
Chain 1: 2500 -19810.660 0.024 0.015
Chain 1: 2600 -19439.989 0.023 0.015
Chain 1: 2700 -19396.713 0.018 0.012
Chain 1: 2800 -19113.141 0.019 0.015
Chain 1: 2900 -19394.775 0.019 0.015
Chain 1: 3000 -19380.954 0.011 0.012
Chain 1: 3100 -19466.069 0.011 0.011
Chain 1: 3200 -19156.162 0.011 0.015
Chain 1: 3300 -19361.341 0.010 0.011
Chain 1: 3400 -18835.185 0.012 0.015
Chain 1: 3500 -19448.621 0.014 0.015
Chain 1: 3600 -18753.257 0.016 0.015
Chain 1: 3700 -19141.545 0.018 0.016
Chain 1: 3800 -18098.050 0.022 0.020
Chain 1: 3900 -18094.095 0.021 0.020
Chain 1: 4000 -18211.435 0.021 0.020
Chain 1: 4100 -18125.021 0.021 0.020
Chain 1: 4200 -17940.567 0.021 0.020
Chain 1: 4300 -18079.469 0.020 0.020
Chain 1: 4400 -18035.707 0.018 0.010
Chain 1: 4500 -17938.137 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001317 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.17 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48345.944 1.000 1.000
Chain 1: 200 -23451.930 1.031 1.061
Chain 1: 300 -29379.352 0.754 1.000
Chain 1: 400 -19373.565 0.695 1.000
Chain 1: 500 -18369.974 0.567 0.516
Chain 1: 600 -13072.859 0.540 0.516
Chain 1: 700 -13483.816 0.467 0.405
Chain 1: 800 -17120.071 0.435 0.405
Chain 1: 900 -16323.204 0.392 0.212
Chain 1: 1000 -11494.765 0.395 0.405
Chain 1: 1100 -10235.445 0.307 0.212
Chain 1: 1200 -10003.129 0.204 0.202
Chain 1: 1300 -10902.309 0.192 0.123
Chain 1: 1400 -13546.717 0.160 0.123
Chain 1: 1500 -9936.952 0.190 0.195
Chain 1: 1600 -10614.427 0.156 0.123
Chain 1: 1700 -10258.693 0.157 0.123
Chain 1: 1800 -11167.060 0.144 0.082
Chain 1: 1900 -10074.780 0.150 0.108
Chain 1: 2000 -10542.394 0.112 0.082
Chain 1: 2100 -8789.869 0.120 0.082
Chain 1: 2200 -9404.148 0.124 0.082
Chain 1: 2300 -9674.677 0.118 0.081
Chain 1: 2400 -8997.005 0.106 0.075
Chain 1: 2500 -9123.549 0.071 0.065
Chain 1: 2600 -9145.822 0.065 0.065
Chain 1: 2700 -10119.218 0.071 0.075
Chain 1: 2800 -9619.104 0.069 0.065
Chain 1: 2900 -9292.087 0.061 0.052
Chain 1: 3000 -8554.762 0.065 0.065
Chain 1: 3100 -8365.738 0.048 0.052
Chain 1: 3200 -8546.378 0.043 0.035
Chain 1: 3300 -15624.091 0.086 0.052
Chain 1: 3400 -8627.120 0.159 0.052
Chain 1: 3500 -9059.540 0.163 0.052
Chain 1: 3600 -9233.065 0.164 0.052
Chain 1: 3700 -8364.348 0.165 0.052
Chain 1: 3800 -8588.626 0.163 0.048
Chain 1: 3900 -8451.293 0.161 0.048
Chain 1: 4000 -9349.803 0.162 0.048
Chain 1: 4100 -8686.518 0.167 0.076
Chain 1: 4200 -9748.432 0.176 0.096
Chain 1: 4300 -13525.889 0.158 0.096
Chain 1: 4400 -10921.670 0.101 0.096
Chain 1: 4500 -8512.952 0.125 0.104
Chain 1: 4600 -13036.147 0.158 0.109
Chain 1: 4700 -8307.945 0.204 0.238
Chain 1: 4800 -8305.860 0.201 0.238
Chain 1: 4900 -12551.514 0.234 0.279
Chain 1: 5000 -11124.022 0.237 0.279
Chain 1: 5100 -8752.878 0.256 0.279
Chain 1: 5200 -8442.793 0.249 0.279
Chain 1: 5300 -14162.253 0.262 0.283
Chain 1: 5400 -10769.363 0.269 0.315
Chain 1: 5500 -8210.201 0.272 0.315
Chain 1: 5600 -9027.029 0.246 0.312
Chain 1: 5700 -13080.703 0.221 0.310
Chain 1: 5800 -8220.470 0.280 0.312
Chain 1: 5900 -10024.289 0.264 0.310
Chain 1: 6000 -8197.370 0.273 0.310
Chain 1: 6100 -9483.300 0.260 0.310
Chain 1: 6200 -9560.360 0.257 0.310
Chain 1: 6300 -10163.896 0.222 0.223
Chain 1: 6400 -11323.128 0.201 0.180
Chain 1: 6500 -11037.168 0.173 0.136
Chain 1: 6600 -8276.737 0.197 0.180
Chain 1: 6700 -12420.827 0.199 0.180
Chain 1: 6800 -8277.147 0.190 0.180
Chain 1: 6900 -8402.077 0.174 0.136
Chain 1: 7000 -9090.539 0.159 0.102
Chain 1: 7100 -11947.655 0.169 0.102
Chain 1: 7200 -8307.848 0.212 0.239
Chain 1: 7300 -9249.683 0.217 0.239
Chain 1: 7400 -12161.147 0.230 0.239
Chain 1: 7500 -9551.239 0.255 0.273
Chain 1: 7600 -8270.745 0.237 0.239
Chain 1: 7700 -8202.998 0.205 0.239
Chain 1: 7800 -9529.158 0.168 0.155
Chain 1: 7900 -8060.144 0.185 0.182
Chain 1: 8000 -8105.474 0.178 0.182
Chain 1: 8100 -8479.852 0.159 0.155
Chain 1: 8200 -8160.881 0.119 0.139
Chain 1: 8300 -9543.468 0.123 0.145
Chain 1: 8400 -11712.442 0.118 0.145
Chain 1: 8500 -8306.827 0.131 0.145
Chain 1: 8600 -8170.864 0.118 0.139
Chain 1: 8700 -8230.780 0.117 0.139
Chain 1: 8800 -7860.088 0.108 0.047
Chain 1: 8900 -8291.809 0.095 0.047
Chain 1: 9000 -9410.173 0.107 0.052
Chain 1: 9100 -10171.436 0.110 0.075
Chain 1: 9200 -12200.324 0.122 0.119
Chain 1: 9300 -9932.707 0.131 0.119
Chain 1: 9400 -8008.389 0.136 0.119
Chain 1: 9500 -8696.640 0.103 0.079
Chain 1: 9600 -8863.787 0.103 0.079
Chain 1: 9700 -8429.695 0.108 0.079
Chain 1: 9800 -8517.065 0.104 0.079
Chain 1: 9900 -8185.638 0.103 0.079
Chain 1: 10000 -8095.663 0.092 0.075
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001429 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.29 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56484.754 1.000 1.000
Chain 1: 200 -16886.511 1.672 2.345
Chain 1: 300 -8473.880 1.446 1.000
Chain 1: 400 -8616.002 1.089 1.000
Chain 1: 500 -7948.965 0.888 0.993
Chain 1: 600 -8364.047 0.748 0.993
Chain 1: 700 -7763.070 0.652 0.084
Chain 1: 800 -7908.924 0.573 0.084
Chain 1: 900 -7798.048 0.511 0.077
Chain 1: 1000 -7667.180 0.461 0.077
Chain 1: 1100 -7558.848 0.363 0.050
Chain 1: 1200 -7711.066 0.130 0.020
Chain 1: 1300 -7550.433 0.033 0.020
Chain 1: 1400 -7572.345 0.032 0.020
Chain 1: 1500 -7546.458 0.024 0.018
Chain 1: 1600 -7454.057 0.020 0.017
Chain 1: 1700 -7442.455 0.013 0.014
Chain 1: 1800 -7467.249 0.011 0.014
Chain 1: 1900 -7538.439 0.011 0.012
Chain 1: 2000 -7519.687 0.009 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002815 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 28.15 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85799.836 1.000 1.000
Chain 1: 200 -12998.804 3.300 5.601
Chain 1: 300 -9468.352 2.324 1.000
Chain 1: 400 -10379.490 1.765 1.000
Chain 1: 500 -8367.999 1.460 0.373
Chain 1: 600 -8035.760 1.224 0.373
Chain 1: 700 -8267.438 1.053 0.240
Chain 1: 800 -8468.185 0.924 0.240
Chain 1: 900 -8330.753 0.823 0.088
Chain 1: 1000 -8067.983 0.744 0.088
Chain 1: 1100 -8347.087 0.648 0.041 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8008.019 0.092 0.041
Chain 1: 1300 -8162.681 0.057 0.033
Chain 1: 1400 -8258.426 0.049 0.033
Chain 1: 1500 -8116.058 0.027 0.028
Chain 1: 1600 -8215.679 0.024 0.024
Chain 1: 1700 -8302.638 0.022 0.019
Chain 1: 1800 -7921.689 0.024 0.019
Chain 1: 1900 -8023.059 0.024 0.019
Chain 1: 2000 -7992.573 0.021 0.018
Chain 1: 2100 -8131.607 0.019 0.017
Chain 1: 2200 -7913.132 0.018 0.017
Chain 1: 2300 -8055.266 0.018 0.017
Chain 1: 2400 -8065.316 0.017 0.017
Chain 1: 2500 -8030.202 0.016 0.013
Chain 1: 2600 -8027.378 0.014 0.013
Chain 1: 2700 -7937.638 0.014 0.013
Chain 1: 2800 -7918.044 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002604 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 26.04 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8398310.281 1.000 1.000
Chain 1: 200 -1585265.415 2.649 4.298
Chain 1: 300 -890012.153 2.026 1.000
Chain 1: 400 -456444.701 1.757 1.000
Chain 1: 500 -356808.036 1.462 0.950
Chain 1: 600 -232019.266 1.308 0.950
Chain 1: 700 -118495.457 1.258 0.950
Chain 1: 800 -85740.826 1.148 0.950
Chain 1: 900 -66131.858 1.054 0.781
Chain 1: 1000 -50956.652 0.978 0.781
Chain 1: 1100 -38459.116 0.911 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37635.313 0.483 0.382
Chain 1: 1300 -25630.829 0.452 0.382
Chain 1: 1400 -25350.405 0.358 0.325
Chain 1: 1500 -21947.726 0.345 0.325
Chain 1: 1600 -21166.115 0.295 0.298
Chain 1: 1700 -20045.503 0.205 0.297
Chain 1: 1800 -19990.546 0.167 0.155
Chain 1: 1900 -20316.069 0.139 0.056
Chain 1: 2000 -18831.155 0.117 0.056
Chain 1: 2100 -19069.409 0.086 0.037
Chain 1: 2200 -19294.849 0.085 0.037
Chain 1: 2300 -18913.096 0.040 0.020
Chain 1: 2400 -18685.492 0.040 0.020
Chain 1: 2500 -18487.241 0.026 0.016
Chain 1: 2600 -18118.372 0.024 0.016
Chain 1: 2700 -18075.637 0.019 0.012
Chain 1: 2800 -17792.657 0.020 0.016
Chain 1: 2900 -18073.569 0.020 0.016
Chain 1: 3000 -18059.886 0.012 0.012
Chain 1: 3100 -18144.737 0.011 0.012
Chain 1: 3200 -17835.919 0.012 0.016
Chain 1: 3300 -18040.261 0.011 0.012
Chain 1: 3400 -17515.974 0.013 0.016
Chain 1: 3500 -18126.561 0.015 0.016
Chain 1: 3600 -17434.955 0.017 0.016
Chain 1: 3700 -17820.448 0.019 0.017
Chain 1: 3800 -16782.707 0.024 0.022
Chain 1: 3900 -16778.889 0.022 0.022
Chain 1: 4000 -16896.224 0.023 0.022
Chain 1: 4100 -16810.064 0.023 0.022
Chain 1: 4200 -16626.912 0.022 0.022
Chain 1: 4300 -16764.926 0.022 0.022
Chain 1: 4400 -16722.221 0.019 0.011
Chain 1: 4500 -16624.810 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001356 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.56 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -11983.296 1.000 1.000
Chain 1: 200 -8902.119 0.673 1.000
Chain 1: 300 -7916.141 0.490 0.346
Chain 1: 400 -7993.422 0.370 0.346
Chain 1: 500 -7828.289 0.300 0.125
Chain 1: 600 -7720.280 0.253 0.125
Chain 1: 700 -7652.914 0.218 0.021
Chain 1: 800 -7671.792 0.191 0.021
Chain 1: 900 -7762.171 0.171 0.014
Chain 1: 1000 -7680.693 0.155 0.014
Chain 1: 1100 -7781.855 0.056 0.013
Chain 1: 1200 -7687.777 0.023 0.012
Chain 1: 1300 -7623.005 0.011 0.012
Chain 1: 1400 -7644.584 0.011 0.012
Chain 1: 1500 -7730.188 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001378 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.78 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -62682.839 1.000 1.000
Chain 1: 200 -17691.030 1.772 2.543
Chain 1: 300 -8562.558 1.536 1.066
Chain 1: 400 -8166.922 1.164 1.066
Chain 1: 500 -8423.957 0.938 1.000
Chain 1: 600 -8569.879 0.784 1.000
Chain 1: 700 -7767.044 0.687 0.103
Chain 1: 800 -7791.780 0.601 0.103
Chain 1: 900 -7769.043 0.535 0.048
Chain 1: 1000 -7693.708 0.482 0.048
Chain 1: 1100 -7611.768 0.384 0.031
Chain 1: 1200 -7578.758 0.130 0.017
Chain 1: 1300 -7623.730 0.024 0.011
Chain 1: 1400 -7847.571 0.022 0.011
Chain 1: 1500 -7569.132 0.022 0.011
Chain 1: 1600 -7474.765 0.022 0.011
Chain 1: 1700 -7461.218 0.012 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002938 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.38 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85771.212 1.000 1.000
Chain 1: 200 -13058.368 3.284 5.568
Chain 1: 300 -9526.171 2.313 1.000
Chain 1: 400 -10450.285 1.757 1.000
Chain 1: 500 -8458.549 1.453 0.371
Chain 1: 600 -8332.291 1.213 0.371
Chain 1: 700 -8321.255 1.040 0.235
Chain 1: 800 -8880.474 0.918 0.235
Chain 1: 900 -8330.960 0.823 0.088
Chain 1: 1000 -8232.796 0.742 0.088
Chain 1: 1100 -8433.476 0.644 0.066 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8154.658 0.091 0.063
Chain 1: 1300 -8318.161 0.056 0.034
Chain 1: 1400 -8278.828 0.048 0.024
Chain 1: 1500 -8153.550 0.026 0.020
Chain 1: 1600 -8254.507 0.025 0.020
Chain 1: 1700 -8345.471 0.026 0.020
Chain 1: 1800 -7959.549 0.025 0.020
Chain 1: 1900 -8061.802 0.019 0.015
Chain 1: 2000 -8031.895 0.019 0.015
Chain 1: 2100 -8166.972 0.018 0.015
Chain 1: 2200 -7950.924 0.017 0.015
Chain 1: 2300 -8092.200 0.017 0.015
Chain 1: 2400 -8102.949 0.017 0.015
Chain 1: 2500 -8071.300 0.015 0.013
Chain 1: 2600 -8069.253 0.014 0.013
Chain 1: 2700 -7978.619 0.014 0.013
Chain 1: 2800 -7957.206 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002534 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.34 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8423285.821 1.000 1.000
Chain 1: 200 -1587203.522 2.653 4.307
Chain 1: 300 -890054.942 2.030 1.000
Chain 1: 400 -456532.972 1.760 1.000
Chain 1: 500 -356542.042 1.464 0.950
Chain 1: 600 -231600.849 1.310 0.950
Chain 1: 700 -118298.770 1.260 0.950
Chain 1: 800 -85621.207 1.150 0.950
Chain 1: 900 -66054.619 1.055 0.783
Chain 1: 1000 -50920.286 0.979 0.783
Chain 1: 1100 -38464.392 0.912 0.539 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37645.333 0.483 0.382
Chain 1: 1300 -25678.155 0.451 0.382
Chain 1: 1400 -25401.816 0.358 0.324
Chain 1: 1500 -22009.555 0.345 0.324
Chain 1: 1600 -21231.548 0.295 0.297
Chain 1: 1700 -20115.121 0.204 0.296
Chain 1: 1800 -20061.362 0.166 0.154
Chain 1: 1900 -20386.963 0.138 0.056
Chain 1: 2000 -18904.465 0.117 0.056
Chain 1: 2100 -19142.317 0.085 0.037
Chain 1: 2200 -19367.597 0.084 0.037
Chain 1: 2300 -18986.044 0.040 0.020
Chain 1: 2400 -18758.487 0.040 0.020
Chain 1: 2500 -18560.181 0.026 0.016
Chain 1: 2600 -18191.187 0.024 0.016
Chain 1: 2700 -18148.515 0.019 0.012
Chain 1: 2800 -17865.469 0.020 0.016
Chain 1: 2900 -18146.397 0.020 0.015
Chain 1: 3000 -18132.646 0.012 0.012
Chain 1: 3100 -18217.498 0.011 0.012
Chain 1: 3200 -17908.654 0.012 0.015
Chain 1: 3300 -18113.069 0.011 0.012
Chain 1: 3400 -17588.650 0.013 0.015
Chain 1: 3500 -18199.349 0.015 0.016
Chain 1: 3600 -17507.634 0.017 0.016
Chain 1: 3700 -17893.160 0.019 0.017
Chain 1: 3800 -16855.204 0.024 0.022
Chain 1: 3900 -16851.401 0.022 0.022
Chain 1: 4000 -16968.733 0.023 0.022
Chain 1: 4100 -16882.540 0.023 0.022
Chain 1: 4200 -16699.383 0.022 0.022
Chain 1: 4300 -16837.392 0.022 0.022
Chain 1: 4400 -16794.638 0.019 0.011
Chain 1: 4500 -16697.253 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001281 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.81 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -11979.001 1.000 1.000
Chain 1: 200 -8933.070 0.670 1.000
Chain 1: 300 -7882.570 0.491 0.341
Chain 1: 400 -7998.739 0.372 0.341
Chain 1: 500 -7832.144 0.302 0.133
Chain 1: 600 -7755.931 0.253 0.133
Chain 1: 700 -7688.313 0.218 0.021
Chain 1: 800 -7642.421 0.192 0.021
Chain 1: 900 -7670.337 0.171 0.015
Chain 1: 1000 -7751.812 0.155 0.015
Chain 1: 1100 -7793.533 0.055 0.011
Chain 1: 1200 -7706.995 0.022 0.011
Chain 1: 1300 -7663.856 0.010 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001418 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.18 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56427.212 1.000 1.000
Chain 1: 200 -16977.398 1.662 2.324
Chain 1: 300 -8491.942 1.441 1.000
Chain 1: 400 -8819.347 1.090 1.000
Chain 1: 500 -8439.269 0.881 0.999
Chain 1: 600 -9088.492 0.746 0.999
Chain 1: 700 -7889.441 0.661 0.152
Chain 1: 800 -8258.222 0.584 0.152
Chain 1: 900 -7649.830 0.528 0.080
Chain 1: 1000 -7791.249 0.477 0.080
Chain 1: 1100 -7484.363 0.381 0.071
Chain 1: 1200 -7753.594 0.152 0.045
Chain 1: 1300 -7620.317 0.054 0.045
Chain 1: 1400 -7840.889 0.053 0.045
Chain 1: 1500 -7541.172 0.053 0.041
Chain 1: 1600 -7453.425 0.047 0.040
Chain 1: 1700 -7448.958 0.032 0.035
Chain 1: 1800 -7480.788 0.028 0.028
Chain 1: 1900 -7414.134 0.020 0.018
Chain 1: 2000 -7524.020 0.020 0.017
Chain 1: 2100 -7591.668 0.017 0.015
Chain 1: 2200 -7600.960 0.014 0.012
Chain 1: 2300 -7496.724 0.013 0.012
Chain 1: 2400 -7528.392 0.011 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00321 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.1 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85745.529 1.000 1.000
Chain 1: 200 -13051.759 3.285 5.570
Chain 1: 300 -9535.140 2.313 1.000
Chain 1: 400 -10323.633 1.754 1.000
Chain 1: 500 -8473.854 1.447 0.369
Chain 1: 600 -8324.786 1.209 0.369
Chain 1: 700 -8354.343 1.036 0.218
Chain 1: 800 -8384.229 0.907 0.218
Chain 1: 900 -8389.366 0.807 0.076
Chain 1: 1000 -8111.612 0.729 0.076
Chain 1: 1100 -8460.225 0.633 0.041 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8199.634 0.080 0.034
Chain 1: 1300 -8329.841 0.044 0.032
Chain 1: 1400 -8300.484 0.037 0.018
Chain 1: 1500 -8184.172 0.017 0.016
Chain 1: 1600 -8279.041 0.016 0.014
Chain 1: 1700 -8382.741 0.017 0.014
Chain 1: 1800 -7992.935 0.021 0.016
Chain 1: 1900 -8093.513 0.023 0.016
Chain 1: 2000 -8063.627 0.020 0.014
Chain 1: 2100 -8204.215 0.017 0.014
Chain 1: 2200 -7984.608 0.017 0.014
Chain 1: 2300 -8127.134 0.017 0.014
Chain 1: 2400 -8015.582 0.018 0.014
Chain 1: 2500 -8070.217 0.017 0.014
Chain 1: 2600 -8083.894 0.016 0.014
Chain 1: 2700 -8006.020 0.016 0.014
Chain 1: 2800 -7988.493 0.011 0.012
Chain 1: 2900 -8008.690 0.010 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003157 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.57 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8397578.350 1.000 1.000
Chain 1: 200 -1589122.062 2.642 4.284
Chain 1: 300 -891954.636 2.022 1.000
Chain 1: 400 -457340.081 1.754 1.000
Chain 1: 500 -357046.731 1.459 0.950
Chain 1: 600 -231981.606 1.306 0.950
Chain 1: 700 -118480.457 1.256 0.950
Chain 1: 800 -85723.507 1.147 0.950
Chain 1: 900 -66129.841 1.053 0.782
Chain 1: 1000 -50972.202 0.977 0.782
Chain 1: 1100 -38488.925 0.909 0.539 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37668.858 0.483 0.382
Chain 1: 1300 -25678.663 0.452 0.382
Chain 1: 1400 -25400.483 0.358 0.324
Chain 1: 1500 -22000.913 0.345 0.324
Chain 1: 1600 -21220.330 0.295 0.297
Chain 1: 1700 -20101.316 0.205 0.296
Chain 1: 1800 -20046.803 0.167 0.155
Chain 1: 1900 -20372.284 0.139 0.056
Chain 1: 2000 -18888.239 0.117 0.056
Chain 1: 2100 -19126.507 0.086 0.037
Chain 1: 2200 -19351.736 0.085 0.037
Chain 1: 2300 -18970.200 0.040 0.020
Chain 1: 2400 -18742.584 0.040 0.020
Chain 1: 2500 -18544.293 0.026 0.016
Chain 1: 2600 -18175.503 0.024 0.016
Chain 1: 2700 -18132.850 0.019 0.012
Chain 1: 2800 -17849.776 0.020 0.016
Chain 1: 2900 -18130.730 0.020 0.015
Chain 1: 3000 -18117.074 0.012 0.012
Chain 1: 3100 -18201.874 0.011 0.012
Chain 1: 3200 -17893.128 0.012 0.015
Chain 1: 3300 -18097.462 0.011 0.012
Chain 1: 3400 -17573.186 0.013 0.015
Chain 1: 3500 -18183.699 0.015 0.016
Chain 1: 3600 -17492.261 0.017 0.016
Chain 1: 3700 -17877.557 0.019 0.017
Chain 1: 3800 -16840.043 0.024 0.022
Chain 1: 3900 -16836.236 0.022 0.022
Chain 1: 4000 -16953.588 0.023 0.022
Chain 1: 4100 -16867.369 0.023 0.022
Chain 1: 4200 -16684.316 0.022 0.022
Chain 1: 4300 -16822.273 0.022 0.022
Chain 1: 4400 -16779.607 0.019 0.011
Chain 1: 4500 -16682.195 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001394 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.94 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49255.491 1.000 1.000
Chain 1: 200 -16531.946 1.490 1.979
Chain 1: 300 -21224.020 1.067 1.000
Chain 1: 400 -17110.617 0.860 1.000
Chain 1: 500 -24033.259 0.746 0.288
Chain 1: 600 -12440.442 0.777 0.932
Chain 1: 700 -12518.939 0.667 0.288
Chain 1: 800 -14318.892 0.599 0.288
Chain 1: 900 -13983.716 0.535 0.240
Chain 1: 1000 -12558.137 0.493 0.240
Chain 1: 1100 -10539.728 0.412 0.221
Chain 1: 1200 -12777.305 0.232 0.192
Chain 1: 1300 -13276.367 0.213 0.175
Chain 1: 1400 -11406.360 0.206 0.164
Chain 1: 1500 -10317.916 0.187 0.126
Chain 1: 1600 -10485.780 0.096 0.114
Chain 1: 1700 -16950.956 0.133 0.126
Chain 1: 1800 -16463.251 0.124 0.114
Chain 1: 1900 -11135.618 0.169 0.164
Chain 1: 2000 -10169.962 0.167 0.164
Chain 1: 2100 -9795.555 0.152 0.105
Chain 1: 2200 -10000.565 0.137 0.095
Chain 1: 2300 -13543.677 0.159 0.105
Chain 1: 2400 -9352.314 0.187 0.105
Chain 1: 2500 -19057.964 0.228 0.262
Chain 1: 2600 -10291.755 0.311 0.381
Chain 1: 2700 -9234.511 0.285 0.262
Chain 1: 2800 -19079.303 0.333 0.448
Chain 1: 2900 -9640.848 0.383 0.448
Chain 1: 3000 -10724.063 0.384 0.448
Chain 1: 3100 -11487.843 0.387 0.448
Chain 1: 3200 -9170.323 0.410 0.448
Chain 1: 3300 -17944.004 0.433 0.489
Chain 1: 3400 -12638.085 0.430 0.489
Chain 1: 3500 -18909.738 0.412 0.420
Chain 1: 3600 -9931.860 0.417 0.420
Chain 1: 3700 -8979.671 0.417 0.420
Chain 1: 3800 -9327.482 0.369 0.332
Chain 1: 3900 -13406.978 0.301 0.304
Chain 1: 4000 -10716.530 0.316 0.304
Chain 1: 4100 -9527.404 0.322 0.304
Chain 1: 4200 -12263.357 0.319 0.304
Chain 1: 4300 -10295.715 0.289 0.251
Chain 1: 4400 -14673.678 0.277 0.251
Chain 1: 4500 -11982.060 0.266 0.225
Chain 1: 4600 -9658.389 0.200 0.225
Chain 1: 4700 -13534.238 0.218 0.241
Chain 1: 4800 -8850.653 0.267 0.251
Chain 1: 4900 -12574.436 0.267 0.251
Chain 1: 5000 -9396.103 0.275 0.286
Chain 1: 5100 -8568.569 0.272 0.286
Chain 1: 5200 -9249.415 0.257 0.286
Chain 1: 5300 -13602.318 0.270 0.296
Chain 1: 5400 -13995.585 0.243 0.286
Chain 1: 5500 -14027.732 0.221 0.286
Chain 1: 5600 -12855.364 0.206 0.286
Chain 1: 5700 -12706.630 0.179 0.097
Chain 1: 5800 -9129.160 0.165 0.097
Chain 1: 5900 -9136.525 0.135 0.091
Chain 1: 6000 -9144.495 0.102 0.074
Chain 1: 6100 -11410.069 0.112 0.074
Chain 1: 6200 -8464.508 0.139 0.091
Chain 1: 6300 -8461.786 0.107 0.028
Chain 1: 6400 -10541.731 0.124 0.091
Chain 1: 6500 -9813.407 0.131 0.091
Chain 1: 6600 -8621.342 0.136 0.138
Chain 1: 6700 -9489.521 0.144 0.138
Chain 1: 6800 -11892.993 0.125 0.138
Chain 1: 6900 -12549.759 0.130 0.138
Chain 1: 7000 -9645.830 0.160 0.197
Chain 1: 7100 -10513.423 0.149 0.138
Chain 1: 7200 -8622.540 0.136 0.138
Chain 1: 7300 -10937.993 0.157 0.197
Chain 1: 7400 -8776.128 0.162 0.202
Chain 1: 7500 -8435.393 0.159 0.202
Chain 1: 7600 -10382.501 0.163 0.202
Chain 1: 7700 -8432.889 0.177 0.212
Chain 1: 7800 -8693.182 0.160 0.212
Chain 1: 7900 -8511.862 0.157 0.212
Chain 1: 8000 -8511.877 0.127 0.188
Chain 1: 8100 -8383.174 0.120 0.188
Chain 1: 8200 -11212.422 0.124 0.188
Chain 1: 8300 -8337.899 0.137 0.188
Chain 1: 8400 -8687.220 0.116 0.040
Chain 1: 8500 -12530.353 0.143 0.188
Chain 1: 8600 -8393.754 0.173 0.231
Chain 1: 8700 -8782.387 0.155 0.044
Chain 1: 8800 -10454.505 0.168 0.160
Chain 1: 8900 -11579.151 0.175 0.160
Chain 1: 9000 -8686.200 0.209 0.252
Chain 1: 9100 -11323.891 0.230 0.252
Chain 1: 9200 -10749.762 0.211 0.233
Chain 1: 9300 -9120.849 0.194 0.179
Chain 1: 9400 -10353.355 0.202 0.179
Chain 1: 9500 -8310.340 0.196 0.179
Chain 1: 9600 -9632.189 0.160 0.160
Chain 1: 9700 -8519.647 0.169 0.160
Chain 1: 9800 -11351.595 0.178 0.179
Chain 1: 9900 -10099.091 0.180 0.179
Chain 1: 10000 -8211.770 0.170 0.179
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001392 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.92 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -62245.505 1.000 1.000
Chain 1: 200 -18175.471 1.712 2.425
Chain 1: 300 -8996.096 1.482 1.020
Chain 1: 400 -9610.333 1.127 1.020
Chain 1: 500 -8518.165 0.927 1.000
Chain 1: 600 -8684.070 0.776 1.000
Chain 1: 700 -7943.721 0.679 0.128
Chain 1: 800 -8231.353 0.598 0.128
Chain 1: 900 -8167.898 0.532 0.093
Chain 1: 1000 -7797.372 0.484 0.093
Chain 1: 1100 -7849.230 0.385 0.064
Chain 1: 1200 -7757.866 0.143 0.048
Chain 1: 1300 -7840.515 0.042 0.035
Chain 1: 1400 -7810.406 0.036 0.019
Chain 1: 1500 -7585.521 0.026 0.019
Chain 1: 1600 -7779.314 0.027 0.025
Chain 1: 1700 -7524.193 0.021 0.025
Chain 1: 1800 -7574.762 0.018 0.012
Chain 1: 1900 -7578.511 0.018 0.012
Chain 1: 2000 -7643.795 0.014 0.011
Chain 1: 2100 -7570.731 0.014 0.011
Chain 1: 2200 -7749.002 0.015 0.011
Chain 1: 2300 -7566.784 0.016 0.023
Chain 1: 2400 -7620.546 0.017 0.023
Chain 1: 2500 -7608.602 0.014 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002595 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.95 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85577.190 1.000 1.000
Chain 1: 200 -13765.386 3.108 5.217
Chain 1: 300 -10084.500 2.194 1.000
Chain 1: 400 -11419.058 1.675 1.000
Chain 1: 500 -9044.801 1.392 0.365
Chain 1: 600 -8978.461 1.161 0.365
Chain 1: 700 -8680.681 1.000 0.262
Chain 1: 800 -8923.039 0.879 0.262
Chain 1: 900 -8786.903 0.783 0.117
Chain 1: 1000 -8858.892 0.705 0.117
Chain 1: 1100 -8841.619 0.606 0.034 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8358.340 0.090 0.034
Chain 1: 1300 -8759.757 0.058 0.034
Chain 1: 1400 -8760.347 0.046 0.027
Chain 1: 1500 -8599.445 0.022 0.019
Chain 1: 1600 -8717.546 0.022 0.019
Chain 1: 1700 -8781.540 0.020 0.015
Chain 1: 1800 -8349.230 0.022 0.015
Chain 1: 1900 -8452.526 0.022 0.014
Chain 1: 2000 -8427.779 0.021 0.014
Chain 1: 2100 -8565.862 0.023 0.016
Chain 1: 2200 -8358.169 0.019 0.016
Chain 1: 2300 -8506.412 0.016 0.016
Chain 1: 2400 -8357.295 0.018 0.017
Chain 1: 2500 -8426.465 0.017 0.016
Chain 1: 2600 -8340.006 0.017 0.016
Chain 1: 2700 -8372.641 0.017 0.016
Chain 1: 2800 -8333.984 0.012 0.012
Chain 1: 2900 -8425.795 0.012 0.011
Chain 1: 3000 -8251.365 0.014 0.016
Chain 1: 3100 -8415.911 0.014 0.017
Chain 1: 3200 -8288.876 0.013 0.015
Chain 1: 3300 -8298.254 0.011 0.011
Chain 1: 3400 -8449.236 0.011 0.011
Chain 1: 3500 -8437.930 0.011 0.011
Chain 1: 3600 -8246.277 0.012 0.015
Chain 1: 3700 -8389.066 0.013 0.017
Chain 1: 3800 -8253.153 0.014 0.017
Chain 1: 3900 -8188.447 0.014 0.017
Chain 1: 4000 -8262.783 0.013 0.016
Chain 1: 4100 -8253.638 0.011 0.015
Chain 1: 4200 -8242.138 0.010 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003101 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.01 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8399207.756 1.000 1.000
Chain 1: 200 -1581008.889 2.656 4.313
Chain 1: 300 -890950.544 2.029 1.000
Chain 1: 400 -458317.602 1.758 1.000
Chain 1: 500 -358852.569 1.462 0.944
Chain 1: 600 -233864.766 1.307 0.944
Chain 1: 700 -119815.369 1.256 0.944
Chain 1: 800 -86976.283 1.147 0.944
Chain 1: 900 -67259.308 1.052 0.775
Chain 1: 1000 -52014.397 0.976 0.775
Chain 1: 1100 -39448.437 0.908 0.534 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38621.728 0.479 0.378
Chain 1: 1300 -26523.648 0.447 0.378
Chain 1: 1400 -26239.090 0.353 0.319
Chain 1: 1500 -22813.050 0.341 0.319
Chain 1: 1600 -22026.532 0.291 0.293
Chain 1: 1700 -20893.189 0.201 0.293
Chain 1: 1800 -20836.085 0.164 0.150
Chain 1: 1900 -21162.538 0.136 0.054
Chain 1: 2000 -19669.922 0.114 0.054
Chain 1: 2100 -19908.121 0.083 0.036
Chain 1: 2200 -20135.576 0.082 0.036
Chain 1: 2300 -19751.954 0.039 0.019
Chain 1: 2400 -19523.919 0.039 0.019
Chain 1: 2500 -19326.290 0.025 0.015
Chain 1: 2600 -18955.634 0.023 0.015
Chain 1: 2700 -18912.482 0.018 0.012
Chain 1: 2800 -18629.313 0.019 0.015
Chain 1: 2900 -18910.859 0.019 0.015
Chain 1: 3000 -18896.889 0.012 0.012
Chain 1: 3100 -18981.907 0.011 0.012
Chain 1: 3200 -18672.276 0.012 0.015
Chain 1: 3300 -18877.315 0.011 0.012
Chain 1: 3400 -18351.751 0.012 0.015
Chain 1: 3500 -18964.369 0.015 0.015
Chain 1: 3600 -18270.245 0.016 0.015
Chain 1: 3700 -18657.650 0.018 0.017
Chain 1: 3800 -17616.067 0.023 0.021
Chain 1: 3900 -17612.289 0.021 0.021
Chain 1: 4000 -17729.527 0.022 0.021
Chain 1: 4100 -17643.189 0.022 0.021
Chain 1: 4200 -17459.243 0.021 0.021
Chain 1: 4300 -17597.724 0.021 0.021
Chain 1: 4400 -17554.302 0.018 0.011
Chain 1: 4500 -17456.910 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001296 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.96 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12519.074 1.000 1.000
Chain 1: 200 -9491.187 0.660 1.000
Chain 1: 300 -8179.673 0.493 0.319
Chain 1: 400 -8377.055 0.376 0.319
Chain 1: 500 -8228.990 0.304 0.160
Chain 1: 600 -8142.923 0.255 0.160
Chain 1: 700 -8053.921 0.220 0.024
Chain 1: 800 -8093.802 0.193 0.024
Chain 1: 900 -8221.309 0.174 0.018
Chain 1: 1000 -8133.177 0.157 0.018
Chain 1: 1100 -8087.216 0.058 0.016
Chain 1: 1200 -8074.142 0.026 0.011
Chain 1: 1300 -8025.342 0.011 0.011
Chain 1: 1400 -8054.032 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001393 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.93 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -45698.432 1.000 1.000
Chain 1: 200 -15656.787 1.459 1.919
Chain 1: 300 -8761.175 1.235 1.000
Chain 1: 400 -8669.865 0.929 1.000
Chain 1: 500 -8710.699 0.744 0.787
Chain 1: 600 -8548.860 0.623 0.787
Chain 1: 700 -7842.293 0.547 0.090
Chain 1: 800 -8172.373 0.484 0.090
Chain 1: 900 -7999.696 0.432 0.040
Chain 1: 1000 -7867.660 0.391 0.040
Chain 1: 1100 -7967.565 0.292 0.022
Chain 1: 1200 -7677.642 0.104 0.022
Chain 1: 1300 -7894.525 0.028 0.022
Chain 1: 1400 -7849.782 0.028 0.022
Chain 1: 1500 -7668.601 0.029 0.024
Chain 1: 1600 -7665.011 0.028 0.024
Chain 1: 1700 -7579.830 0.020 0.022
Chain 1: 1800 -7626.662 0.016 0.017
Chain 1: 1900 -7653.378 0.015 0.013
Chain 1: 2000 -7663.262 0.013 0.011
Chain 1: 2100 -7649.872 0.012 0.006 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00259 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.9 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87266.612 1.000 1.000
Chain 1: 200 -13633.904 3.200 5.401
Chain 1: 300 -10005.925 2.254 1.000
Chain 1: 400 -10744.349 1.708 1.000
Chain 1: 500 -8974.449 1.406 0.363
Chain 1: 600 -8527.277 1.180 0.363
Chain 1: 700 -8418.362 1.014 0.197
Chain 1: 800 -9049.280 0.896 0.197
Chain 1: 900 -8879.365 0.798 0.070
Chain 1: 1000 -8596.515 0.722 0.070
Chain 1: 1100 -8847.028 0.624 0.069 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8503.333 0.088 0.052
Chain 1: 1300 -8723.554 0.055 0.040
Chain 1: 1400 -8726.592 0.048 0.033
Chain 1: 1500 -8581.947 0.030 0.028
Chain 1: 1600 -8694.526 0.026 0.025
Chain 1: 1700 -8780.692 0.026 0.025
Chain 1: 1800 -8373.231 0.023 0.025
Chain 1: 1900 -8469.205 0.023 0.025
Chain 1: 2000 -8441.666 0.020 0.017
Chain 1: 2100 -8562.626 0.018 0.014
Chain 1: 2200 -8454.403 0.016 0.013
Chain 1: 2300 -8509.341 0.014 0.013
Chain 1: 2400 -8398.417 0.015 0.013
Chain 1: 2500 -8445.509 0.014 0.013
Chain 1: 2600 -8472.725 0.013 0.011
Chain 1: 2700 -8388.926 0.013 0.011
Chain 1: 2800 -8357.879 0.008 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002477 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 24.77 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8414985.120 1.000 1.000
Chain 1: 200 -1586046.573 2.653 4.306
Chain 1: 300 -891977.191 2.028 1.000
Chain 1: 400 -457826.638 1.758 1.000
Chain 1: 500 -358152.396 1.462 0.948
Chain 1: 600 -232946.974 1.308 0.948
Chain 1: 700 -119309.586 1.257 0.948
Chain 1: 800 -86501.168 1.147 0.948
Chain 1: 900 -66859.355 1.053 0.778
Chain 1: 1000 -51665.657 0.977 0.778
Chain 1: 1100 -39150.451 0.909 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38328.712 0.480 0.379
Chain 1: 1300 -26300.897 0.448 0.379
Chain 1: 1400 -26019.512 0.354 0.320
Chain 1: 1500 -22610.646 0.342 0.320
Chain 1: 1600 -21827.940 0.292 0.294
Chain 1: 1700 -20704.278 0.202 0.294
Chain 1: 1800 -20648.909 0.164 0.151
Chain 1: 1900 -20974.911 0.136 0.054
Chain 1: 2000 -19487.563 0.114 0.054
Chain 1: 2100 -19725.880 0.084 0.036
Chain 1: 2200 -19951.934 0.083 0.036
Chain 1: 2300 -19569.557 0.039 0.020
Chain 1: 2400 -19341.776 0.039 0.020
Chain 1: 2500 -19143.539 0.025 0.016
Chain 1: 2600 -18774.039 0.023 0.016
Chain 1: 2700 -18731.164 0.018 0.012
Chain 1: 2800 -18447.962 0.019 0.015
Chain 1: 2900 -18729.179 0.019 0.015
Chain 1: 3000 -18715.344 0.012 0.012
Chain 1: 3100 -18800.300 0.011 0.012
Chain 1: 3200 -18491.112 0.012 0.015
Chain 1: 3300 -18695.779 0.011 0.012
Chain 1: 3400 -18170.819 0.012 0.015
Chain 1: 3500 -18782.382 0.015 0.015
Chain 1: 3600 -18089.561 0.017 0.015
Chain 1: 3700 -18475.987 0.018 0.017
Chain 1: 3800 -17436.296 0.023 0.021
Chain 1: 3900 -17432.458 0.021 0.021
Chain 1: 4000 -17549.783 0.022 0.021
Chain 1: 4100 -17463.501 0.022 0.021
Chain 1: 4200 -17279.953 0.021 0.021
Chain 1: 4300 -17418.239 0.021 0.021
Chain 1: 4400 -17375.214 0.018 0.011
Chain 1: 4500 -17277.739 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001395 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.95 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13014.113 1.000 1.000
Chain 1: 200 -9945.907 0.654 1.000
Chain 1: 300 -8394.059 0.498 0.308
Chain 1: 400 -8614.930 0.380 0.308
Chain 1: 500 -8333.448 0.311 0.185
Chain 1: 600 -8256.863 0.260 0.185
Chain 1: 700 -8301.416 0.224 0.034
Chain 1: 800 -8288.018 0.196 0.034
Chain 1: 900 -8207.251 0.175 0.026
Chain 1: 1000 -8236.380 0.158 0.026
Chain 1: 1100 -8375.049 0.060 0.017
Chain 1: 1200 -8247.499 0.031 0.015
Chain 1: 1300 -8197.870 0.013 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001413 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.13 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49787.170 1.000 1.000
Chain 1: 200 -16752.816 1.486 1.972
Chain 1: 300 -9176.986 1.266 1.000
Chain 1: 400 -9742.870 0.964 1.000
Chain 1: 500 -8901.321 0.790 0.826
Chain 1: 600 -8793.353 0.660 0.826
Chain 1: 700 -8309.984 0.574 0.095
Chain 1: 800 -8591.798 0.507 0.095
Chain 1: 900 -8249.982 0.455 0.058
Chain 1: 1000 -7848.090 0.415 0.058
Chain 1: 1100 -7900.226 0.315 0.058
Chain 1: 1200 -8149.299 0.121 0.051
Chain 1: 1300 -8142.631 0.039 0.041
Chain 1: 1400 -7932.870 0.035 0.033
Chain 1: 1500 -7742.654 0.028 0.031
Chain 1: 1600 -7862.954 0.029 0.031
Chain 1: 1700 -7787.524 0.024 0.026
Chain 1: 1800 -7803.836 0.021 0.025
Chain 1: 1900 -7832.533 0.017 0.015
Chain 1: 2000 -7880.411 0.013 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002505 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.05 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86875.544 1.000 1.000
Chain 1: 200 -14241.731 3.050 5.100
Chain 1: 300 -10418.989 2.156 1.000
Chain 1: 400 -12244.506 1.654 1.000
Chain 1: 500 -9189.260 1.390 0.367
Chain 1: 600 -8734.590 1.167 0.367
Chain 1: 700 -9254.840 1.008 0.332
Chain 1: 800 -9167.546 0.883 0.332
Chain 1: 900 -9068.744 0.786 0.149
Chain 1: 1000 -9273.538 0.710 0.149
Chain 1: 1100 -9145.994 0.611 0.056 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8706.990 0.106 0.052
Chain 1: 1300 -9019.079 0.073 0.050
Chain 1: 1400 -8891.433 0.060 0.035
Chain 1: 1500 -8898.674 0.026 0.022
Chain 1: 1600 -8962.774 0.022 0.014
Chain 1: 1700 -9020.879 0.017 0.014
Chain 1: 1800 -8558.850 0.021 0.014
Chain 1: 1900 -8678.130 0.022 0.014
Chain 1: 2000 -8697.644 0.020 0.014
Chain 1: 2100 -8783.109 0.019 0.014
Chain 1: 2200 -8562.251 0.017 0.014
Chain 1: 2300 -8769.212 0.016 0.014
Chain 1: 2400 -8575.892 0.017 0.014
Chain 1: 2500 -8648.363 0.017 0.014
Chain 1: 2600 -8558.653 0.018 0.014
Chain 1: 2700 -8592.290 0.017 0.014
Chain 1: 2800 -8542.956 0.013 0.010
Chain 1: 2900 -8658.175 0.013 0.010
Chain 1: 3000 -8568.533 0.013 0.010
Chain 1: 3100 -8534.970 0.013 0.010
Chain 1: 3200 -8506.264 0.011 0.010
Chain 1: 3300 -8768.380 0.011 0.010
Chain 1: 3400 -8813.336 0.009 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003138 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.38 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8408635.466 1.000 1.000
Chain 1: 200 -1583827.495 2.655 4.309
Chain 1: 300 -893097.638 2.027 1.000
Chain 1: 400 -459279.836 1.757 1.000
Chain 1: 500 -359796.709 1.461 0.945
Chain 1: 600 -234532.086 1.306 0.945
Chain 1: 700 -120421.782 1.255 0.945
Chain 1: 800 -87503.125 1.145 0.945
Chain 1: 900 -67773.995 1.050 0.773
Chain 1: 1000 -52521.913 0.974 0.773
Chain 1: 1100 -39943.826 0.906 0.534 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39122.815 0.477 0.376
Chain 1: 1300 -27015.816 0.444 0.376
Chain 1: 1400 -26732.179 0.351 0.315
Chain 1: 1500 -23301.776 0.338 0.315
Chain 1: 1600 -22513.923 0.288 0.291
Chain 1: 1700 -21379.795 0.199 0.290
Chain 1: 1800 -21322.613 0.161 0.147
Chain 1: 1900 -21649.545 0.134 0.053
Chain 1: 2000 -20154.849 0.112 0.053
Chain 1: 2100 -20393.705 0.082 0.035
Chain 1: 2200 -20621.227 0.081 0.035
Chain 1: 2300 -20237.304 0.038 0.019
Chain 1: 2400 -20009.029 0.038 0.019
Chain 1: 2500 -19811.079 0.024 0.015
Chain 1: 2600 -19440.276 0.023 0.015
Chain 1: 2700 -19396.995 0.018 0.012
Chain 1: 2800 -19113.421 0.019 0.015
Chain 1: 2900 -19395.235 0.019 0.015
Chain 1: 3000 -19381.324 0.011 0.012
Chain 1: 3100 -19466.416 0.011 0.011
Chain 1: 3200 -19156.476 0.011 0.015
Chain 1: 3300 -19361.723 0.010 0.011
Chain 1: 3400 -18835.508 0.012 0.015
Chain 1: 3500 -19449.033 0.014 0.015
Chain 1: 3600 -18753.689 0.016 0.015
Chain 1: 3700 -19142.021 0.018 0.016
Chain 1: 3800 -18098.451 0.022 0.020
Chain 1: 3900 -18094.536 0.021 0.020
Chain 1: 4000 -18211.848 0.021 0.020
Chain 1: 4100 -18125.364 0.021 0.020
Chain 1: 4200 -17940.963 0.021 0.020
Chain 1: 4300 -18079.821 0.020 0.020
Chain 1: 4400 -18036.089 0.018 0.010
Chain 1: 4500 -17938.521 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001302 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.02 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49302.195 1.000 1.000
Chain 1: 200 -17151.023 1.437 1.875
Chain 1: 300 -16345.236 0.975 1.000
Chain 1: 400 -19103.770 0.767 1.000
Chain 1: 500 -16178.172 0.650 0.181
Chain 1: 600 -12060.335 0.598 0.341
Chain 1: 700 -14696.432 0.539 0.181
Chain 1: 800 -12928.222 0.488 0.181
Chain 1: 900 -17690.319 0.464 0.181
Chain 1: 1000 -12744.507 0.456 0.269
Chain 1: 1100 -12692.212 0.357 0.181
Chain 1: 1200 -12880.326 0.171 0.179
Chain 1: 1300 -10644.095 0.187 0.181
Chain 1: 1400 -10443.567 0.174 0.181
Chain 1: 1500 -10465.594 0.156 0.179
Chain 1: 1600 -10209.981 0.125 0.137
Chain 1: 1700 -12218.176 0.123 0.137
Chain 1: 1800 -10905.743 0.122 0.120
Chain 1: 1900 -10989.977 0.096 0.025
Chain 1: 2000 -18588.898 0.098 0.025
Chain 1: 2100 -10467.883 0.175 0.120
Chain 1: 2200 -20345.174 0.222 0.164
Chain 1: 2300 -9710.124 0.310 0.164
Chain 1: 2400 -9967.171 0.311 0.164
Chain 1: 2500 -10091.214 0.312 0.164
Chain 1: 2600 -12695.714 0.330 0.205
Chain 1: 2700 -15547.896 0.332 0.205
Chain 1: 2800 -20890.558 0.346 0.256
Chain 1: 2900 -16104.306 0.374 0.297
Chain 1: 3000 -9141.157 0.410 0.297
Chain 1: 3100 -10404.007 0.344 0.256
Chain 1: 3200 -10179.852 0.298 0.205
Chain 1: 3300 -9957.867 0.191 0.183
Chain 1: 3400 -9488.437 0.193 0.183
Chain 1: 3500 -9716.749 0.194 0.183
Chain 1: 3600 -10246.654 0.179 0.121
Chain 1: 3700 -10289.712 0.161 0.052
Chain 1: 3800 -10004.481 0.138 0.049
Chain 1: 3900 -9466.782 0.114 0.049
Chain 1: 4000 -20807.216 0.092 0.049
Chain 1: 4100 -9131.722 0.208 0.049
Chain 1: 4200 -10455.142 0.219 0.052
Chain 1: 4300 -14617.302 0.245 0.057
Chain 1: 4400 -13019.388 0.252 0.123
Chain 1: 4500 -11267.878 0.265 0.127
Chain 1: 4600 -11494.240 0.262 0.127
Chain 1: 4700 -9022.128 0.289 0.155
Chain 1: 4800 -9409.146 0.290 0.155
Chain 1: 4900 -11412.784 0.302 0.176
Chain 1: 5000 -10392.421 0.258 0.155
Chain 1: 5100 -9257.157 0.142 0.127
Chain 1: 5200 -12173.868 0.153 0.155
Chain 1: 5300 -12421.262 0.127 0.123
Chain 1: 5400 -12108.714 0.117 0.123
Chain 1: 5500 -10754.007 0.114 0.123
Chain 1: 5600 -10290.493 0.117 0.123
Chain 1: 5700 -11436.179 0.099 0.100
Chain 1: 5800 -11174.582 0.098 0.100
Chain 1: 5900 -9654.755 0.096 0.100
Chain 1: 6000 -9509.217 0.088 0.100
Chain 1: 6100 -11370.115 0.092 0.100
Chain 1: 6200 -10450.061 0.076 0.088
Chain 1: 6300 -8825.350 0.093 0.100
Chain 1: 6400 -13270.444 0.124 0.126
Chain 1: 6500 -12773.803 0.115 0.100
Chain 1: 6600 -9226.641 0.149 0.157
Chain 1: 6700 -12971.562 0.168 0.164
Chain 1: 6800 -12550.111 0.169 0.164
Chain 1: 6900 -8690.192 0.198 0.184
Chain 1: 7000 -8641.522 0.197 0.184
Chain 1: 7100 -8536.729 0.181 0.184
Chain 1: 7200 -8879.430 0.177 0.184
Chain 1: 7300 -11392.879 0.180 0.221
Chain 1: 7400 -8937.015 0.174 0.221
Chain 1: 7500 -10254.333 0.183 0.221
Chain 1: 7600 -11995.417 0.159 0.145
Chain 1: 7700 -9074.578 0.163 0.145
Chain 1: 7800 -8992.577 0.160 0.145
Chain 1: 7900 -8981.120 0.116 0.128
Chain 1: 8000 -8607.442 0.120 0.128
Chain 1: 8100 -9311.868 0.126 0.128
Chain 1: 8200 -9360.402 0.123 0.128
Chain 1: 8300 -10402.401 0.111 0.100
Chain 1: 8400 -11755.448 0.095 0.100
Chain 1: 8500 -9899.808 0.100 0.100
Chain 1: 8600 -11339.578 0.099 0.100
Chain 1: 8700 -9626.399 0.084 0.100
Chain 1: 8800 -9955.576 0.087 0.100
Chain 1: 8900 -8473.604 0.104 0.115
Chain 1: 9000 -11858.649 0.128 0.127
Chain 1: 9100 -8351.289 0.163 0.175
Chain 1: 9200 -9047.323 0.170 0.175
Chain 1: 9300 -9132.195 0.161 0.175
Chain 1: 9400 -11235.162 0.168 0.178
Chain 1: 9500 -11291.499 0.150 0.175
Chain 1: 9600 -8529.489 0.169 0.178
Chain 1: 9700 -8930.404 0.156 0.175
Chain 1: 9800 -8617.254 0.156 0.175
Chain 1: 9900 -9598.636 0.149 0.102
Chain 1: 10000 -9200.966 0.125 0.077
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001374 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.74 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -62075.655 1.000 1.000
Chain 1: 200 -18159.117 1.709 2.418
Chain 1: 300 -9054.615 1.475 1.006
Chain 1: 400 -9677.719 1.122 1.006
Chain 1: 500 -8577.698 0.923 1.000
Chain 1: 600 -8525.301 0.770 1.000
Chain 1: 700 -8293.847 0.664 0.128
Chain 1: 800 -8335.800 0.582 0.128
Chain 1: 900 -7728.967 0.526 0.079
Chain 1: 1000 -7870.069 0.475 0.079
Chain 1: 1100 -7754.829 0.377 0.064
Chain 1: 1200 -7705.078 0.135 0.028
Chain 1: 1300 -7863.627 0.037 0.020
Chain 1: 1400 -7786.955 0.032 0.018
Chain 1: 1500 -7616.926 0.021 0.018
Chain 1: 1600 -7771.222 0.022 0.020
Chain 1: 1700 -7538.327 0.023 0.020
Chain 1: 1800 -7674.904 0.024 0.020
Chain 1: 1900 -7747.482 0.017 0.018
Chain 1: 2000 -7749.783 0.015 0.018
Chain 1: 2100 -7637.773 0.015 0.018
Chain 1: 2200 -7802.350 0.017 0.020
Chain 1: 2300 -7603.293 0.017 0.020
Chain 1: 2400 -7601.662 0.016 0.020
Chain 1: 2500 -7676.736 0.015 0.018
Chain 1: 2600 -7594.321 0.014 0.015
Chain 1: 2700 -7534.057 0.012 0.011
Chain 1: 2800 -7572.030 0.011 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002515 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.15 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85844.968 1.000 1.000
Chain 1: 200 -13834.582 3.103 5.205
Chain 1: 300 -10149.601 2.189 1.000
Chain 1: 400 -11317.176 1.668 1.000
Chain 1: 500 -9147.471 1.382 0.363
Chain 1: 600 -8788.777 1.158 0.363
Chain 1: 700 -8819.584 0.993 0.237
Chain 1: 800 -9391.169 0.877 0.237
Chain 1: 900 -8895.526 0.785 0.103
Chain 1: 1000 -8649.718 0.710 0.103
Chain 1: 1100 -8945.363 0.613 0.061 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8393.875 0.099 0.061
Chain 1: 1300 -8801.004 0.067 0.056
Chain 1: 1400 -8833.585 0.058 0.046
Chain 1: 1500 -8659.271 0.036 0.041
Chain 1: 1600 -8770.866 0.033 0.033
Chain 1: 1700 -8833.036 0.033 0.033
Chain 1: 1800 -8396.430 0.032 0.033
Chain 1: 1900 -8501.553 0.028 0.028
Chain 1: 2000 -8477.407 0.026 0.020
Chain 1: 2100 -8434.232 0.023 0.013
Chain 1: 2200 -8420.823 0.016 0.012
Chain 1: 2300 -8557.745 0.013 0.012
Chain 1: 2400 -8402.356 0.015 0.013
Chain 1: 2500 -8471.506 0.014 0.012
Chain 1: 2600 -8389.515 0.013 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003286 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.86 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8394705.870 1.000 1.000
Chain 1: 200 -1580304.011 2.656 4.312
Chain 1: 300 -890280.487 2.029 1.000
Chain 1: 400 -457886.418 1.758 1.000
Chain 1: 500 -358525.886 1.462 0.944
Chain 1: 600 -233586.635 1.307 0.944
Chain 1: 700 -119707.001 1.256 0.944
Chain 1: 800 -86932.610 1.146 0.944
Chain 1: 900 -67247.128 1.052 0.775
Chain 1: 1000 -52028.479 0.976 0.775
Chain 1: 1100 -39487.488 0.907 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38664.630 0.478 0.377
Chain 1: 1300 -26584.214 0.446 0.377
Chain 1: 1400 -26302.700 0.353 0.318
Chain 1: 1500 -22880.929 0.340 0.318
Chain 1: 1600 -22096.205 0.290 0.293
Chain 1: 1700 -20964.361 0.201 0.293
Chain 1: 1800 -20907.774 0.163 0.150
Chain 1: 1900 -21234.314 0.135 0.054
Chain 1: 2000 -19742.388 0.114 0.054
Chain 1: 2100 -19980.639 0.083 0.036
Chain 1: 2200 -20208.033 0.082 0.036
Chain 1: 2300 -19824.384 0.039 0.019
Chain 1: 2400 -19596.302 0.039 0.019
Chain 1: 2500 -19398.710 0.025 0.015
Chain 1: 2600 -19028.069 0.023 0.015
Chain 1: 2700 -18984.882 0.018 0.012
Chain 1: 2800 -18701.726 0.019 0.015
Chain 1: 2900 -18983.198 0.019 0.015
Chain 1: 3000 -18969.258 0.012 0.012
Chain 1: 3100 -19054.325 0.011 0.012
Chain 1: 3200 -18744.668 0.011 0.015
Chain 1: 3300 -18949.703 0.011 0.012
Chain 1: 3400 -18424.150 0.012 0.015
Chain 1: 3500 -19036.831 0.014 0.015
Chain 1: 3600 -18342.502 0.016 0.015
Chain 1: 3700 -18730.061 0.018 0.017
Chain 1: 3800 -17688.308 0.023 0.021
Chain 1: 3900 -17684.501 0.021 0.021
Chain 1: 4000 -17801.724 0.022 0.021
Chain 1: 4100 -17715.431 0.022 0.021
Chain 1: 4200 -17531.408 0.021 0.021
Chain 1: 4300 -17669.947 0.021 0.021
Chain 1: 4400 -17626.479 0.018 0.010
Chain 1: 4500 -17529.051 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001251 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.51 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48645.438 1.000 1.000
Chain 1: 200 -22586.428 1.077 1.154
Chain 1: 300 -25875.039 0.760 1.000
Chain 1: 400 -28777.918 0.595 1.000
Chain 1: 500 -22626.821 0.531 0.272
Chain 1: 600 -20939.191 0.456 0.272
Chain 1: 700 -11549.211 0.507 0.272
Chain 1: 800 -13801.358 0.464 0.272
Chain 1: 900 -16755.845 0.432 0.176
Chain 1: 1000 -13049.933 0.417 0.272
Chain 1: 1100 -22327.489 0.359 0.272
Chain 1: 1200 -10796.570 0.350 0.272
Chain 1: 1300 -11595.941 0.344 0.272
Chain 1: 1400 -9857.729 0.352 0.272
Chain 1: 1500 -16848.481 0.366 0.284
Chain 1: 1600 -11151.873 0.409 0.415
Chain 1: 1700 -9454.007 0.346 0.284
Chain 1: 1800 -13092.963 0.357 0.284
Chain 1: 1900 -9219.870 0.382 0.415
Chain 1: 2000 -14855.481 0.391 0.415
Chain 1: 2100 -9501.980 0.406 0.415
Chain 1: 2200 -10428.213 0.308 0.379
Chain 1: 2300 -9374.079 0.312 0.379
Chain 1: 2400 -10685.387 0.307 0.379
Chain 1: 2500 -9034.735 0.284 0.278
Chain 1: 2600 -9978.554 0.242 0.183
Chain 1: 2700 -10411.075 0.228 0.183
Chain 1: 2800 -9873.201 0.206 0.123
Chain 1: 2900 -9241.817 0.171 0.112
Chain 1: 3000 -9824.691 0.139 0.095
Chain 1: 3100 -9072.205 0.091 0.089
Chain 1: 3200 -9199.924 0.083 0.083
Chain 1: 3300 -9335.176 0.073 0.068
Chain 1: 3400 -9921.040 0.067 0.059
Chain 1: 3500 -8752.908 0.062 0.059
Chain 1: 3600 -9326.311 0.059 0.059
Chain 1: 3700 -8812.133 0.061 0.059
Chain 1: 3800 -9935.552 0.066 0.061
Chain 1: 3900 -11995.991 0.077 0.061
Chain 1: 4000 -8546.646 0.111 0.083
Chain 1: 4100 -13326.874 0.139 0.113
Chain 1: 4200 -8823.424 0.188 0.133
Chain 1: 4300 -12005.309 0.213 0.172
Chain 1: 4400 -10694.140 0.220 0.172
Chain 1: 4500 -8894.804 0.227 0.202
Chain 1: 4600 -11583.620 0.244 0.232
Chain 1: 4700 -9277.542 0.263 0.249
Chain 1: 4800 -8386.857 0.262 0.249
Chain 1: 4900 -13397.350 0.282 0.265
Chain 1: 5000 -9384.889 0.285 0.265
Chain 1: 5100 -16055.083 0.290 0.265
Chain 1: 5200 -8658.230 0.325 0.265
Chain 1: 5300 -9133.580 0.304 0.249
Chain 1: 5400 -11133.789 0.309 0.249
Chain 1: 5500 -9236.253 0.310 0.249
Chain 1: 5600 -10946.122 0.302 0.249
Chain 1: 5700 -12716.190 0.291 0.205
Chain 1: 5800 -8680.144 0.327 0.374
Chain 1: 5900 -11748.375 0.316 0.261
Chain 1: 6000 -10713.901 0.283 0.205
Chain 1: 6100 -8404.679 0.268 0.205
Chain 1: 6200 -8278.905 0.185 0.180
Chain 1: 6300 -8346.435 0.180 0.180
Chain 1: 6400 -11064.370 0.187 0.205
Chain 1: 6500 -12844.604 0.180 0.156
Chain 1: 6600 -8332.048 0.219 0.246
Chain 1: 6700 -9782.521 0.219 0.246
Chain 1: 6800 -10341.096 0.178 0.148
Chain 1: 6900 -8284.211 0.177 0.148
Chain 1: 7000 -10366.979 0.188 0.201
Chain 1: 7100 -8437.411 0.183 0.201
Chain 1: 7200 -8386.348 0.182 0.201
Chain 1: 7300 -9338.397 0.191 0.201
Chain 1: 7400 -10142.407 0.175 0.148
Chain 1: 7500 -11862.348 0.175 0.148
Chain 1: 7600 -9307.986 0.149 0.148
Chain 1: 7700 -8479.459 0.144 0.145
Chain 1: 7800 -11864.467 0.167 0.201
Chain 1: 7900 -8234.030 0.186 0.201
Chain 1: 8000 -8254.774 0.166 0.145
Chain 1: 8100 -11650.410 0.172 0.145
Chain 1: 8200 -8166.628 0.215 0.274
Chain 1: 8300 -8086.501 0.205 0.274
Chain 1: 8400 -8376.695 0.201 0.274
Chain 1: 8500 -8765.319 0.191 0.274
Chain 1: 8600 -8458.791 0.167 0.098
Chain 1: 8700 -8549.926 0.158 0.044
Chain 1: 8800 -8599.326 0.130 0.036
Chain 1: 8900 -12122.902 0.115 0.036
Chain 1: 9000 -9312.084 0.145 0.044
Chain 1: 9100 -8153.567 0.130 0.044
Chain 1: 9200 -8205.913 0.088 0.036
Chain 1: 9300 -8097.078 0.089 0.036
Chain 1: 9400 -11615.336 0.115 0.044
Chain 1: 9500 -9597.920 0.132 0.142
Chain 1: 9600 -9720.740 0.130 0.142
Chain 1: 9700 -10499.754 0.136 0.142
Chain 1: 9800 -8957.582 0.153 0.172
Chain 1: 9900 -9836.197 0.133 0.142
Chain 1: 10000 -10809.889 0.111 0.090
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001417 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.17 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57674.210 1.000 1.000
Chain 1: 200 -17442.605 1.653 2.307
Chain 1: 300 -8567.201 1.447 1.036
Chain 1: 400 -8120.534 1.099 1.036
Chain 1: 500 -8338.796 0.885 1.000
Chain 1: 600 -8910.575 0.748 1.000
Chain 1: 700 -7721.780 0.663 0.154
Chain 1: 800 -7948.643 0.584 0.154
Chain 1: 900 -7900.084 0.520 0.064
Chain 1: 1000 -7700.429 0.470 0.064
Chain 1: 1100 -7815.395 0.372 0.055
Chain 1: 1200 -7601.168 0.144 0.029
Chain 1: 1300 -7563.540 0.041 0.028
Chain 1: 1400 -7846.175 0.039 0.028
Chain 1: 1500 -7589.594 0.040 0.029
Chain 1: 1600 -7525.033 0.034 0.028
Chain 1: 1700 -7487.289 0.019 0.026
Chain 1: 1800 -7535.699 0.017 0.015
Chain 1: 1900 -7551.243 0.017 0.015
Chain 1: 2000 -7546.689 0.014 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004143 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 41.43 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86175.761 1.000 1.000
Chain 1: 200 -13261.039 3.249 5.498
Chain 1: 300 -9697.495 2.289 1.000
Chain 1: 400 -10506.797 1.736 1.000
Chain 1: 500 -8598.928 1.433 0.367
Chain 1: 600 -8261.433 1.201 0.367
Chain 1: 700 -8173.779 1.031 0.222
Chain 1: 800 -8633.366 0.909 0.222
Chain 1: 900 -8515.296 0.809 0.077
Chain 1: 1000 -8351.357 0.730 0.077
Chain 1: 1100 -8604.059 0.633 0.053 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8129.701 0.089 0.053
Chain 1: 1300 -8286.188 0.054 0.041
Chain 1: 1400 -8441.197 0.049 0.029
Chain 1: 1500 -8307.806 0.028 0.020
Chain 1: 1600 -8416.895 0.025 0.019
Chain 1: 1700 -8497.560 0.025 0.019
Chain 1: 1800 -8106.505 0.025 0.019
Chain 1: 1900 -8209.063 0.024 0.019
Chain 1: 2000 -8179.092 0.023 0.018
Chain 1: 2100 -8306.455 0.021 0.016
Chain 1: 2200 -8092.773 0.018 0.016
Chain 1: 2300 -8237.711 0.018 0.016
Chain 1: 2400 -8252.824 0.016 0.015
Chain 1: 2500 -8219.377 0.015 0.013
Chain 1: 2600 -8221.032 0.014 0.012
Chain 1: 2700 -8128.125 0.014 0.012
Chain 1: 2800 -8101.751 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003616 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 36.16 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8382358.032 1.000 1.000
Chain 1: 200 -1582872.855 2.648 4.296
Chain 1: 300 -890758.364 2.024 1.000
Chain 1: 400 -457294.904 1.755 1.000
Chain 1: 500 -357789.647 1.460 0.948
Chain 1: 600 -232811.138 1.306 0.948
Chain 1: 700 -119034.433 1.256 0.948
Chain 1: 800 -86196.396 1.147 0.948
Chain 1: 900 -66539.883 1.052 0.777
Chain 1: 1000 -51326.858 0.976 0.777
Chain 1: 1100 -38792.953 0.909 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37965.829 0.481 0.381
Chain 1: 1300 -25925.134 0.450 0.381
Chain 1: 1400 -25642.501 0.356 0.323
Chain 1: 1500 -22229.431 0.344 0.323
Chain 1: 1600 -21444.825 0.294 0.296
Chain 1: 1700 -20319.834 0.204 0.295
Chain 1: 1800 -20263.866 0.166 0.154
Chain 1: 1900 -20589.586 0.138 0.055
Chain 1: 2000 -19101.722 0.116 0.055
Chain 1: 2100 -19340.298 0.085 0.037
Chain 1: 2200 -19566.181 0.084 0.037
Chain 1: 2300 -19183.947 0.040 0.020
Chain 1: 2400 -18956.187 0.040 0.020
Chain 1: 2500 -18758.011 0.025 0.016
Chain 1: 2600 -18388.914 0.024 0.016
Chain 1: 2700 -18346.050 0.019 0.012
Chain 1: 2800 -18063.006 0.020 0.016
Chain 1: 2900 -18344.041 0.020 0.015
Chain 1: 3000 -18330.383 0.012 0.012
Chain 1: 3100 -18415.269 0.011 0.012
Chain 1: 3200 -18106.301 0.012 0.015
Chain 1: 3300 -18310.724 0.011 0.012
Chain 1: 3400 -17786.192 0.013 0.015
Chain 1: 3500 -18397.228 0.015 0.016
Chain 1: 3600 -17705.018 0.017 0.016
Chain 1: 3700 -18090.990 0.019 0.017
Chain 1: 3800 -17052.368 0.023 0.021
Chain 1: 3900 -17048.513 0.022 0.021
Chain 1: 4000 -17165.848 0.022 0.021
Chain 1: 4100 -17079.675 0.022 0.021
Chain 1: 4200 -16896.276 0.022 0.021
Chain 1: 4300 -17034.459 0.022 0.021
Chain 1: 4400 -16991.609 0.019 0.011
Chain 1: 4500 -16894.144 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004333 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 43.33 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12201.518 1.000 1.000
Chain 1: 200 -9127.683 0.668 1.000
Chain 1: 300 -7858.020 0.499 0.337
Chain 1: 400 -8045.933 0.380 0.337
Chain 1: 500 -7907.807 0.308 0.162
Chain 1: 600 -7815.678 0.258 0.162
Chain 1: 700 -7727.544 0.223 0.023
Chain 1: 800 -7767.674 0.196 0.023
Chain 1: 900 -7895.041 0.176 0.017
Chain 1: 1000 -7807.470 0.159 0.017
Chain 1: 1100 -7824.134 0.060 0.016
Chain 1: 1200 -7746.982 0.027 0.012
Chain 1: 1300 -7696.015 0.012 0.011
Chain 1: 1400 -7718.476 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001554 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.54 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56654.034 1.000 1.000
Chain 1: 200 -17200.783 1.647 2.294
Chain 1: 300 -8597.869 1.431 1.001
Chain 1: 400 -8213.005 1.085 1.001
Chain 1: 500 -8036.852 0.873 1.000
Chain 1: 600 -8567.295 0.737 1.000
Chain 1: 700 -7898.378 0.644 0.085
Chain 1: 800 -8188.608 0.568 0.085
Chain 1: 900 -7817.028 0.510 0.062
Chain 1: 1000 -7660.994 0.461 0.062
Chain 1: 1100 -7699.675 0.362 0.048
Chain 1: 1200 -7572.618 0.134 0.047
Chain 1: 1300 -7751.629 0.036 0.035
Chain 1: 1400 -7603.525 0.034 0.023
Chain 1: 1500 -7531.711 0.032 0.023
Chain 1: 1600 -7640.419 0.028 0.020
Chain 1: 1700 -7450.650 0.022 0.020
Chain 1: 1800 -7506.391 0.019 0.019
Chain 1: 1900 -7496.063 0.014 0.017
Chain 1: 2000 -7525.367 0.013 0.014
Chain 1: 2100 -7520.993 0.012 0.014
Chain 1: 2200 -7623.984 0.012 0.014
Chain 1: 2300 -7550.492 0.011 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003583 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.83 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86459.584 1.000 1.000
Chain 1: 200 -13278.140 3.256 5.511
Chain 1: 300 -9673.389 2.295 1.000
Chain 1: 400 -10395.511 1.738 1.000
Chain 1: 500 -8635.066 1.431 0.373
Chain 1: 600 -8277.708 1.200 0.373
Chain 1: 700 -8275.925 1.029 0.204
Chain 1: 800 -8570.916 0.904 0.204
Chain 1: 900 -8414.131 0.806 0.069
Chain 1: 1000 -8259.282 0.727 0.069
Chain 1: 1100 -8374.601 0.629 0.043 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8185.672 0.080 0.034
Chain 1: 1300 -8392.074 0.045 0.025
Chain 1: 1400 -8377.686 0.038 0.023
Chain 1: 1500 -8251.593 0.019 0.019
Chain 1: 1600 -8361.194 0.016 0.019
Chain 1: 1700 -8448.022 0.017 0.019
Chain 1: 1800 -8042.982 0.019 0.019
Chain 1: 1900 -8140.321 0.018 0.015
Chain 1: 2000 -8112.173 0.017 0.014
Chain 1: 2100 -8232.397 0.017 0.015
Chain 1: 2200 -8028.663 0.017 0.015
Chain 1: 2300 -8176.775 0.016 0.015
Chain 1: 2400 -8183.637 0.016 0.015
Chain 1: 2500 -8153.895 0.015 0.013
Chain 1: 2600 -8152.519 0.014 0.012
Chain 1: 2700 -8065.090 0.014 0.012
Chain 1: 2800 -8030.992 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003108 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.08 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8403593.642 1.000 1.000
Chain 1: 200 -1583428.509 2.654 4.307
Chain 1: 300 -889876.481 2.029 1.000
Chain 1: 400 -456813.876 1.759 1.000
Chain 1: 500 -357157.143 1.463 0.948
Chain 1: 600 -232397.472 1.308 0.948
Chain 1: 700 -118819.179 1.258 0.948
Chain 1: 800 -86082.940 1.148 0.948
Chain 1: 900 -66461.576 1.054 0.779
Chain 1: 1000 -51285.258 0.978 0.779
Chain 1: 1100 -38785.005 0.910 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37964.504 0.481 0.380
Chain 1: 1300 -25943.522 0.450 0.380
Chain 1: 1400 -25664.586 0.356 0.322
Chain 1: 1500 -22257.643 0.344 0.322
Chain 1: 1600 -21475.813 0.293 0.296
Chain 1: 1700 -20352.171 0.203 0.295
Chain 1: 1800 -20296.882 0.166 0.153
Chain 1: 1900 -20622.892 0.138 0.055
Chain 1: 2000 -19135.811 0.116 0.055
Chain 1: 2100 -19374.088 0.085 0.036
Chain 1: 2200 -19600.208 0.084 0.036
Chain 1: 2300 -19217.760 0.040 0.020
Chain 1: 2400 -18989.930 0.040 0.020
Chain 1: 2500 -18791.880 0.025 0.016
Chain 1: 2600 -18422.329 0.024 0.016
Chain 1: 2700 -18379.420 0.018 0.012
Chain 1: 2800 -18096.329 0.020 0.016
Chain 1: 2900 -18377.470 0.020 0.015
Chain 1: 3000 -18363.700 0.012 0.012
Chain 1: 3100 -18448.636 0.011 0.012
Chain 1: 3200 -18139.469 0.012 0.015
Chain 1: 3300 -18344.101 0.011 0.012
Chain 1: 3400 -17819.262 0.013 0.015
Chain 1: 3500 -18430.732 0.015 0.016
Chain 1: 3600 -17737.965 0.017 0.016
Chain 1: 3700 -18124.334 0.019 0.017
Chain 1: 3800 -17084.856 0.023 0.021
Chain 1: 3900 -17081.015 0.022 0.021
Chain 1: 4000 -17198.327 0.022 0.021
Chain 1: 4100 -17112.100 0.022 0.021
Chain 1: 4200 -16928.557 0.022 0.021
Chain 1: 4300 -17066.820 0.021 0.021
Chain 1: 4400 -17023.797 0.019 0.011
Chain 1: 4500 -16926.343 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001308 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.08 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12826.751 1.000 1.000
Chain 1: 200 -9719.596 0.660 1.000
Chain 1: 300 -8228.300 0.500 0.320
Chain 1: 400 -8466.957 0.382 0.320
Chain 1: 500 -8354.919 0.309 0.181
Chain 1: 600 -8251.790 0.259 0.181
Chain 1: 700 -8060.146 0.226 0.028
Chain 1: 800 -8057.285 0.197 0.028
Chain 1: 900 -8124.655 0.176 0.024
Chain 1: 1000 -8172.879 0.159 0.024
Chain 1: 1100 -8142.966 0.060 0.013
Chain 1: 1200 -8088.463 0.028 0.012
Chain 1: 1300 -8007.121 0.011 0.010
Chain 1: 1400 -8032.222 0.009 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001497 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.97 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -46979.582 1.000 1.000
Chain 1: 200 -16095.276 1.459 1.919
Chain 1: 300 -8991.120 1.236 1.000
Chain 1: 400 -8278.805 0.949 1.000
Chain 1: 500 -8730.817 0.769 0.790
Chain 1: 600 -8743.160 0.641 0.790
Chain 1: 700 -8631.020 0.552 0.086
Chain 1: 800 -8623.485 0.483 0.086
Chain 1: 900 -8243.582 0.434 0.052
Chain 1: 1000 -8011.780 0.394 0.052
Chain 1: 1100 -7989.483 0.294 0.046
Chain 1: 1200 -7775.039 0.105 0.029
Chain 1: 1300 -7830.975 0.027 0.028
Chain 1: 1400 -7768.692 0.019 0.013
Chain 1: 1500 -7719.113 0.014 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003668 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 36.68 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86503.804 1.000 1.000
Chain 1: 200 -14019.489 3.085 5.170
Chain 1: 300 -10230.612 2.180 1.000
Chain 1: 400 -12007.766 1.672 1.000
Chain 1: 500 -8698.404 1.414 0.380
Chain 1: 600 -8760.509 1.179 0.380
Chain 1: 700 -9086.442 1.016 0.370
Chain 1: 800 -8789.047 0.893 0.370
Chain 1: 900 -8945.678 0.796 0.148
Chain 1: 1000 -9213.647 0.719 0.148
Chain 1: 1100 -8928.098 0.622 0.036 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8632.147 0.109 0.034
Chain 1: 1300 -8858.463 0.074 0.034
Chain 1: 1400 -8687.016 0.062 0.032
Chain 1: 1500 -8691.945 0.024 0.029
Chain 1: 1600 -8804.502 0.024 0.029
Chain 1: 1700 -8849.570 0.021 0.026
Chain 1: 1800 -8391.076 0.023 0.026
Chain 1: 1900 -8502.848 0.023 0.026
Chain 1: 2000 -8516.897 0.020 0.020
Chain 1: 2100 -8608.146 0.018 0.013
Chain 1: 2200 -8392.588 0.017 0.013
Chain 1: 2300 -8600.589 0.017 0.013
Chain 1: 2400 -8397.900 0.017 0.013
Chain 1: 2500 -8474.610 0.018 0.013
Chain 1: 2600 -8384.548 0.018 0.013
Chain 1: 2700 -8417.596 0.018 0.013
Chain 1: 2800 -8368.651 0.013 0.011
Chain 1: 2900 -8483.082 0.013 0.011
Chain 1: 3000 -8399.648 0.014 0.011
Chain 1: 3100 -8360.953 0.013 0.011
Chain 1: 3200 -8333.201 0.011 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003704 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 37.04 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8366451.121 1.000 1.000
Chain 1: 200 -1580854.599 2.646 4.292
Chain 1: 300 -891120.599 2.022 1.000
Chain 1: 400 -457919.656 1.753 1.000
Chain 1: 500 -358732.721 1.458 0.946
Chain 1: 600 -233781.998 1.304 0.946
Chain 1: 700 -119974.015 1.253 0.946
Chain 1: 800 -87125.320 1.144 0.946
Chain 1: 900 -67457.281 1.049 0.774
Chain 1: 1000 -52242.241 0.973 0.774
Chain 1: 1100 -39690.385 0.905 0.534 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38875.332 0.478 0.377
Chain 1: 1300 -26789.794 0.445 0.377
Chain 1: 1400 -26509.417 0.352 0.316
Chain 1: 1500 -23083.440 0.339 0.316
Chain 1: 1600 -22297.248 0.289 0.292
Chain 1: 1700 -21165.265 0.200 0.291
Chain 1: 1800 -21108.748 0.162 0.148
Chain 1: 1900 -21435.718 0.135 0.053
Chain 1: 2000 -19942.078 0.113 0.053
Chain 1: 2100 -20181.010 0.082 0.035
Chain 1: 2200 -20408.272 0.081 0.035
Chain 1: 2300 -20024.532 0.038 0.019
Chain 1: 2400 -19796.251 0.038 0.019
Chain 1: 2500 -19598.263 0.025 0.015
Chain 1: 2600 -19227.640 0.023 0.015
Chain 1: 2700 -19184.426 0.018 0.012
Chain 1: 2800 -18900.808 0.019 0.015
Chain 1: 2900 -19182.600 0.019 0.015
Chain 1: 3000 -19168.743 0.012 0.012
Chain 1: 3100 -19253.789 0.011 0.012
Chain 1: 3200 -18943.980 0.011 0.015
Chain 1: 3300 -19149.147 0.011 0.012
Chain 1: 3400 -18623.048 0.012 0.015
Chain 1: 3500 -19236.458 0.014 0.015
Chain 1: 3600 -18541.265 0.016 0.015
Chain 1: 3700 -18929.419 0.018 0.016
Chain 1: 3800 -17886.157 0.022 0.021
Chain 1: 3900 -17882.236 0.021 0.021
Chain 1: 4000 -17999.551 0.021 0.021
Chain 1: 4100 -17913.074 0.022 0.021
Chain 1: 4200 -17728.744 0.021 0.021
Chain 1: 4300 -17867.578 0.021 0.021
Chain 1: 4400 -17823.884 0.018 0.010
Chain 1: 4500 -17726.313 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001434 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.34 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12612.061 1.000 1.000
Chain 1: 200 -9420.375 0.669 1.000
Chain 1: 300 -8027.783 0.504 0.339
Chain 1: 400 -8157.600 0.382 0.339
Chain 1: 500 -7992.837 0.310 0.173
Chain 1: 600 -7920.151 0.260 0.173
Chain 1: 700 -7813.470 0.225 0.021
Chain 1: 800 -7820.037 0.197 0.021
Chain 1: 900 -7764.738 0.176 0.016
Chain 1: 1000 -7943.359 0.160 0.021
Chain 1: 1100 -7953.818 0.060 0.016
Chain 1: 1200 -7836.571 0.028 0.015
Chain 1: 1300 -7803.185 0.011 0.014
Chain 1: 1400 -7814.513 0.010 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001607 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 16.07 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49619.008 1.000 1.000
Chain 1: 200 -16007.986 1.550 2.100
Chain 1: 300 -8722.820 1.312 1.000
Chain 1: 400 -8533.102 0.989 1.000
Chain 1: 500 -8528.345 0.792 0.835
Chain 1: 600 -8464.169 0.661 0.835
Chain 1: 700 -7717.465 0.580 0.097
Chain 1: 800 -8063.742 0.513 0.097
Chain 1: 900 -7867.077 0.459 0.043
Chain 1: 1000 -7847.902 0.413 0.043
Chain 1: 1100 -7850.754 0.313 0.025
Chain 1: 1200 -7653.051 0.106 0.025
Chain 1: 1300 -7783.814 0.024 0.022
Chain 1: 1400 -7706.453 0.023 0.017
Chain 1: 1500 -7546.305 0.025 0.021
Chain 1: 1600 -7829.320 0.028 0.025
Chain 1: 1700 -7581.724 0.021 0.025
Chain 1: 1800 -7653.762 0.018 0.021
Chain 1: 1900 -7640.066 0.016 0.017
Chain 1: 2000 -7624.343 0.016 0.017
Chain 1: 2100 -7581.009 0.016 0.017
Chain 1: 2200 -7722.795 0.015 0.017
Chain 1: 2300 -7603.794 0.015 0.016
Chain 1: 2400 -7639.959 0.015 0.016
Chain 1: 2500 -7594.639 0.013 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003102 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.02 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87374.811 1.000 1.000
Chain 1: 200 -13567.796 3.220 5.440
Chain 1: 300 -9862.900 2.272 1.000
Chain 1: 400 -10937.629 1.728 1.000
Chain 1: 500 -8854.643 1.430 0.376
Chain 1: 600 -8328.572 1.202 0.376
Chain 1: 700 -8753.893 1.037 0.235
Chain 1: 800 -9477.540 0.917 0.235
Chain 1: 900 -8622.844 0.826 0.099
Chain 1: 1000 -8300.568 0.748 0.099
Chain 1: 1100 -8691.761 0.652 0.098 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8249.351 0.113 0.076
Chain 1: 1300 -8559.845 0.079 0.063
Chain 1: 1400 -8507.130 0.070 0.054
Chain 1: 1500 -8401.073 0.048 0.049
Chain 1: 1600 -8510.212 0.043 0.045
Chain 1: 1700 -8581.213 0.039 0.039
Chain 1: 1800 -8150.267 0.037 0.039
Chain 1: 1900 -8254.418 0.028 0.036
Chain 1: 2000 -8229.552 0.024 0.013
Chain 1: 2100 -8366.523 0.021 0.013
Chain 1: 2200 -8159.534 0.019 0.013
Chain 1: 2300 -8258.557 0.016 0.013
Chain 1: 2400 -8321.233 0.016 0.013
Chain 1: 2500 -8261.890 0.016 0.013
Chain 1: 2600 -8267.613 0.015 0.012
Chain 1: 2700 -8182.391 0.015 0.012
Chain 1: 2800 -8138.737 0.010 0.010
Chain 1: 2900 -8227.619 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.005015 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 50.15 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8427149.376 1.000 1.000
Chain 1: 200 -1588641.066 2.652 4.305
Chain 1: 300 -891012.664 2.029 1.000
Chain 1: 400 -457142.176 1.759 1.000
Chain 1: 500 -357156.855 1.463 0.949
Chain 1: 600 -232077.091 1.309 0.949
Chain 1: 700 -118804.313 1.258 0.949
Chain 1: 800 -86119.587 1.149 0.949
Chain 1: 900 -66564.315 1.054 0.783
Chain 1: 1000 -51446.894 0.978 0.783
Chain 1: 1100 -38997.872 0.910 0.539 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38188.473 0.481 0.380
Chain 1: 1300 -26217.869 0.449 0.380
Chain 1: 1400 -25945.048 0.355 0.319
Chain 1: 1500 -22550.612 0.342 0.319
Chain 1: 1600 -21772.839 0.291 0.294
Chain 1: 1700 -20655.313 0.202 0.294
Chain 1: 1800 -20601.699 0.164 0.151
Chain 1: 1900 -20928.137 0.136 0.054
Chain 1: 2000 -19443.167 0.114 0.054
Chain 1: 2100 -19681.498 0.084 0.036
Chain 1: 2200 -19907.292 0.083 0.036
Chain 1: 2300 -19524.995 0.039 0.020
Chain 1: 2400 -19297.087 0.039 0.020
Chain 1: 2500 -19098.624 0.025 0.016
Chain 1: 2600 -18728.999 0.023 0.016
Chain 1: 2700 -18686.060 0.018 0.012
Chain 1: 2800 -18402.586 0.019 0.015
Chain 1: 2900 -18683.868 0.019 0.015
Chain 1: 3000 -18670.142 0.012 0.012
Chain 1: 3100 -18755.132 0.011 0.012
Chain 1: 3200 -18445.737 0.012 0.015
Chain 1: 3300 -18650.540 0.011 0.012
Chain 1: 3400 -18125.140 0.012 0.015
Chain 1: 3500 -18737.311 0.015 0.015
Chain 1: 3600 -18043.598 0.017 0.015
Chain 1: 3700 -18430.627 0.018 0.017
Chain 1: 3800 -17389.565 0.023 0.021
Chain 1: 3900 -17385.617 0.021 0.021
Chain 1: 4000 -17503.002 0.022 0.021
Chain 1: 4100 -17416.667 0.022 0.021
Chain 1: 4200 -17232.777 0.021 0.021
Chain 1: 4300 -17371.330 0.021 0.021
Chain 1: 4400 -17328.035 0.019 0.011
Chain 1: 4500 -17230.473 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001297 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.97 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48044.270 1.000 1.000
Chain 1: 200 -14345.258 1.675 2.349
Chain 1: 300 -16210.682 1.155 1.000
Chain 1: 400 -11545.010 0.967 1.000
Chain 1: 500 -19784.742 0.857 0.416
Chain 1: 600 -17382.165 0.737 0.416
Chain 1: 700 -25988.964 0.679 0.404
Chain 1: 800 -21321.855 0.622 0.404
Chain 1: 900 -12560.840 0.630 0.404
Chain 1: 1000 -12368.422 0.569 0.404
Chain 1: 1100 -17172.325 0.497 0.331
Chain 1: 1200 -13924.642 0.285 0.280
Chain 1: 1300 -11054.232 0.299 0.280
Chain 1: 1400 -9764.443 0.272 0.260
Chain 1: 1500 -9530.056 0.233 0.233
Chain 1: 1600 -11379.520 0.235 0.233
Chain 1: 1700 -18184.128 0.240 0.233
Chain 1: 1800 -14471.494 0.244 0.257
Chain 1: 1900 -8745.366 0.239 0.257
Chain 1: 2000 -15791.916 0.282 0.260
Chain 1: 2100 -9324.857 0.324 0.260
Chain 1: 2200 -9048.873 0.303 0.260
Chain 1: 2300 -8656.164 0.282 0.257
Chain 1: 2400 -8635.786 0.269 0.257
Chain 1: 2500 -14349.823 0.306 0.374
Chain 1: 2600 -8756.518 0.354 0.398
Chain 1: 2700 -8331.252 0.322 0.398
Chain 1: 2800 -16447.658 0.345 0.446
Chain 1: 2900 -8780.525 0.367 0.446
Chain 1: 3000 -13494.552 0.358 0.398
Chain 1: 3100 -11669.289 0.304 0.349
Chain 1: 3200 -12526.336 0.308 0.349
Chain 1: 3300 -9977.202 0.329 0.349
Chain 1: 3400 -9308.443 0.336 0.349
Chain 1: 3500 -8616.366 0.304 0.255
Chain 1: 3600 -8764.576 0.242 0.156
Chain 1: 3700 -8168.398 0.244 0.156
Chain 1: 3800 -11777.606 0.225 0.156
Chain 1: 3900 -8875.455 0.171 0.156
Chain 1: 4000 -8727.287 0.137 0.080
Chain 1: 4100 -8428.439 0.125 0.073
Chain 1: 4200 -8248.879 0.121 0.073
Chain 1: 4300 -11435.912 0.123 0.073
Chain 1: 4400 -8136.294 0.156 0.080
Chain 1: 4500 -10459.085 0.170 0.222
Chain 1: 4600 -9194.802 0.182 0.222
Chain 1: 4700 -8087.073 0.189 0.222
Chain 1: 4800 -10188.628 0.179 0.206
Chain 1: 4900 -9635.248 0.152 0.137
Chain 1: 5000 -9734.555 0.151 0.137
Chain 1: 5100 -8082.496 0.168 0.204
Chain 1: 5200 -10499.308 0.189 0.206
Chain 1: 5300 -12429.550 0.177 0.204
Chain 1: 5400 -8440.886 0.183 0.204
Chain 1: 5500 -7983.740 0.167 0.155
Chain 1: 5600 -12555.364 0.189 0.204
Chain 1: 5700 -12452.507 0.177 0.204
Chain 1: 5800 -8535.028 0.202 0.204
Chain 1: 5900 -7973.727 0.203 0.204
Chain 1: 6000 -9157.418 0.215 0.204
Chain 1: 6100 -8070.453 0.208 0.155
Chain 1: 6200 -8771.063 0.193 0.135
Chain 1: 6300 -11604.747 0.202 0.135
Chain 1: 6400 -11999.317 0.158 0.129
Chain 1: 6500 -8625.980 0.191 0.135
Chain 1: 6600 -8239.793 0.160 0.129
Chain 1: 6700 -8087.298 0.161 0.129
Chain 1: 6800 -9708.955 0.132 0.129
Chain 1: 6900 -8403.157 0.140 0.135
Chain 1: 7000 -7958.541 0.133 0.135
Chain 1: 7100 -8097.809 0.121 0.080
Chain 1: 7200 -7821.293 0.116 0.056
Chain 1: 7300 -7697.706 0.094 0.047
Chain 1: 7400 -8255.327 0.097 0.056
Chain 1: 7500 -7660.089 0.066 0.056
Chain 1: 7600 -10500.431 0.088 0.068
Chain 1: 7700 -7863.900 0.120 0.078
Chain 1: 7800 -9267.755 0.118 0.078
Chain 1: 7900 -7962.187 0.119 0.078
Chain 1: 8000 -7599.724 0.118 0.078
Chain 1: 8100 -7812.813 0.119 0.078
Chain 1: 8200 -7730.399 0.117 0.078
Chain 1: 8300 -9240.983 0.132 0.151
Chain 1: 8400 -8866.805 0.129 0.151
Chain 1: 8500 -7829.303 0.135 0.151
Chain 1: 8600 -10957.498 0.136 0.151
Chain 1: 8700 -10106.098 0.111 0.133
Chain 1: 8800 -7653.210 0.128 0.133
Chain 1: 8900 -7930.643 0.115 0.084
Chain 1: 9000 -8654.223 0.118 0.084
Chain 1: 9100 -8277.337 0.120 0.084
Chain 1: 9200 -8954.989 0.127 0.084
Chain 1: 9300 -10589.984 0.126 0.084
Chain 1: 9400 -11169.234 0.127 0.084
Chain 1: 9500 -7723.226 0.158 0.084
Chain 1: 9600 -7730.647 0.130 0.084
Chain 1: 9700 -8693.568 0.132 0.084
Chain 1: 9800 -7732.562 0.113 0.084
Chain 1: 9900 -9133.488 0.125 0.111
Chain 1: 10000 -9227.398 0.117 0.111
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002017 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 20.17 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56200.214 1.000 1.000
Chain 1: 200 -16617.793 1.691 2.382
Chain 1: 300 -8355.006 1.457 1.000
Chain 1: 400 -7981.017 1.104 1.000
Chain 1: 500 -8195.767 0.889 0.989
Chain 1: 600 -8707.561 0.750 0.989
Chain 1: 700 -7702.051 0.662 0.131
Chain 1: 800 -7855.316 0.582 0.131
Chain 1: 900 -7498.052 0.522 0.059
Chain 1: 1000 -7593.094 0.471 0.059
Chain 1: 1100 -7537.426 0.372 0.048
Chain 1: 1200 -7495.712 0.134 0.047
Chain 1: 1300 -7589.902 0.037 0.026
Chain 1: 1400 -7737.071 0.034 0.020
Chain 1: 1500 -7519.289 0.034 0.020
Chain 1: 1600 -7433.108 0.030 0.019
Chain 1: 1700 -7415.260 0.017 0.013
Chain 1: 1800 -7449.665 0.015 0.012
Chain 1: 1900 -7498.510 0.011 0.012
Chain 1: 2000 -7487.738 0.010 0.007 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003492 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.92 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85711.100 1.000 1.000
Chain 1: 200 -12694.681 3.376 5.752
Chain 1: 300 -9226.268 2.376 1.000
Chain 1: 400 -9758.897 1.796 1.000
Chain 1: 500 -8140.535 1.476 0.376
Chain 1: 600 -8231.541 1.232 0.376
Chain 1: 700 -8049.200 1.059 0.199
Chain 1: 800 -8190.115 0.929 0.199
Chain 1: 900 -8154.953 0.826 0.055
Chain 1: 1000 -7876.792 0.747 0.055
Chain 1: 1100 -8136.396 0.650 0.035 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -7859.374 0.079 0.035
Chain 1: 1300 -8036.920 0.043 0.032
Chain 1: 1400 -7964.677 0.039 0.023
Chain 1: 1500 -7912.646 0.020 0.022
Chain 1: 1600 -7904.814 0.019 0.022
Chain 1: 1700 -7855.100 0.017 0.017
Chain 1: 1800 -7734.601 0.017 0.016
Chain 1: 1900 -7843.977 0.018 0.016
Chain 1: 2000 -7810.050 0.015 0.014
Chain 1: 2100 -7959.063 0.013 0.014
Chain 1: 2200 -7736.272 0.013 0.014
Chain 1: 2300 -7817.517 0.011 0.010
Chain 1: 2400 -7882.087 0.011 0.010
Chain 1: 2500 -7845.259 0.011 0.010
Chain 1: 2600 -7838.000 0.011 0.010
Chain 1: 2700 -7750.412 0.012 0.011
Chain 1: 2800 -7738.739 0.010 0.010
Chain 1: 2900 -7743.668 0.009 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004088 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 40.88 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8414654.313 1.000 1.000
Chain 1: 200 -1587973.079 2.649 4.299
Chain 1: 300 -891290.255 2.027 1.000
Chain 1: 400 -457258.166 1.757 1.000
Chain 1: 500 -357181.020 1.462 0.949
Chain 1: 600 -231950.906 1.308 0.949
Chain 1: 700 -118245.884 1.259 0.949
Chain 1: 800 -85464.873 1.149 0.949
Chain 1: 900 -65830.173 1.055 0.782
Chain 1: 1000 -50635.973 0.979 0.782
Chain 1: 1100 -38134.762 0.912 0.540 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37304.306 0.484 0.384
Chain 1: 1300 -25305.031 0.454 0.384
Chain 1: 1400 -25022.289 0.360 0.328
Chain 1: 1500 -21622.554 0.348 0.328
Chain 1: 1600 -20841.091 0.297 0.300
Chain 1: 1700 -19721.632 0.207 0.298
Chain 1: 1800 -19666.612 0.169 0.157
Chain 1: 1900 -19991.692 0.141 0.057
Chain 1: 2000 -18508.346 0.119 0.057
Chain 1: 2100 -18746.303 0.087 0.037
Chain 1: 2200 -18971.522 0.086 0.037
Chain 1: 2300 -18590.088 0.041 0.021
Chain 1: 2400 -18362.671 0.041 0.021
Chain 1: 2500 -18164.480 0.026 0.016
Chain 1: 2600 -17795.957 0.025 0.016
Chain 1: 2700 -17753.275 0.019 0.013
Chain 1: 2800 -17470.567 0.020 0.016
Chain 1: 2900 -17751.239 0.020 0.016
Chain 1: 3000 -17737.535 0.012 0.013
Chain 1: 3100 -17822.365 0.012 0.012
Chain 1: 3200 -17513.788 0.012 0.016
Chain 1: 3300 -17717.919 0.011 0.012
Chain 1: 3400 -17194.131 0.013 0.016
Chain 1: 3500 -17803.993 0.015 0.016
Chain 1: 3600 -17113.295 0.017 0.016
Chain 1: 3700 -17498.139 0.019 0.018
Chain 1: 3800 -16461.850 0.024 0.022
Chain 1: 3900 -16458.081 0.022 0.022
Chain 1: 4000 -16575.401 0.023 0.022
Chain 1: 4100 -16489.364 0.023 0.022
Chain 1: 4200 -16306.486 0.023 0.022
Chain 1: 4300 -16444.274 0.022 0.022
Chain 1: 4400 -16401.814 0.019 0.011
Chain 1: 4500 -16304.481 0.017 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001303 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.03 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48877.479 1.000 1.000
Chain 1: 200 -14625.798 1.671 2.342
Chain 1: 300 -20800.116 1.213 1.000
Chain 1: 400 -72398.294 1.088 1.000
Chain 1: 500 -11267.508 1.955 1.000
Chain 1: 600 -16180.892 1.680 1.000
Chain 1: 700 -10888.672 1.509 0.713
Chain 1: 800 -13834.715 1.347 0.713
Chain 1: 900 -17174.524 1.219 0.486
Chain 1: 1000 -13042.300 1.129 0.486
Chain 1: 1100 -10003.503 1.059 0.317 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -11537.349 0.839 0.304 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1300 -12212.865 0.814 0.304 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1400 -12023.556 0.745 0.304 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1500 -9990.220 0.223 0.213
Chain 1: 1600 -15344.862 0.227 0.213
Chain 1: 1700 -15721.857 0.181 0.204
Chain 1: 1800 -18298.563 0.174 0.194
Chain 1: 1900 -9852.828 0.240 0.204
Chain 1: 2000 -16366.946 0.248 0.204
Chain 1: 2100 -9002.054 0.299 0.204
Chain 1: 2200 -13982.387 0.322 0.349
Chain 1: 2300 -13262.671 0.322 0.349
Chain 1: 2400 -9121.859 0.366 0.356
Chain 1: 2500 -15254.671 0.385 0.398
Chain 1: 2600 -9034.642 0.419 0.402
Chain 1: 2700 -13855.118 0.452 0.402
Chain 1: 2800 -9766.158 0.479 0.419
Chain 1: 2900 -9414.460 0.397 0.402
Chain 1: 3000 -9398.652 0.358 0.402
Chain 1: 3100 -12466.018 0.301 0.356
Chain 1: 3200 -8439.141 0.313 0.402
Chain 1: 3300 -9593.246 0.319 0.402
Chain 1: 3400 -12273.513 0.296 0.348
Chain 1: 3500 -13503.821 0.265 0.246
Chain 1: 3600 -9459.812 0.239 0.246
Chain 1: 3700 -8695.446 0.213 0.218
Chain 1: 3800 -8588.899 0.172 0.120
Chain 1: 3900 -8641.284 0.169 0.120
Chain 1: 4000 -8493.666 0.170 0.120
Chain 1: 4100 -9114.456 0.153 0.091
Chain 1: 4200 -9896.154 0.113 0.088
Chain 1: 4300 -9608.314 0.104 0.079
Chain 1: 4400 -8192.874 0.099 0.079
Chain 1: 4500 -8645.480 0.095 0.068
Chain 1: 4600 -8336.438 0.056 0.052
Chain 1: 4700 -11774.540 0.077 0.052
Chain 1: 4800 -10392.213 0.089 0.068
Chain 1: 4900 -14078.125 0.114 0.079
Chain 1: 5000 -16100.250 0.125 0.126
Chain 1: 5100 -8271.058 0.213 0.133
Chain 1: 5200 -8461.972 0.207 0.133
Chain 1: 5300 -8972.534 0.210 0.133
Chain 1: 5400 -11959.710 0.218 0.133
Chain 1: 5500 -8713.305 0.250 0.250
Chain 1: 5600 -8210.822 0.252 0.250
Chain 1: 5700 -10039.567 0.241 0.182
Chain 1: 5800 -8525.949 0.246 0.182
Chain 1: 5900 -12835.121 0.253 0.182
Chain 1: 6000 -9349.850 0.278 0.250
Chain 1: 6100 -9547.244 0.185 0.182
Chain 1: 6200 -12975.263 0.209 0.250
Chain 1: 6300 -12298.264 0.209 0.250
Chain 1: 6400 -12123.389 0.186 0.182
Chain 1: 6500 -14664.399 0.166 0.178
Chain 1: 6600 -9136.146 0.220 0.182
Chain 1: 6700 -9820.303 0.209 0.178
Chain 1: 6800 -8140.586 0.212 0.206
Chain 1: 6900 -9510.830 0.193 0.173
Chain 1: 7000 -14336.092 0.189 0.173
Chain 1: 7100 -8434.217 0.257 0.206
Chain 1: 7200 -9787.860 0.244 0.173
Chain 1: 7300 -10801.047 0.248 0.173
Chain 1: 7400 -9005.589 0.267 0.199
Chain 1: 7500 -8204.039 0.259 0.199
Chain 1: 7600 -8682.376 0.204 0.144
Chain 1: 7700 -8808.842 0.199 0.144
Chain 1: 7800 -8182.782 0.186 0.138
Chain 1: 7900 -8067.682 0.173 0.098
Chain 1: 8000 -8437.887 0.143 0.094
Chain 1: 8100 -8168.208 0.077 0.077
Chain 1: 8200 -12385.661 0.097 0.077
Chain 1: 8300 -8617.678 0.131 0.077
Chain 1: 8400 -8236.029 0.116 0.055
Chain 1: 8500 -8084.774 0.108 0.046
Chain 1: 8600 -8693.217 0.109 0.046
Chain 1: 8700 -8246.278 0.113 0.054
Chain 1: 8800 -7959.416 0.109 0.046
Chain 1: 8900 -8284.337 0.112 0.046
Chain 1: 9000 -9665.306 0.122 0.054
Chain 1: 9100 -8053.832 0.139 0.070
Chain 1: 9200 -8691.508 0.112 0.070
Chain 1: 9300 -8346.342 0.072 0.054
Chain 1: 9400 -11281.732 0.094 0.070
Chain 1: 9500 -8017.540 0.132 0.073
Chain 1: 9600 -8047.745 0.126 0.073
Chain 1: 9700 -8194.218 0.122 0.073
Chain 1: 9800 -8494.340 0.122 0.073
Chain 1: 9900 -9706.143 0.131 0.125
Chain 1: 10000 -8377.378 0.132 0.125
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001391 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.91 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56755.481 1.000 1.000
Chain 1: 200 -17101.567 1.659 2.319
Chain 1: 300 -8554.654 1.439 1.000
Chain 1: 400 -7807.572 1.103 1.000
Chain 1: 500 -8452.891 0.898 0.999
Chain 1: 600 -7963.752 0.759 0.999
Chain 1: 700 -8045.430 0.652 0.096
Chain 1: 800 -8067.500 0.571 0.096
Chain 1: 900 -7722.878 0.512 0.076
Chain 1: 1000 -7752.809 0.461 0.076
Chain 1: 1100 -7568.353 0.364 0.061
Chain 1: 1200 -7592.850 0.132 0.045
Chain 1: 1300 -7670.122 0.033 0.024
Chain 1: 1400 -7759.379 0.025 0.012
Chain 1: 1500 -7537.007 0.020 0.012
Chain 1: 1600 -7496.798 0.015 0.010
Chain 1: 1700 -7435.648 0.014 0.010
Chain 1: 1800 -7508.018 0.015 0.010
Chain 1: 1900 -7495.716 0.011 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003261 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.61 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86241.617 1.000 1.000
Chain 1: 200 -13175.923 3.273 5.545
Chain 1: 300 -9618.878 2.305 1.000
Chain 1: 400 -10619.720 1.752 1.000
Chain 1: 500 -8537.443 1.451 0.370
Chain 1: 600 -8127.811 1.217 0.370
Chain 1: 700 -8214.554 1.045 0.244
Chain 1: 800 -8444.396 0.918 0.244
Chain 1: 900 -8453.937 0.816 0.094
Chain 1: 1000 -8195.313 0.737 0.094
Chain 1: 1100 -8494.059 0.641 0.050 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8321.591 0.088 0.035
Chain 1: 1300 -8388.261 0.052 0.032
Chain 1: 1400 -8346.709 0.043 0.027
Chain 1: 1500 -8232.909 0.020 0.021
Chain 1: 1600 -8332.155 0.017 0.014
Chain 1: 1700 -8415.431 0.016 0.014
Chain 1: 1800 -8024.723 0.019 0.014
Chain 1: 1900 -8127.114 0.020 0.014
Chain 1: 2000 -8097.320 0.017 0.013
Chain 1: 2100 -8224.636 0.015 0.013
Chain 1: 2200 -8010.850 0.016 0.013
Chain 1: 2300 -8155.960 0.017 0.014
Chain 1: 2400 -8171.153 0.016 0.014
Chain 1: 2500 -8137.666 0.015 0.013
Chain 1: 2600 -8139.344 0.014 0.013
Chain 1: 2700 -8046.412 0.014 0.013
Chain 1: 2800 -8019.943 0.010 0.012 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003459 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.59 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8443038.110 1.000 1.000
Chain 1: 200 -1591747.980 2.652 4.304
Chain 1: 300 -890697.012 2.030 1.000
Chain 1: 400 -456728.675 1.760 1.000
Chain 1: 500 -356597.945 1.464 0.950
Chain 1: 600 -231597.588 1.310 0.950
Chain 1: 700 -118296.968 1.260 0.950
Chain 1: 800 -85670.759 1.150 0.950
Chain 1: 900 -66117.571 1.055 0.787
Chain 1: 1000 -51002.940 0.979 0.787
Chain 1: 1100 -38564.948 0.912 0.540 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37745.789 0.483 0.381
Chain 1: 1300 -25789.417 0.451 0.381
Chain 1: 1400 -25515.050 0.357 0.323
Chain 1: 1500 -22126.090 0.344 0.323
Chain 1: 1600 -21349.156 0.294 0.296
Chain 1: 1700 -20233.607 0.204 0.296
Chain 1: 1800 -20179.981 0.166 0.153
Chain 1: 1900 -20505.754 0.138 0.055
Chain 1: 2000 -19023.284 0.116 0.055
Chain 1: 2100 -19261.240 0.085 0.036
Chain 1: 2200 -19486.653 0.084 0.036
Chain 1: 2300 -19104.871 0.040 0.020
Chain 1: 2400 -18877.209 0.040 0.020
Chain 1: 2500 -18678.974 0.025 0.016
Chain 1: 2600 -18309.933 0.024 0.016
Chain 1: 2700 -18267.079 0.019 0.012
Chain 1: 2800 -17984.093 0.020 0.016
Chain 1: 2900 -18264.938 0.020 0.015
Chain 1: 3000 -18251.239 0.012 0.012
Chain 1: 3100 -18336.186 0.011 0.012
Chain 1: 3200 -18027.210 0.012 0.015
Chain 1: 3300 -18231.631 0.011 0.012
Chain 1: 3400 -17707.131 0.013 0.015
Chain 1: 3500 -18318.068 0.015 0.016
Chain 1: 3600 -17625.852 0.017 0.016
Chain 1: 3700 -18011.797 0.019 0.017
Chain 1: 3800 -16973.248 0.023 0.021
Chain 1: 3900 -16969.372 0.022 0.021
Chain 1: 4000 -17086.711 0.022 0.021
Chain 1: 4100 -17000.605 0.023 0.021
Chain 1: 4200 -16817.178 0.022 0.021
Chain 1: 4300 -16955.365 0.022 0.021
Chain 1: 4400 -16912.480 0.019 0.011
Chain 1: 4500 -16815.024 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001236 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.36 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48895.349 1.000 1.000
Chain 1: 200 -16777.472 1.457 1.914
Chain 1: 300 -20887.421 1.037 1.000
Chain 1: 400 -16468.934 0.845 1.000
Chain 1: 500 -14312.634 0.706 0.268
Chain 1: 600 -13426.076 0.599 0.268
Chain 1: 700 -16260.932 0.539 0.197
Chain 1: 800 -13445.577 0.497 0.209
Chain 1: 900 -11768.730 0.458 0.197
Chain 1: 1000 -12630.612 0.419 0.197
Chain 1: 1100 -13094.919 0.323 0.174
Chain 1: 1200 -11314.518 0.147 0.157
Chain 1: 1300 -13193.733 0.141 0.151
Chain 1: 1400 -13422.894 0.116 0.142
Chain 1: 1500 -10979.817 0.124 0.142
Chain 1: 1600 -10461.616 0.122 0.142
Chain 1: 1700 -18814.222 0.149 0.142
Chain 1: 1800 -22606.413 0.145 0.142
Chain 1: 1900 -10418.863 0.247 0.157
Chain 1: 2000 -17813.127 0.282 0.168
Chain 1: 2100 -21137.599 0.294 0.168
Chain 1: 2200 -9556.418 0.400 0.223
Chain 1: 2300 -9705.959 0.387 0.223
Chain 1: 2400 -12776.525 0.409 0.240
Chain 1: 2500 -9826.638 0.417 0.300
Chain 1: 2600 -17123.611 0.455 0.415
Chain 1: 2700 -9404.683 0.492 0.415
Chain 1: 2800 -10739.121 0.488 0.415
Chain 1: 2900 -9939.594 0.379 0.300
Chain 1: 3000 -16564.850 0.378 0.300
Chain 1: 3100 -9697.244 0.433 0.400
Chain 1: 3200 -10362.602 0.318 0.300
Chain 1: 3300 -9295.646 0.328 0.300
Chain 1: 3400 -9743.491 0.308 0.300
Chain 1: 3500 -11239.476 0.292 0.133
Chain 1: 3600 -10286.157 0.258 0.124
Chain 1: 3700 -14462.558 0.205 0.124
Chain 1: 3800 -8812.921 0.257 0.133
Chain 1: 3900 -8927.477 0.250 0.133
Chain 1: 4000 -10743.425 0.227 0.133
Chain 1: 4100 -9054.764 0.175 0.133
Chain 1: 4200 -9909.043 0.177 0.133
Chain 1: 4300 -14203.778 0.196 0.169
Chain 1: 4400 -9379.080 0.243 0.186
Chain 1: 4500 -9044.134 0.233 0.186
Chain 1: 4600 -8602.178 0.229 0.186
Chain 1: 4700 -10644.826 0.219 0.186
Chain 1: 4800 -9096.841 0.172 0.170
Chain 1: 4900 -12230.611 0.197 0.186
Chain 1: 5000 -14883.714 0.197 0.186
Chain 1: 5100 -8709.807 0.250 0.192
Chain 1: 5200 -9046.275 0.245 0.192
Chain 1: 5300 -11632.555 0.237 0.192
Chain 1: 5400 -8461.728 0.223 0.192
Chain 1: 5500 -13484.582 0.256 0.222
Chain 1: 5600 -9704.879 0.290 0.256
Chain 1: 5700 -9209.980 0.276 0.256
Chain 1: 5800 -8744.125 0.265 0.256
Chain 1: 5900 -12953.986 0.272 0.325
Chain 1: 6000 -9071.263 0.297 0.372
Chain 1: 6100 -9019.077 0.226 0.325
Chain 1: 6200 -8352.091 0.230 0.325
Chain 1: 6300 -9320.088 0.219 0.325
Chain 1: 6400 -8582.148 0.190 0.104
Chain 1: 6500 -9066.486 0.158 0.086
Chain 1: 6600 -11329.049 0.139 0.086
Chain 1: 6700 -10563.870 0.141 0.086
Chain 1: 6800 -13938.675 0.160 0.104
Chain 1: 6900 -9872.943 0.168 0.104
Chain 1: 7000 -8263.238 0.145 0.104
Chain 1: 7100 -8939.090 0.152 0.104
Chain 1: 7200 -11275.918 0.165 0.195
Chain 1: 7300 -8359.234 0.189 0.200
Chain 1: 7400 -13959.010 0.221 0.207
Chain 1: 7500 -9401.184 0.264 0.242
Chain 1: 7600 -9183.251 0.246 0.242
Chain 1: 7700 -10975.699 0.255 0.242
Chain 1: 7800 -9430.290 0.248 0.207
Chain 1: 7900 -8420.973 0.218 0.195
Chain 1: 8000 -9236.119 0.208 0.164
Chain 1: 8100 -8533.169 0.208 0.164
Chain 1: 8200 -8355.429 0.190 0.163
Chain 1: 8300 -8382.410 0.155 0.120
Chain 1: 8400 -8496.676 0.116 0.088
Chain 1: 8500 -8338.751 0.070 0.082
Chain 1: 8600 -8784.444 0.073 0.082
Chain 1: 8700 -9245.340 0.061 0.051
Chain 1: 8800 -8647.890 0.052 0.051
Chain 1: 8900 -9283.618 0.047 0.051
Chain 1: 9000 -11361.413 0.056 0.051
Chain 1: 9100 -8474.733 0.082 0.051
Chain 1: 9200 -9634.020 0.092 0.068
Chain 1: 9300 -9510.380 0.093 0.068
Chain 1: 9400 -8332.073 0.106 0.069
Chain 1: 9500 -11345.009 0.130 0.120
Chain 1: 9600 -8536.547 0.158 0.141
Chain 1: 9700 -8814.373 0.156 0.141
Chain 1: 9800 -8507.323 0.153 0.141
Chain 1: 9900 -9176.122 0.153 0.141
Chain 1: 10000 -8238.325 0.146 0.120
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001395 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.95 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57284.696 1.000 1.000
Chain 1: 200 -17642.010 1.624 2.247
Chain 1: 300 -8786.794 1.418 1.008
Chain 1: 400 -8261.488 1.080 1.008
Chain 1: 500 -8521.222 0.870 1.000
Chain 1: 600 -8261.433 0.730 1.000
Chain 1: 700 -7940.112 0.632 0.064
Chain 1: 800 -8390.565 0.559 0.064
Chain 1: 900 -7703.948 0.507 0.064
Chain 1: 1000 -7809.093 0.458 0.064
Chain 1: 1100 -7709.980 0.359 0.054
Chain 1: 1200 -7777.338 0.135 0.040
Chain 1: 1300 -7637.943 0.036 0.031
Chain 1: 1400 -7813.778 0.032 0.030
Chain 1: 1500 -7592.796 0.032 0.029
Chain 1: 1600 -7718.431 0.030 0.023
Chain 1: 1700 -7568.214 0.028 0.020
Chain 1: 1800 -7630.678 0.024 0.018
Chain 1: 1900 -7581.154 0.016 0.016
Chain 1: 2000 -7581.530 0.014 0.016
Chain 1: 2100 -7514.515 0.014 0.016
Chain 1: 2200 -7724.953 0.016 0.018
Chain 1: 2300 -7521.729 0.017 0.020
Chain 1: 2400 -7612.533 0.016 0.016
Chain 1: 2500 -7617.981 0.013 0.012
Chain 1: 2600 -7540.341 0.012 0.010
Chain 1: 2700 -7570.133 0.010 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002882 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 28.82 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86554.124 1.000 1.000
Chain 1: 200 -13628.314 3.176 5.351
Chain 1: 300 -9963.248 2.240 1.000
Chain 1: 400 -10958.983 1.702 1.000
Chain 1: 500 -8873.311 1.409 0.368
Chain 1: 600 -8425.314 1.183 0.368
Chain 1: 700 -8547.443 1.016 0.235
Chain 1: 800 -8696.722 0.891 0.235
Chain 1: 900 -8715.287 0.792 0.091
Chain 1: 1000 -8686.418 0.713 0.091
Chain 1: 1100 -8648.662 0.614 0.053 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8384.572 0.082 0.031
Chain 1: 1300 -8711.903 0.049 0.031
Chain 1: 1400 -8655.309 0.041 0.017
Chain 1: 1500 -8496.724 0.019 0.017
Chain 1: 1600 -8610.878 0.015 0.014
Chain 1: 1700 -8687.190 0.014 0.013
Chain 1: 1800 -8259.910 0.018 0.013
Chain 1: 1900 -8362.640 0.019 0.013
Chain 1: 2000 -8337.552 0.019 0.013
Chain 1: 2100 -8464.621 0.020 0.015
Chain 1: 2200 -8263.509 0.019 0.015
Chain 1: 2300 -8358.011 0.016 0.013
Chain 1: 2400 -8425.845 0.017 0.013
Chain 1: 2500 -8372.034 0.015 0.012
Chain 1: 2600 -8374.525 0.014 0.011
Chain 1: 2700 -8290.710 0.014 0.011
Chain 1: 2800 -8249.191 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003618 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 36.18 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8412129.449 1.000 1.000
Chain 1: 200 -1587420.747 2.650 4.299
Chain 1: 300 -891367.400 2.027 1.000
Chain 1: 400 -457815.730 1.757 1.000
Chain 1: 500 -357992.204 1.461 0.947
Chain 1: 600 -232865.771 1.307 0.947
Chain 1: 700 -119194.982 1.257 0.947
Chain 1: 800 -86464.089 1.147 0.947
Chain 1: 900 -66837.788 1.052 0.781
Chain 1: 1000 -51667.830 0.976 0.781
Chain 1: 1100 -39170.289 0.908 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38350.720 0.480 0.379
Chain 1: 1300 -26324.082 0.448 0.379
Chain 1: 1400 -26045.624 0.354 0.319
Chain 1: 1500 -22637.792 0.342 0.319
Chain 1: 1600 -21856.202 0.291 0.294
Chain 1: 1700 -20731.491 0.201 0.294
Chain 1: 1800 -20676.117 0.164 0.151
Chain 1: 1900 -21002.494 0.136 0.054
Chain 1: 2000 -19514.262 0.114 0.054
Chain 1: 2100 -19752.514 0.084 0.036
Chain 1: 2200 -19979.089 0.083 0.036
Chain 1: 2300 -19596.162 0.039 0.020
Chain 1: 2400 -19368.204 0.039 0.020
Chain 1: 2500 -19170.252 0.025 0.016
Chain 1: 2600 -18800.254 0.023 0.016
Chain 1: 2700 -18757.175 0.018 0.012
Chain 1: 2800 -18473.991 0.019 0.015
Chain 1: 2900 -18755.256 0.019 0.015
Chain 1: 3000 -18741.455 0.012 0.012
Chain 1: 3100 -18826.478 0.011 0.012
Chain 1: 3200 -18517.042 0.012 0.015
Chain 1: 3300 -18721.864 0.011 0.012
Chain 1: 3400 -18196.592 0.012 0.015
Chain 1: 3500 -18808.760 0.015 0.015
Chain 1: 3600 -18115.056 0.017 0.015
Chain 1: 3700 -18502.115 0.018 0.017
Chain 1: 3800 -17461.258 0.023 0.021
Chain 1: 3900 -17457.397 0.021 0.021
Chain 1: 4000 -17574.693 0.022 0.021
Chain 1: 4100 -17488.442 0.022 0.021
Chain 1: 4200 -17304.564 0.021 0.021
Chain 1: 4300 -17443.041 0.021 0.021
Chain 1: 4400 -17399.745 0.018 0.011
Chain 1: 4500 -17302.272 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001126 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 11.26 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13232.777 1.000 1.000
Chain 1: 200 -9684.130 0.683 1.000
Chain 1: 300 -8696.357 0.493 0.366
Chain 1: 400 -8122.988 0.388 0.366
Chain 1: 500 -8214.949 0.312 0.114
Chain 1: 600 -8037.261 0.264 0.114
Chain 1: 700 -8009.815 0.227 0.071
Chain 1: 800 -7955.184 0.199 0.071
Chain 1: 900 -8034.430 0.178 0.022
Chain 1: 1000 -7994.255 0.161 0.022
Chain 1: 1100 -8102.695 0.062 0.013
Chain 1: 1200 -7935.177 0.028 0.013
Chain 1: 1300 -7920.098 0.017 0.011
Chain 1: 1400 -7928.129 0.010 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001382 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.82 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58865.420 1.000 1.000
Chain 1: 200 -18218.296 1.616 2.231
Chain 1: 300 -8914.196 1.425 1.044
Chain 1: 400 -8040.729 1.096 1.044
Chain 1: 500 -8719.590 0.892 1.000
Chain 1: 600 -8270.909 0.753 1.000
Chain 1: 700 -7965.461 0.651 0.109
Chain 1: 800 -7978.756 0.569 0.109
Chain 1: 900 -7797.649 0.509 0.078
Chain 1: 1000 -7584.534 0.461 0.078
Chain 1: 1100 -7797.858 0.363 0.054
Chain 1: 1200 -7840.132 0.141 0.038
Chain 1: 1300 -7579.230 0.040 0.034
Chain 1: 1400 -7827.222 0.032 0.032
Chain 1: 1500 -7612.218 0.027 0.028
Chain 1: 1600 -7786.415 0.024 0.028
Chain 1: 1700 -7632.114 0.022 0.027
Chain 1: 1800 -7689.330 0.023 0.027
Chain 1: 1900 -7596.604 0.022 0.027
Chain 1: 2000 -7696.500 0.020 0.022
Chain 1: 2100 -7578.481 0.019 0.020
Chain 1: 2200 -7781.283 0.021 0.022
Chain 1: 2300 -7542.211 0.021 0.022
Chain 1: 2400 -7718.670 0.020 0.022
Chain 1: 2500 -7620.486 0.018 0.020
Chain 1: 2600 -7528.610 0.017 0.016
Chain 1: 2700 -7519.952 0.016 0.013
Chain 1: 2800 -7520.087 0.015 0.013
Chain 1: 2900 -7373.330 0.016 0.016
Chain 1: 3000 -7528.646 0.016 0.020
Chain 1: 3100 -7522.975 0.015 0.020
Chain 1: 3200 -7742.173 0.015 0.020
Chain 1: 3300 -7461.621 0.016 0.020
Chain 1: 3400 -7701.437 0.016 0.020
Chain 1: 3500 -7435.454 0.019 0.021
Chain 1: 3600 -7500.063 0.018 0.021
Chain 1: 3700 -7452.713 0.019 0.021
Chain 1: 3800 -7452.509 0.019 0.021
Chain 1: 3900 -7407.987 0.018 0.021
Chain 1: 4000 -7401.512 0.016 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003127 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.27 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86701.838 1.000 1.000
Chain 1: 200 -13944.808 3.109 5.217
Chain 1: 300 -10138.226 2.198 1.000
Chain 1: 400 -11938.996 1.686 1.000
Chain 1: 500 -8537.068 1.428 0.398
Chain 1: 600 -8402.845 1.193 0.398
Chain 1: 700 -8741.075 1.028 0.375
Chain 1: 800 -9273.760 0.907 0.375
Chain 1: 900 -8749.110 0.813 0.151
Chain 1: 1000 -8759.260 0.732 0.151
Chain 1: 1100 -8904.769 0.633 0.060 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8355.086 0.118 0.060
Chain 1: 1300 -8744.479 0.085 0.057
Chain 1: 1400 -8622.917 0.071 0.045
Chain 1: 1500 -8572.047 0.032 0.039
Chain 1: 1600 -8691.102 0.032 0.039
Chain 1: 1700 -8731.482 0.028 0.016
Chain 1: 1800 -8267.432 0.028 0.016
Chain 1: 1900 -8381.602 0.024 0.014
Chain 1: 2000 -8401.859 0.024 0.014
Chain 1: 2100 -8495.954 0.023 0.014
Chain 1: 2200 -8266.792 0.019 0.014
Chain 1: 2300 -8487.138 0.018 0.014
Chain 1: 2400 -8272.723 0.019 0.014
Chain 1: 2500 -8350.040 0.019 0.014
Chain 1: 2600 -8258.944 0.019 0.014
Chain 1: 2700 -8294.769 0.019 0.014
Chain 1: 2800 -8246.576 0.014 0.011
Chain 1: 2900 -8360.863 0.014 0.011
Chain 1: 3000 -8270.199 0.015 0.011
Chain 1: 3100 -8237.622 0.014 0.011
Chain 1: 3200 -8208.471 0.011 0.011
Chain 1: 3300 -8472.098 0.012 0.011
Chain 1: 3400 -8518.733 0.010 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003307 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.07 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8418859.176 1.000 1.000
Chain 1: 200 -1585567.737 2.655 4.310
Chain 1: 300 -890660.872 2.030 1.000
Chain 1: 400 -458467.327 1.758 1.000
Chain 1: 500 -358646.015 1.462 0.943
Chain 1: 600 -233594.480 1.308 0.943
Chain 1: 700 -119726.706 1.257 0.943
Chain 1: 800 -86943.722 1.147 0.943
Chain 1: 900 -67274.797 1.052 0.780
Chain 1: 1000 -52082.521 0.976 0.780
Chain 1: 1100 -39564.279 0.907 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38747.712 0.479 0.377
Chain 1: 1300 -26694.765 0.446 0.377
Chain 1: 1400 -26417.300 0.353 0.316
Chain 1: 1500 -23001.734 0.340 0.316
Chain 1: 1600 -22219.060 0.290 0.292
Chain 1: 1700 -21090.375 0.200 0.292
Chain 1: 1800 -21034.727 0.162 0.148
Chain 1: 1900 -21361.709 0.135 0.054
Chain 1: 2000 -19870.176 0.113 0.054
Chain 1: 2100 -20108.806 0.083 0.035
Chain 1: 2200 -20335.999 0.082 0.035
Chain 1: 2300 -19952.297 0.038 0.019
Chain 1: 2400 -19724.003 0.038 0.019
Chain 1: 2500 -19526.102 0.025 0.015
Chain 1: 2600 -19155.349 0.023 0.015
Chain 1: 2700 -19112.029 0.018 0.012
Chain 1: 2800 -18828.479 0.019 0.015
Chain 1: 2900 -19110.186 0.019 0.015
Chain 1: 3000 -19096.253 0.012 0.012
Chain 1: 3100 -19181.413 0.011 0.012
Chain 1: 3200 -18871.499 0.011 0.015
Chain 1: 3300 -19076.687 0.011 0.012
Chain 1: 3400 -18550.586 0.012 0.015
Chain 1: 3500 -19164.067 0.014 0.015
Chain 1: 3600 -18468.579 0.016 0.015
Chain 1: 3700 -18856.975 0.018 0.016
Chain 1: 3800 -17813.424 0.022 0.021
Chain 1: 3900 -17809.462 0.021 0.021
Chain 1: 4000 -17926.768 0.022 0.021
Chain 1: 4100 -17840.382 0.022 0.021
Chain 1: 4200 -17655.873 0.021 0.021
Chain 1: 4300 -17794.792 0.021 0.021
Chain 1: 4400 -17751.025 0.018 0.010
Chain 1: 4500 -17653.428 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001348 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.48 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49551.342 1.000 1.000
Chain 1: 200 -18387.940 1.347 1.695
Chain 1: 300 -20739.758 0.936 1.000
Chain 1: 400 -29913.482 0.779 1.000
Chain 1: 500 -26855.113 0.646 0.307
Chain 1: 600 -23401.438 0.563 0.307
Chain 1: 700 -14924.121 0.563 0.307
Chain 1: 800 -12072.782 0.523 0.307
Chain 1: 900 -12042.240 0.465 0.236
Chain 1: 1000 -14141.609 0.433 0.236
Chain 1: 1100 -11537.659 0.356 0.226
Chain 1: 1200 -15291.303 0.211 0.226
Chain 1: 1300 -13620.971 0.212 0.226
Chain 1: 1400 -12190.238 0.193 0.148
Chain 1: 1500 -11388.146 0.188 0.148
Chain 1: 1600 -10350.914 0.184 0.148
Chain 1: 1700 -9934.487 0.131 0.123
Chain 1: 1800 -27386.743 0.171 0.123
Chain 1: 1900 -11172.931 0.316 0.148
Chain 1: 2000 -10733.701 0.305 0.123
Chain 1: 2100 -9946.530 0.291 0.117
Chain 1: 2200 -11494.520 0.280 0.117
Chain 1: 2300 -10024.286 0.282 0.117
Chain 1: 2400 -23749.866 0.328 0.135
Chain 1: 2500 -11107.922 0.435 0.147
Chain 1: 2600 -10299.368 0.433 0.147
Chain 1: 2700 -10518.596 0.431 0.147
Chain 1: 2800 -10525.224 0.367 0.135
Chain 1: 2900 -9732.090 0.230 0.081
Chain 1: 3000 -10020.406 0.229 0.081
Chain 1: 3100 -19615.830 0.270 0.135
Chain 1: 3200 -9824.672 0.356 0.147
Chain 1: 3300 -11025.377 0.352 0.109
Chain 1: 3400 -10423.287 0.300 0.081
Chain 1: 3500 -9583.241 0.195 0.081
Chain 1: 3600 -11969.966 0.207 0.088
Chain 1: 3700 -10180.784 0.223 0.109
Chain 1: 3800 -19809.720 0.271 0.176
Chain 1: 3900 -9229.426 0.378 0.199
Chain 1: 4000 -9685.932 0.379 0.199
Chain 1: 4100 -10383.171 0.337 0.176
Chain 1: 4200 -10594.253 0.240 0.109
Chain 1: 4300 -10488.336 0.230 0.088
Chain 1: 4400 -9076.350 0.240 0.156
Chain 1: 4500 -9289.850 0.233 0.156
Chain 1: 4600 -9033.705 0.216 0.067
Chain 1: 4700 -9202.748 0.200 0.047
Chain 1: 4800 -9424.062 0.154 0.028
Chain 1: 4900 -15580.758 0.079 0.028
Chain 1: 5000 -10000.262 0.130 0.028
Chain 1: 5100 -8964.794 0.135 0.028
Chain 1: 5200 -10309.188 0.146 0.116
Chain 1: 5300 -12800.499 0.164 0.130
Chain 1: 5400 -9840.800 0.179 0.130
Chain 1: 5500 -15091.531 0.211 0.195
Chain 1: 5600 -14486.636 0.213 0.195
Chain 1: 5700 -8829.909 0.275 0.301
Chain 1: 5800 -15128.821 0.314 0.348
Chain 1: 5900 -11481.914 0.306 0.318
Chain 1: 6000 -11098.986 0.254 0.301
Chain 1: 6100 -9183.577 0.263 0.301
Chain 1: 6200 -10208.188 0.260 0.301
Chain 1: 6300 -8876.978 0.256 0.301
Chain 1: 6400 -9077.656 0.228 0.209
Chain 1: 6500 -9574.953 0.198 0.150
Chain 1: 6600 -9636.369 0.195 0.150
Chain 1: 6700 -13537.891 0.160 0.150
Chain 1: 6800 -9600.033 0.159 0.150
Chain 1: 6900 -9983.030 0.131 0.100
Chain 1: 7000 -16144.930 0.166 0.150
Chain 1: 7100 -8620.272 0.232 0.150
Chain 1: 7200 -8929.196 0.226 0.150
Chain 1: 7300 -8541.370 0.215 0.052
Chain 1: 7400 -8948.334 0.218 0.052
Chain 1: 7500 -10748.720 0.229 0.167
Chain 1: 7600 -10001.663 0.236 0.167
Chain 1: 7700 -9380.736 0.214 0.075
Chain 1: 7800 -12664.571 0.199 0.075
Chain 1: 7900 -8827.200 0.238 0.167
Chain 1: 8000 -8693.202 0.202 0.075
Chain 1: 8100 -8958.043 0.117 0.066
Chain 1: 8200 -10147.919 0.126 0.075
Chain 1: 8300 -11869.228 0.136 0.117
Chain 1: 8400 -8809.033 0.166 0.145
Chain 1: 8500 -9713.509 0.158 0.117
Chain 1: 8600 -8754.655 0.162 0.117
Chain 1: 8700 -9211.157 0.160 0.117
Chain 1: 8800 -14561.000 0.171 0.117
Chain 1: 8900 -9593.035 0.179 0.117
Chain 1: 9000 -8486.192 0.191 0.130
Chain 1: 9100 -8415.872 0.189 0.130
Chain 1: 9200 -9199.648 0.185 0.130
Chain 1: 9300 -8939.742 0.174 0.110
Chain 1: 9400 -9592.409 0.146 0.093
Chain 1: 9500 -10313.830 0.144 0.085
Chain 1: 9600 -10267.581 0.133 0.070
Chain 1: 9700 -8536.052 0.148 0.085
Chain 1: 9800 -11928.260 0.140 0.085
Chain 1: 9900 -8560.937 0.128 0.085
Chain 1: 10000 -9342.425 0.123 0.084
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001438 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.38 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58681.284 1.000 1.000
Chain 1: 200 -18478.444 1.588 2.176
Chain 1: 300 -9053.851 1.406 1.041
Chain 1: 400 -8107.930 1.083 1.041
Chain 1: 500 -8933.110 0.885 1.000
Chain 1: 600 -8478.348 0.747 1.000
Chain 1: 700 -8526.707 0.641 0.117
Chain 1: 800 -8637.285 0.562 0.117
Chain 1: 900 -8188.062 0.506 0.092
Chain 1: 1000 -7738.747 0.461 0.092
Chain 1: 1100 -7828.563 0.362 0.058
Chain 1: 1200 -8007.007 0.147 0.055
Chain 1: 1300 -7911.061 0.044 0.054
Chain 1: 1400 -8035.975 0.034 0.022
Chain 1: 1500 -7520.281 0.032 0.022
Chain 1: 1600 -7741.768 0.029 0.022
Chain 1: 1700 -7455.654 0.032 0.029
Chain 1: 1800 -7584.669 0.033 0.029
Chain 1: 1900 -7603.732 0.027 0.022
Chain 1: 2000 -7732.358 0.023 0.017
Chain 1: 2100 -7582.098 0.024 0.020
Chain 1: 2200 -7657.333 0.023 0.017
Chain 1: 2300 -7480.100 0.024 0.020
Chain 1: 2400 -7626.082 0.024 0.020
Chain 1: 2500 -7505.344 0.019 0.019
Chain 1: 2600 -7540.819 0.017 0.017
Chain 1: 2700 -7512.255 0.013 0.017
Chain 1: 2800 -7550.781 0.012 0.016
Chain 1: 2900 -7376.178 0.014 0.017
Chain 1: 3000 -7531.896 0.015 0.019
Chain 1: 3100 -7521.820 0.013 0.016
Chain 1: 3200 -7731.810 0.015 0.019
Chain 1: 3300 -7394.647 0.017 0.019
Chain 1: 3400 -7656.277 0.018 0.021
Chain 1: 3500 -7467.482 0.019 0.024
Chain 1: 3600 -7463.555 0.019 0.024
Chain 1: 3700 -7439.601 0.019 0.024
Chain 1: 3800 -7412.021 0.019 0.024
Chain 1: 3900 -7429.067 0.016 0.021
Chain 1: 4000 -7386.631 0.015 0.006 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003249 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.49 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86973.307 1.000 1.000
Chain 1: 200 -14281.379 3.045 5.090
Chain 1: 300 -10441.563 2.153 1.000
Chain 1: 400 -12560.161 1.657 1.000
Chain 1: 500 -8836.584 1.410 0.421
Chain 1: 600 -8658.903 1.178 0.421
Chain 1: 700 -8753.510 1.011 0.368
Chain 1: 800 -9393.365 0.893 0.368
Chain 1: 900 -9014.842 0.799 0.169
Chain 1: 1000 -8901.332 0.720 0.169
Chain 1: 1100 -9183.663 0.623 0.068 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8650.566 0.120 0.062
Chain 1: 1300 -8965.732 0.087 0.042
Chain 1: 1400 -8891.156 0.071 0.035
Chain 1: 1500 -8868.584 0.029 0.031
Chain 1: 1600 -8925.885 0.028 0.031
Chain 1: 1700 -8988.126 0.027 0.031
Chain 1: 1800 -8521.268 0.026 0.031
Chain 1: 1900 -8631.021 0.023 0.013
Chain 1: 2000 -8648.259 0.022 0.013
Chain 1: 2100 -8753.118 0.020 0.012
Chain 1: 2200 -8507.439 0.017 0.012
Chain 1: 2300 -8614.158 0.015 0.012
Chain 1: 2400 -8683.145 0.015 0.012
Chain 1: 2500 -8622.836 0.015 0.012
Chain 1: 2600 -8662.094 0.015 0.012
Chain 1: 2700 -8551.208 0.016 0.012
Chain 1: 2800 -8497.661 0.011 0.012
Chain 1: 2900 -8604.470 0.011 0.012
Chain 1: 3000 -8518.819 0.011 0.012
Chain 1: 3100 -8484.124 0.011 0.010
Chain 1: 3200 -8450.340 0.008 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003257 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.57 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8359977.669 1.000 1.000
Chain 1: 200 -1576447.418 2.652 4.303
Chain 1: 300 -890235.841 2.025 1.000
Chain 1: 400 -458000.315 1.754 1.000
Chain 1: 500 -359374.288 1.458 0.944
Chain 1: 600 -234479.262 1.304 0.944
Chain 1: 700 -120472.598 1.253 0.944
Chain 1: 800 -87614.118 1.143 0.944
Chain 1: 900 -67886.975 1.049 0.771
Chain 1: 1000 -52630.905 0.973 0.771
Chain 1: 1100 -40046.321 0.904 0.533 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39223.657 0.476 0.375
Chain 1: 1300 -27094.280 0.444 0.375
Chain 1: 1400 -26809.669 0.350 0.314
Chain 1: 1500 -23374.136 0.337 0.314
Chain 1: 1600 -22585.769 0.288 0.291
Chain 1: 1700 -21447.994 0.198 0.290
Chain 1: 1800 -21390.164 0.161 0.147
Chain 1: 1900 -21717.448 0.134 0.053
Chain 1: 2000 -20220.560 0.112 0.053
Chain 1: 2100 -20459.381 0.082 0.035
Chain 1: 2200 -20687.626 0.081 0.035
Chain 1: 2300 -20302.935 0.038 0.019
Chain 1: 2400 -20074.522 0.038 0.019
Chain 1: 2500 -19876.883 0.024 0.015
Chain 1: 2600 -19505.629 0.023 0.015
Chain 1: 2700 -19462.127 0.018 0.012
Chain 1: 2800 -19178.712 0.019 0.015
Chain 1: 2900 -19460.546 0.019 0.014
Chain 1: 3000 -19446.522 0.011 0.012
Chain 1: 3100 -19531.747 0.011 0.011
Chain 1: 3200 -19221.596 0.011 0.014
Chain 1: 3300 -19426.963 0.010 0.011
Chain 1: 3400 -18900.572 0.012 0.014
Chain 1: 3500 -19514.597 0.014 0.015
Chain 1: 3600 -18818.444 0.016 0.015
Chain 1: 3700 -19207.470 0.018 0.016
Chain 1: 3800 -18162.901 0.022 0.020
Chain 1: 3900 -18158.983 0.021 0.020
Chain 1: 4000 -18276.226 0.021 0.020
Chain 1: 4100 -18189.845 0.021 0.020
Chain 1: 4200 -18005.117 0.021 0.020
Chain 1: 4300 -18144.173 0.020 0.020
Chain 1: 4400 -18100.243 0.018 0.010
Chain 1: 4500 -18002.652 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001299 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.99 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48623.484 1.000 1.000
Chain 1: 200 -19629.450 1.239 1.477
Chain 1: 300 -19150.391 0.834 1.000
Chain 1: 400 -15989.447 0.675 1.000
Chain 1: 500 -12611.078 0.594 0.268
Chain 1: 600 -17162.101 0.539 0.268
Chain 1: 700 -14517.697 0.488 0.265
Chain 1: 800 -21264.603 0.467 0.268
Chain 1: 900 -11013.065 0.518 0.268
Chain 1: 1000 -24743.282 0.522 0.317
Chain 1: 1100 -22656.814 0.431 0.268
Chain 1: 1200 -10767.440 0.394 0.268
Chain 1: 1300 -12833.740 0.407 0.268
Chain 1: 1400 -12391.901 0.391 0.268
Chain 1: 1500 -11874.670 0.369 0.265
Chain 1: 1600 -12159.667 0.345 0.182
Chain 1: 1700 -10913.730 0.338 0.161
Chain 1: 1800 -10010.886 0.315 0.114
Chain 1: 1900 -15924.279 0.259 0.114
Chain 1: 2000 -11112.553 0.247 0.114
Chain 1: 2100 -10561.665 0.243 0.114
Chain 1: 2200 -9484.056 0.144 0.114
Chain 1: 2300 -9007.419 0.133 0.090
Chain 1: 2400 -10490.502 0.144 0.114
Chain 1: 2500 -9054.491 0.155 0.114
Chain 1: 2600 -9901.327 0.161 0.114
Chain 1: 2700 -8990.972 0.160 0.114
Chain 1: 2800 -9245.313 0.154 0.114
Chain 1: 2900 -9410.347 0.118 0.101
Chain 1: 3000 -10001.052 0.081 0.086
Chain 1: 3100 -9022.367 0.087 0.101
Chain 1: 3200 -8911.312 0.076 0.086
Chain 1: 3300 -12645.480 0.101 0.101
Chain 1: 3400 -8652.847 0.133 0.101
Chain 1: 3500 -9094.239 0.122 0.086
Chain 1: 3600 -9619.882 0.119 0.059
Chain 1: 3700 -8718.920 0.119 0.059
Chain 1: 3800 -13720.404 0.153 0.103
Chain 1: 3900 -9086.158 0.202 0.108
Chain 1: 4000 -10656.038 0.211 0.147
Chain 1: 4100 -13504.176 0.221 0.211
Chain 1: 4200 -12812.590 0.225 0.211
Chain 1: 4300 -9815.354 0.226 0.211
Chain 1: 4400 -11373.754 0.194 0.147
Chain 1: 4500 -9350.715 0.210 0.211
Chain 1: 4600 -11917.251 0.226 0.215
Chain 1: 4700 -8606.667 0.255 0.216
Chain 1: 4800 -8407.784 0.220 0.215
Chain 1: 4900 -10912.197 0.192 0.215
Chain 1: 5000 -15509.530 0.207 0.216
Chain 1: 5100 -11999.110 0.215 0.230
Chain 1: 5200 -9956.584 0.231 0.230
Chain 1: 5300 -9494.954 0.205 0.216
Chain 1: 5400 -10084.247 0.197 0.216
Chain 1: 5500 -9013.731 0.187 0.215
Chain 1: 5600 -10651.246 0.181 0.205
Chain 1: 5700 -12743.393 0.159 0.164
Chain 1: 5800 -9552.148 0.190 0.205
Chain 1: 5900 -13090.173 0.194 0.205
Chain 1: 6000 -11865.934 0.175 0.164
Chain 1: 6100 -8563.493 0.184 0.164
Chain 1: 6200 -8272.860 0.167 0.154
Chain 1: 6300 -13608.209 0.202 0.164
Chain 1: 6400 -9750.359 0.235 0.270
Chain 1: 6500 -8376.617 0.240 0.270
Chain 1: 6600 -11652.974 0.253 0.281
Chain 1: 6700 -8329.868 0.276 0.334
Chain 1: 6800 -11211.036 0.268 0.281
Chain 1: 6900 -10491.364 0.248 0.281
Chain 1: 7000 -8502.269 0.261 0.281
Chain 1: 7100 -10630.375 0.243 0.257
Chain 1: 7200 -11220.545 0.244 0.257
Chain 1: 7300 -8548.922 0.236 0.257
Chain 1: 7400 -8312.577 0.200 0.234
Chain 1: 7500 -8169.089 0.185 0.234
Chain 1: 7600 -10368.120 0.178 0.212
Chain 1: 7700 -8219.964 0.164 0.212
Chain 1: 7800 -11655.628 0.168 0.212
Chain 1: 7900 -9185.985 0.188 0.234
Chain 1: 8000 -8178.206 0.177 0.212
Chain 1: 8100 -8094.432 0.158 0.212
Chain 1: 8200 -9897.910 0.171 0.212
Chain 1: 8300 -11182.804 0.151 0.182
Chain 1: 8400 -8120.058 0.186 0.212
Chain 1: 8500 -8232.818 0.186 0.212
Chain 1: 8600 -8206.561 0.165 0.182
Chain 1: 8700 -8562.647 0.143 0.123
Chain 1: 8800 -8065.271 0.120 0.115
Chain 1: 8900 -8951.011 0.103 0.099
Chain 1: 9000 -8325.847 0.098 0.075
Chain 1: 9100 -9833.925 0.112 0.099
Chain 1: 9200 -9590.404 0.097 0.075
Chain 1: 9300 -9609.844 0.085 0.062
Chain 1: 9400 -8197.044 0.065 0.062
Chain 1: 9500 -11689.356 0.093 0.075
Chain 1: 9600 -8217.492 0.135 0.099
Chain 1: 9700 -8117.243 0.132 0.099
Chain 1: 9800 -10332.352 0.148 0.153
Chain 1: 9900 -9708.070 0.144 0.153
Chain 1: 10000 -8693.154 0.148 0.153
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001375 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.75 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58040.846 1.000 1.000
Chain 1: 200 -17459.450 1.662 2.324
Chain 1: 300 -8580.856 1.453 1.035
Chain 1: 400 -8195.423 1.102 1.035
Chain 1: 500 -8229.975 0.882 1.000
Chain 1: 600 -8539.021 0.741 1.000
Chain 1: 700 -7901.911 0.647 0.081
Chain 1: 800 -7839.899 0.567 0.081
Chain 1: 900 -7925.509 0.505 0.047
Chain 1: 1000 -7813.419 0.456 0.047
Chain 1: 1100 -7699.200 0.357 0.036
Chain 1: 1200 -7713.335 0.125 0.015
Chain 1: 1300 -7760.858 0.022 0.014
Chain 1: 1400 -7664.854 0.019 0.013
Chain 1: 1500 -7597.087 0.019 0.013
Chain 1: 1600 -7573.167 0.016 0.011
Chain 1: 1700 -7533.043 0.009 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003547 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.47 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86015.660 1.000 1.000
Chain 1: 200 -13261.915 3.243 5.486
Chain 1: 300 -9701.971 2.284 1.000
Chain 1: 400 -10713.084 1.737 1.000
Chain 1: 500 -8626.344 1.438 0.367
Chain 1: 600 -8260.369 1.206 0.367
Chain 1: 700 -8457.071 1.037 0.242
Chain 1: 800 -8968.352 0.914 0.242
Chain 1: 900 -8521.441 0.818 0.094
Chain 1: 1000 -8304.892 0.739 0.094
Chain 1: 1100 -8582.090 0.642 0.057 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8215.909 0.098 0.052
Chain 1: 1300 -8273.575 0.062 0.045
Chain 1: 1400 -8267.164 0.053 0.044
Chain 1: 1500 -8302.276 0.029 0.032
Chain 1: 1600 -8308.611 0.025 0.026
Chain 1: 1700 -8239.276 0.023 0.026
Chain 1: 1800 -8119.957 0.019 0.015
Chain 1: 1900 -8238.473 0.015 0.014
Chain 1: 2000 -8198.150 0.013 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003188 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.88 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8398274.109 1.000 1.000
Chain 1: 200 -1582335.625 2.654 4.308
Chain 1: 300 -891147.473 2.028 1.000
Chain 1: 400 -458559.538 1.757 1.000
Chain 1: 500 -358892.145 1.461 0.943
Chain 1: 600 -233547.765 1.307 0.943
Chain 1: 700 -119323.081 1.257 0.943
Chain 1: 800 -86458.570 1.147 0.943
Chain 1: 900 -66714.710 1.053 0.776
Chain 1: 1000 -51447.051 0.977 0.776
Chain 1: 1100 -38876.547 0.909 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38039.280 0.481 0.380
Chain 1: 1300 -25955.102 0.450 0.380
Chain 1: 1400 -25667.436 0.357 0.323
Chain 1: 1500 -22245.804 0.344 0.323
Chain 1: 1600 -21458.972 0.294 0.297
Chain 1: 1700 -20328.244 0.204 0.296
Chain 1: 1800 -20271.055 0.166 0.154
Chain 1: 1900 -20596.673 0.138 0.056
Chain 1: 2000 -19106.638 0.116 0.056
Chain 1: 2100 -19344.872 0.085 0.037
Chain 1: 2200 -19571.546 0.084 0.037
Chain 1: 2300 -19188.683 0.040 0.020
Chain 1: 2400 -18960.916 0.040 0.020
Chain 1: 2500 -18763.254 0.026 0.016
Chain 1: 2600 -18393.656 0.024 0.016
Chain 1: 2700 -18350.616 0.019 0.012
Chain 1: 2800 -18067.898 0.020 0.016
Chain 1: 2900 -18348.944 0.020 0.015
Chain 1: 3000 -18335.028 0.012 0.012
Chain 1: 3100 -18420.022 0.011 0.012
Chain 1: 3200 -18110.949 0.012 0.015
Chain 1: 3300 -18315.441 0.011 0.012
Chain 1: 3400 -17791.005 0.013 0.015
Chain 1: 3500 -18402.076 0.015 0.016
Chain 1: 3600 -17709.756 0.017 0.016
Chain 1: 3700 -18095.899 0.019 0.017
Chain 1: 3800 -17057.287 0.023 0.021
Chain 1: 3900 -17053.511 0.022 0.021
Chain 1: 4000 -17170.753 0.022 0.021
Chain 1: 4100 -17084.691 0.022 0.021
Chain 1: 4200 -16901.227 0.022 0.021
Chain 1: 4300 -17039.357 0.022 0.021
Chain 1: 4400 -16996.477 0.019 0.011
Chain 1: 4500 -16899.098 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001382 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.82 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -14191.915 1.000 1.000
Chain 1: 200 -10649.459 0.666 1.000
Chain 1: 300 -9377.400 0.489 0.333
Chain 1: 400 -8689.603 0.387 0.333
Chain 1: 500 -8617.187 0.311 0.136
Chain 1: 600 -8721.148 0.261 0.136
Chain 1: 700 -8733.452 0.224 0.079
Chain 1: 800 -8631.473 0.198 0.079
Chain 1: 900 -8508.966 0.177 0.014
Chain 1: 1000 -8610.580 0.161 0.014
Chain 1: 1100 -8636.707 0.061 0.012
Chain 1: 1200 -8531.425 0.029 0.012
Chain 1: 1300 -8492.741 0.016 0.012
Chain 1: 1400 -8511.725 0.008 0.012 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001473 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.73 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -64761.399 1.000 1.000
Chain 1: 200 -19500.068 1.661 2.321
Chain 1: 300 -9447.127 1.462 1.064
Chain 1: 400 -8840.994 1.113 1.064
Chain 1: 500 -9081.614 0.896 1.000
Chain 1: 600 -9752.238 0.758 1.000
Chain 1: 700 -8468.834 0.672 0.152
Chain 1: 800 -8528.743 0.588 0.152
Chain 1: 900 -8548.262 0.523 0.069
Chain 1: 1000 -7655.032 0.483 0.117
Chain 1: 1100 -7910.267 0.386 0.069
Chain 1: 1200 -7915.768 0.154 0.069
Chain 1: 1300 -7598.998 0.052 0.042
Chain 1: 1400 -8057.693 0.050 0.042
Chain 1: 1500 -7632.461 0.053 0.056
Chain 1: 1600 -7836.167 0.049 0.042
Chain 1: 1700 -7557.510 0.038 0.037
Chain 1: 1800 -7594.033 0.037 0.037
Chain 1: 1900 -7698.079 0.039 0.037
Chain 1: 2000 -7634.615 0.028 0.032
Chain 1: 2100 -7575.696 0.025 0.026
Chain 1: 2200 -7982.057 0.030 0.037
Chain 1: 2300 -7600.300 0.031 0.037
Chain 1: 2400 -7564.639 0.026 0.026
Chain 1: 2500 -7559.056 0.020 0.014
Chain 1: 2600 -7624.128 0.019 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002954 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.54 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87414.574 1.000 1.000
Chain 1: 200 -14693.902 2.975 4.949
Chain 1: 300 -10826.847 2.102 1.000
Chain 1: 400 -13077.261 1.620 1.000
Chain 1: 500 -9155.268 1.381 0.428
Chain 1: 600 -9225.136 1.152 0.428
Chain 1: 700 -9336.237 0.989 0.357
Chain 1: 800 -9294.295 0.866 0.357
Chain 1: 900 -9535.326 0.773 0.172
Chain 1: 1000 -9595.809 0.696 0.172
Chain 1: 1100 -9473.182 0.598 0.025 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8939.336 0.109 0.025
Chain 1: 1300 -9338.908 0.077 0.025
Chain 1: 1400 -9282.640 0.061 0.013
Chain 1: 1500 -9237.598 0.018 0.012
Chain 1: 1600 -9249.844 0.018 0.012
Chain 1: 1700 -9366.001 0.018 0.012
Chain 1: 1800 -8879.475 0.023 0.013
Chain 1: 1900 -9002.145 0.021 0.013
Chain 1: 2000 -9011.957 0.021 0.013
Chain 1: 2100 -9135.326 0.021 0.014
Chain 1: 2200 -8873.363 0.018 0.014
Chain 1: 2300 -8965.535 0.015 0.012
Chain 1: 2400 -9053.693 0.015 0.012
Chain 1: 2500 -8973.411 0.016 0.012
Chain 1: 2600 -8997.841 0.016 0.012
Chain 1: 2700 -8911.759 0.015 0.010
Chain 1: 2800 -8872.922 0.010 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003273 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.73 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8426767.277 1.000 1.000
Chain 1: 200 -1585036.008 2.658 4.316
Chain 1: 300 -890984.758 2.032 1.000
Chain 1: 400 -458287.827 1.760 1.000
Chain 1: 500 -358446.856 1.464 0.944
Chain 1: 600 -233648.670 1.309 0.944
Chain 1: 700 -120183.983 1.257 0.944
Chain 1: 800 -87473.818 1.146 0.944
Chain 1: 900 -67883.380 1.051 0.779
Chain 1: 1000 -52743.946 0.975 0.779
Chain 1: 1100 -40266.911 0.906 0.534 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39460.798 0.476 0.374
Chain 1: 1300 -27441.751 0.442 0.374
Chain 1: 1400 -27169.570 0.348 0.310
Chain 1: 1500 -23761.928 0.335 0.310
Chain 1: 1600 -22982.257 0.285 0.289
Chain 1: 1700 -21857.358 0.196 0.287
Chain 1: 1800 -21802.835 0.159 0.143
Chain 1: 1900 -22130.172 0.131 0.051
Chain 1: 2000 -20639.909 0.110 0.051
Chain 1: 2100 -20878.493 0.080 0.034
Chain 1: 2200 -21105.583 0.079 0.034
Chain 1: 2300 -20721.953 0.037 0.019
Chain 1: 2400 -20493.619 0.037 0.019
Chain 1: 2500 -20295.480 0.024 0.015
Chain 1: 2600 -19924.574 0.022 0.015
Chain 1: 2700 -19881.265 0.017 0.011
Chain 1: 2800 -19597.477 0.018 0.014
Chain 1: 2900 -19879.288 0.018 0.014
Chain 1: 3000 -19865.412 0.011 0.011
Chain 1: 3100 -19950.561 0.010 0.011
Chain 1: 3200 -19640.470 0.011 0.014
Chain 1: 3300 -19845.835 0.010 0.011
Chain 1: 3400 -19319.287 0.012 0.014
Chain 1: 3500 -19933.282 0.014 0.014
Chain 1: 3600 -19237.194 0.016 0.014
Chain 1: 3700 -19626.000 0.017 0.016
Chain 1: 3800 -18581.370 0.022 0.020
Chain 1: 3900 -18577.369 0.020 0.020
Chain 1: 4000 -18694.713 0.021 0.020
Chain 1: 4100 -18608.221 0.021 0.020
Chain 1: 4200 -18423.531 0.020 0.020
Chain 1: 4300 -18562.613 0.020 0.020
Chain 1: 4400 -18518.662 0.017 0.010
Chain 1: 4500 -18421.015 0.015 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001365 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.65 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12397.789 1.000 1.000
Chain 1: 200 -9275.742 0.668 1.000
Chain 1: 300 -8185.208 0.490 0.337
Chain 1: 400 -8269.517 0.370 0.337
Chain 1: 500 -8136.701 0.299 0.133
Chain 1: 600 -8053.100 0.251 0.133
Chain 1: 700 -7976.394 0.217 0.016
Chain 1: 800 -8026.684 0.190 0.016
Chain 1: 900 -8100.890 0.170 0.010
Chain 1: 1000 -8043.733 0.154 0.010
Chain 1: 1100 -8120.728 0.055 0.010
Chain 1: 1200 -8001.433 0.023 0.010
Chain 1: 1300 -7944.641 0.010 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001674 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 16.74 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56755.408 1.000 1.000
Chain 1: 200 -17400.798 1.631 2.262
Chain 1: 300 -8727.240 1.419 1.000
Chain 1: 400 -8382.838 1.074 1.000
Chain 1: 500 -8644.321 0.865 0.994
Chain 1: 600 -9250.784 0.732 0.994
Chain 1: 700 -7907.746 0.652 0.170
Chain 1: 800 -8203.304 0.575 0.170
Chain 1: 900 -7922.730 0.515 0.066
Chain 1: 1000 -7764.404 0.465 0.066
Chain 1: 1100 -7780.841 0.366 0.041
Chain 1: 1200 -7708.348 0.140 0.036
Chain 1: 1300 -7672.855 0.041 0.035
Chain 1: 1400 -7737.162 0.038 0.030
Chain 1: 1500 -7636.036 0.036 0.020
Chain 1: 1600 -7692.660 0.031 0.013
Chain 1: 1700 -7541.789 0.016 0.013
Chain 1: 1800 -7560.517 0.012 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003368 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.68 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86989.818 1.000 1.000
Chain 1: 200 -13467.786 3.230 5.459
Chain 1: 300 -9880.051 2.274 1.000
Chain 1: 400 -10786.082 1.727 1.000
Chain 1: 500 -8831.186 1.426 0.363
Chain 1: 600 -8619.507 1.192 0.363
Chain 1: 700 -8446.475 1.025 0.221
Chain 1: 800 -9262.700 0.908 0.221
Chain 1: 900 -8699.671 0.814 0.088
Chain 1: 1000 -8486.163 0.735 0.088
Chain 1: 1100 -8773.430 0.638 0.084 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8303.557 0.098 0.065
Chain 1: 1300 -8511.708 0.064 0.057
Chain 1: 1400 -8626.591 0.057 0.033
Chain 1: 1500 -8490.942 0.037 0.025
Chain 1: 1600 -8600.355 0.035 0.025
Chain 1: 1700 -8682.938 0.034 0.025
Chain 1: 1800 -8286.336 0.030 0.025
Chain 1: 1900 -8388.395 0.025 0.024
Chain 1: 2000 -8358.845 0.023 0.016
Chain 1: 2100 -8481.329 0.021 0.014
Chain 1: 2200 -8262.865 0.018 0.014
Chain 1: 2300 -8416.945 0.017 0.014
Chain 1: 2400 -8430.958 0.016 0.014
Chain 1: 2500 -8400.245 0.015 0.013
Chain 1: 2600 -8402.826 0.014 0.012
Chain 1: 2700 -8309.093 0.014 0.012
Chain 1: 2800 -8280.347 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003018 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.18 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8410951.295 1.000 1.000
Chain 1: 200 -1586437.769 2.651 4.302
Chain 1: 300 -890995.128 2.027 1.000
Chain 1: 400 -457381.087 1.758 1.000
Chain 1: 500 -357492.405 1.462 0.948
Chain 1: 600 -232468.325 1.308 0.948
Chain 1: 700 -118946.373 1.257 0.948
Chain 1: 800 -86192.233 1.148 0.948
Chain 1: 900 -66586.428 1.053 0.781
Chain 1: 1000 -51418.123 0.977 0.781
Chain 1: 1100 -38928.181 0.909 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38108.784 0.481 0.380
Chain 1: 1300 -26107.446 0.449 0.380
Chain 1: 1400 -25828.952 0.355 0.321
Chain 1: 1500 -22426.581 0.343 0.321
Chain 1: 1600 -21645.636 0.292 0.295
Chain 1: 1700 -20524.972 0.202 0.294
Chain 1: 1800 -20470.290 0.165 0.152
Chain 1: 1900 -20796.062 0.137 0.055
Chain 1: 2000 -19310.808 0.115 0.055
Chain 1: 2100 -19549.060 0.084 0.036
Chain 1: 2200 -19774.640 0.083 0.036
Chain 1: 2300 -19392.722 0.039 0.020
Chain 1: 2400 -19165.008 0.039 0.020
Chain 1: 2500 -18966.727 0.025 0.016
Chain 1: 2600 -18597.527 0.024 0.016
Chain 1: 2700 -18554.781 0.018 0.012
Chain 1: 2800 -18271.610 0.020 0.015
Chain 1: 2900 -18552.690 0.020 0.015
Chain 1: 3000 -18538.981 0.012 0.012
Chain 1: 3100 -18623.848 0.011 0.012
Chain 1: 3200 -18314.857 0.012 0.015
Chain 1: 3300 -18519.378 0.011 0.012
Chain 1: 3400 -17994.716 0.013 0.015
Chain 1: 3500 -18605.822 0.015 0.015
Chain 1: 3600 -17913.592 0.017 0.015
Chain 1: 3700 -18299.486 0.019 0.017
Chain 1: 3800 -17260.755 0.023 0.021
Chain 1: 3900 -17256.925 0.022 0.021
Chain 1: 4000 -17374.262 0.022 0.021
Chain 1: 4100 -17287.998 0.022 0.021
Chain 1: 4200 -17104.687 0.022 0.021
Chain 1: 4300 -17242.825 0.021 0.021
Chain 1: 4400 -17199.935 0.019 0.011
Chain 1: 4500 -17102.500 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001456 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.56 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49711.195 1.000 1.000
Chain 1: 200 -25659.642 0.969 1.000
Chain 1: 300 -20606.988 0.728 0.937
Chain 1: 400 -19920.910 0.554 0.937
Chain 1: 500 -17178.558 0.475 0.245
Chain 1: 600 -18267.541 0.406 0.245
Chain 1: 700 -15968.750 0.369 0.160
Chain 1: 800 -13076.252 0.350 0.221
Chain 1: 900 -16613.663 0.335 0.213
Chain 1: 1000 -11546.310 0.345 0.221
Chain 1: 1100 -10831.559 0.252 0.213
Chain 1: 1200 -12933.444 0.174 0.163
Chain 1: 1300 -17219.130 0.175 0.163
Chain 1: 1400 -11823.734 0.217 0.213
Chain 1: 1500 -10589.537 0.213 0.213
Chain 1: 1600 -13944.243 0.231 0.221
Chain 1: 1700 -10908.986 0.244 0.241
Chain 1: 1800 -10410.580 0.227 0.241
Chain 1: 1900 -10592.311 0.207 0.241
Chain 1: 2000 -10546.604 0.164 0.163
Chain 1: 2100 -10129.592 0.161 0.163
Chain 1: 2200 -10261.189 0.146 0.117
Chain 1: 2300 -10661.897 0.125 0.048
Chain 1: 2400 -9771.118 0.089 0.048
Chain 1: 2500 -10996.486 0.088 0.048
Chain 1: 2600 -11466.786 0.068 0.041
Chain 1: 2700 -10443.101 0.050 0.041
Chain 1: 2800 -19550.622 0.092 0.041
Chain 1: 2900 -10052.179 0.185 0.091
Chain 1: 3000 -18045.132 0.229 0.098
Chain 1: 3100 -17757.741 0.226 0.098
Chain 1: 3200 -10669.364 0.291 0.111
Chain 1: 3300 -10360.373 0.291 0.111
Chain 1: 3400 -16440.057 0.318 0.370
Chain 1: 3500 -10746.500 0.360 0.443
Chain 1: 3600 -10900.876 0.358 0.443
Chain 1: 3700 -10692.682 0.350 0.443
Chain 1: 3800 -12638.877 0.319 0.370
Chain 1: 3900 -10796.690 0.241 0.171
Chain 1: 4000 -9553.671 0.210 0.154
Chain 1: 4100 -10956.767 0.221 0.154
Chain 1: 4200 -16736.452 0.189 0.154
Chain 1: 4300 -9890.572 0.255 0.171
Chain 1: 4400 -9946.811 0.219 0.154
Chain 1: 4500 -10971.545 0.175 0.130
Chain 1: 4600 -10767.879 0.176 0.130
Chain 1: 4700 -16555.601 0.209 0.154
Chain 1: 4800 -9534.777 0.267 0.171
Chain 1: 4900 -16111.105 0.291 0.345
Chain 1: 5000 -9817.504 0.342 0.350
Chain 1: 5100 -10323.823 0.334 0.350
Chain 1: 5200 -10988.902 0.305 0.350
Chain 1: 5300 -13869.934 0.257 0.208
Chain 1: 5400 -9892.133 0.297 0.350
Chain 1: 5500 -11739.280 0.303 0.350
Chain 1: 5600 -9054.940 0.331 0.350
Chain 1: 5700 -12521.350 0.324 0.296
Chain 1: 5800 -11641.131 0.257 0.277
Chain 1: 5900 -10806.859 0.224 0.208
Chain 1: 6000 -9555.000 0.173 0.157
Chain 1: 6100 -10056.159 0.173 0.157
Chain 1: 6200 -13400.908 0.192 0.208
Chain 1: 6300 -9240.753 0.217 0.250
Chain 1: 6400 -10065.014 0.185 0.157
Chain 1: 6500 -13098.528 0.192 0.232
Chain 1: 6600 -9359.895 0.202 0.232
Chain 1: 6700 -9783.828 0.179 0.131
Chain 1: 6800 -13032.945 0.196 0.232
Chain 1: 6900 -11747.414 0.200 0.232
Chain 1: 7000 -8977.554 0.217 0.249
Chain 1: 7100 -9159.354 0.214 0.249
Chain 1: 7200 -9427.356 0.192 0.232
Chain 1: 7300 -10806.507 0.160 0.128
Chain 1: 7400 -9176.010 0.170 0.178
Chain 1: 7500 -9731.285 0.152 0.128
Chain 1: 7600 -9007.622 0.120 0.109
Chain 1: 7700 -11474.120 0.137 0.128
Chain 1: 7800 -13469.255 0.127 0.128
Chain 1: 7900 -9221.697 0.162 0.148
Chain 1: 8000 -8939.666 0.135 0.128
Chain 1: 8100 -9725.411 0.141 0.128
Chain 1: 8200 -9773.109 0.138 0.128
Chain 1: 8300 -9298.976 0.131 0.081
Chain 1: 8400 -8999.747 0.116 0.080
Chain 1: 8500 -13445.867 0.144 0.081
Chain 1: 8600 -8920.291 0.186 0.148
Chain 1: 8700 -11130.521 0.185 0.148
Chain 1: 8800 -9377.852 0.189 0.187
Chain 1: 8900 -10535.003 0.153 0.110
Chain 1: 9000 -10859.328 0.153 0.110
Chain 1: 9100 -9208.881 0.163 0.179
Chain 1: 9200 -9147.738 0.163 0.179
Chain 1: 9300 -10852.429 0.174 0.179
Chain 1: 9400 -11372.268 0.175 0.179
Chain 1: 9500 -11505.369 0.143 0.157
Chain 1: 9600 -9153.490 0.118 0.157
Chain 1: 9700 -8770.843 0.103 0.110
Chain 1: 9800 -10731.637 0.102 0.110
Chain 1: 9900 -8777.545 0.114 0.157
Chain 1: 10000 -11247.542 0.133 0.179
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001379 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.79 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58865.249 1.000 1.000
Chain 1: 200 -18465.015 1.594 2.188
Chain 1: 300 -9027.527 1.411 1.045
Chain 1: 400 -8073.394 1.088 1.045
Chain 1: 500 -9139.828 0.894 1.000
Chain 1: 600 -9138.914 0.745 1.000
Chain 1: 700 -7949.592 0.660 0.150
Chain 1: 800 -8513.767 0.586 0.150
Chain 1: 900 -8463.128 0.521 0.118
Chain 1: 1000 -7734.179 0.478 0.118
Chain 1: 1100 -7648.109 0.380 0.117
Chain 1: 1200 -7951.740 0.165 0.094
Chain 1: 1300 -8054.875 0.061 0.066
Chain 1: 1400 -7788.149 0.053 0.038
Chain 1: 1500 -7481.270 0.045 0.038
Chain 1: 1600 -7756.345 0.049 0.038
Chain 1: 1700 -7424.878 0.038 0.038
Chain 1: 1800 -7595.911 0.034 0.035
Chain 1: 1900 -7710.867 0.035 0.035
Chain 1: 2000 -7614.087 0.027 0.034
Chain 1: 2100 -7544.002 0.027 0.034
Chain 1: 2200 -7843.857 0.027 0.034
Chain 1: 2300 -7633.438 0.028 0.034
Chain 1: 2400 -7642.489 0.025 0.028
Chain 1: 2500 -7381.929 0.024 0.028
Chain 1: 2600 -7512.918 0.022 0.023
Chain 1: 2700 -7493.910 0.018 0.017
Chain 1: 2800 -7472.423 0.016 0.015
Chain 1: 2900 -7342.250 0.016 0.017
Chain 1: 3000 -7535.779 0.018 0.018
Chain 1: 3100 -7508.051 0.017 0.018
Chain 1: 3200 -7706.280 0.016 0.018
Chain 1: 3300 -7392.985 0.017 0.018
Chain 1: 3400 -7657.407 0.021 0.026
Chain 1: 3500 -7438.921 0.020 0.026
Chain 1: 3600 -7447.387 0.019 0.026
Chain 1: 3700 -7383.075 0.019 0.026
Chain 1: 3800 -7420.599 0.019 0.026
Chain 1: 3900 -7370.355 0.018 0.026
Chain 1: 4000 -7370.896 0.016 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00358 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.8 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86477.727 1.000 1.000
Chain 1: 200 -14275.378 3.029 5.058
Chain 1: 300 -10610.183 2.134 1.000
Chain 1: 400 -11629.698 1.623 1.000
Chain 1: 500 -9601.557 1.340 0.345
Chain 1: 600 -9042.505 1.127 0.345
Chain 1: 700 -9190.344 0.969 0.211
Chain 1: 800 -9720.009 0.854 0.211
Chain 1: 900 -9346.772 0.764 0.088
Chain 1: 1000 -9332.008 0.688 0.088
Chain 1: 1100 -9475.126 0.589 0.062 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -9024.439 0.088 0.054
Chain 1: 1300 -9280.770 0.057 0.050
Chain 1: 1400 -9287.489 0.048 0.040
Chain 1: 1500 -9137.659 0.028 0.028
Chain 1: 1600 -9252.915 0.023 0.016
Chain 1: 1700 -9323.621 0.023 0.016
Chain 1: 1800 -8893.521 0.022 0.016
Chain 1: 1900 -8997.156 0.019 0.015
Chain 1: 2000 -8972.541 0.019 0.015
Chain 1: 2100 -9105.355 0.019 0.015
Chain 1: 2200 -8900.820 0.016 0.015
Chain 1: 2300 -8995.958 0.015 0.012
Chain 1: 2400 -9060.874 0.015 0.012
Chain 1: 2500 -9005.990 0.014 0.012
Chain 1: 2600 -9010.003 0.013 0.011
Chain 1: 2700 -8925.292 0.013 0.011
Chain 1: 2800 -8882.332 0.009 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003486 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.86 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8398030.358 1.000 1.000
Chain 1: 200 -1580951.547 2.656 4.312
Chain 1: 300 -890513.004 2.029 1.000
Chain 1: 400 -458477.725 1.757 1.000
Chain 1: 500 -358889.121 1.461 0.942
Chain 1: 600 -233964.963 1.307 0.942
Chain 1: 700 -120104.914 1.256 0.942
Chain 1: 800 -87326.698 1.146 0.942
Chain 1: 900 -67648.379 1.051 0.775
Chain 1: 1000 -52432.972 0.975 0.775
Chain 1: 1100 -39899.692 0.906 0.534 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39075.690 0.477 0.375
Chain 1: 1300 -27010.682 0.444 0.375
Chain 1: 1400 -26730.105 0.351 0.314
Chain 1: 1500 -23311.972 0.338 0.314
Chain 1: 1600 -22527.770 0.288 0.291
Chain 1: 1700 -21398.171 0.198 0.290
Chain 1: 1800 -21341.904 0.161 0.147
Chain 1: 1900 -21668.325 0.133 0.053
Chain 1: 2000 -20177.510 0.112 0.053
Chain 1: 2100 -20415.866 0.082 0.035
Chain 1: 2200 -20642.899 0.081 0.035
Chain 1: 2300 -20259.567 0.038 0.019
Chain 1: 2400 -20031.497 0.038 0.019
Chain 1: 2500 -19833.773 0.024 0.015
Chain 1: 2600 -19463.443 0.023 0.015
Chain 1: 2700 -19420.294 0.018 0.012
Chain 1: 2800 -19137.152 0.019 0.015
Chain 1: 2900 -19418.547 0.019 0.014
Chain 1: 3000 -19404.667 0.011 0.012
Chain 1: 3100 -19489.698 0.011 0.011
Chain 1: 3200 -19180.178 0.011 0.014
Chain 1: 3300 -19385.071 0.010 0.011
Chain 1: 3400 -18859.699 0.012 0.014
Chain 1: 3500 -19472.084 0.014 0.015
Chain 1: 3600 -18778.105 0.016 0.015
Chain 1: 3700 -19165.413 0.018 0.016
Chain 1: 3800 -18124.158 0.022 0.020
Chain 1: 3900 -18120.312 0.021 0.020
Chain 1: 4000 -18237.593 0.021 0.020
Chain 1: 4100 -18151.315 0.021 0.020
Chain 1: 4200 -17967.347 0.021 0.020
Chain 1: 4300 -18105.867 0.020 0.020
Chain 1: 4400 -18062.499 0.018 0.010
Chain 1: 4500 -17965.028 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001327 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.27 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48398.951 1.000 1.000
Chain 1: 200 -18167.584 1.332 1.664
Chain 1: 300 -38616.913 1.065 1.000
Chain 1: 400 -13814.314 1.247 1.664
Chain 1: 500 -11708.395 1.034 1.000
Chain 1: 600 -24006.759 0.947 1.000
Chain 1: 700 -18127.573 0.858 0.530
Chain 1: 800 -11111.086 0.830 0.631
Chain 1: 900 -19995.820 0.787 0.530
Chain 1: 1000 -9878.869 0.811 0.631
Chain 1: 1100 -16980.024 0.752 0.530 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -19565.891 0.599 0.512 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1300 -12019.293 0.609 0.512 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1400 -13590.690 0.441 0.444
Chain 1: 1500 -10598.348 0.451 0.444
Chain 1: 1600 -25894.140 0.459 0.444
Chain 1: 1700 -9729.146 0.593 0.591 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1800 -10002.923 0.532 0.444 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1900 -9955.796 0.488 0.418
Chain 1: 2000 -9972.109 0.386 0.282
Chain 1: 2100 -9432.597 0.350 0.132
Chain 1: 2200 -9302.833 0.338 0.116
Chain 1: 2300 -9160.017 0.277 0.057
Chain 1: 2400 -9696.076 0.271 0.055
Chain 1: 2500 -10310.433 0.249 0.055
Chain 1: 2600 -11158.229 0.197 0.055
Chain 1: 2700 -10044.315 0.042 0.055
Chain 1: 2800 -14267.692 0.069 0.057
Chain 1: 2900 -9027.139 0.127 0.060
Chain 1: 3000 -10528.218 0.141 0.076
Chain 1: 3100 -9670.916 0.144 0.089
Chain 1: 3200 -13510.689 0.171 0.111
Chain 1: 3300 -10224.493 0.202 0.143
Chain 1: 3400 -9245.846 0.207 0.143
Chain 1: 3500 -9181.415 0.201 0.143
Chain 1: 3600 -10071.877 0.203 0.143
Chain 1: 3700 -8846.058 0.205 0.143
Chain 1: 3800 -10972.565 0.195 0.143
Chain 1: 3900 -9103.004 0.158 0.143
Chain 1: 4000 -8603.004 0.149 0.139
Chain 1: 4100 -9312.350 0.148 0.139
Chain 1: 4200 -8990.529 0.123 0.106
Chain 1: 4300 -8989.292 0.091 0.088
Chain 1: 4400 -9222.874 0.083 0.076
Chain 1: 4500 -8965.763 0.085 0.076
Chain 1: 4600 -12498.609 0.104 0.076
Chain 1: 4700 -15690.252 0.111 0.076
Chain 1: 4800 -8696.129 0.172 0.076
Chain 1: 4900 -8537.860 0.153 0.058
Chain 1: 5000 -9178.854 0.154 0.070
Chain 1: 5100 -13383.353 0.178 0.070
Chain 1: 5200 -8610.897 0.230 0.203
Chain 1: 5300 -12069.731 0.259 0.283
Chain 1: 5400 -9694.550 0.281 0.283
Chain 1: 5500 -9192.061 0.283 0.283
Chain 1: 5600 -13530.687 0.287 0.287
Chain 1: 5700 -11214.529 0.287 0.287
Chain 1: 5800 -10995.960 0.209 0.245
Chain 1: 5900 -9312.920 0.225 0.245
Chain 1: 6000 -8152.997 0.232 0.245
Chain 1: 6100 -9118.129 0.212 0.207
Chain 1: 6200 -8996.282 0.158 0.181
Chain 1: 6300 -8439.842 0.136 0.142
Chain 1: 6400 -9738.424 0.124 0.133
Chain 1: 6500 -14463.533 0.152 0.142
Chain 1: 6600 -8358.190 0.193 0.142
Chain 1: 6700 -8157.274 0.174 0.133
Chain 1: 6800 -11138.164 0.199 0.142
Chain 1: 6900 -11791.042 0.187 0.133
Chain 1: 7000 -12382.914 0.177 0.106
Chain 1: 7100 -8710.726 0.209 0.133
Chain 1: 7200 -11117.622 0.229 0.216
Chain 1: 7300 -8206.803 0.258 0.268
Chain 1: 7400 -8411.492 0.247 0.268
Chain 1: 7500 -8342.574 0.215 0.216
Chain 1: 7600 -8843.705 0.148 0.057
Chain 1: 7700 -8362.577 0.151 0.058
Chain 1: 7800 -13278.387 0.161 0.058
Chain 1: 7900 -8079.825 0.220 0.216
Chain 1: 8000 -10432.343 0.238 0.226
Chain 1: 8100 -8287.565 0.222 0.226
Chain 1: 8200 -8132.801 0.202 0.226
Chain 1: 8300 -8692.949 0.173 0.064
Chain 1: 8400 -8447.860 0.173 0.064
Chain 1: 8500 -8566.048 0.174 0.064
Chain 1: 8600 -11032.538 0.191 0.224
Chain 1: 8700 -8170.000 0.220 0.226
Chain 1: 8800 -10360.278 0.204 0.224
Chain 1: 8900 -11671.842 0.151 0.211
Chain 1: 9000 -10580.332 0.139 0.112
Chain 1: 9100 -8405.037 0.139 0.112
Chain 1: 9200 -8865.945 0.142 0.112
Chain 1: 9300 -10835.302 0.154 0.182
Chain 1: 9400 -8159.791 0.184 0.211
Chain 1: 9500 -10470.792 0.204 0.221
Chain 1: 9600 -8242.334 0.209 0.221
Chain 1: 9700 -8003.218 0.177 0.211
Chain 1: 9800 -9850.290 0.174 0.188
Chain 1: 9900 -8112.642 0.185 0.214
Chain 1: 10000 -8139.893 0.175 0.214
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001632 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 16.32 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57710.526 1.000 1.000
Chain 1: 200 -17377.021 1.661 2.321
Chain 1: 300 -8577.952 1.449 1.026
Chain 1: 400 -8182.543 1.099 1.026
Chain 1: 500 -8432.374 0.885 1.000
Chain 1: 600 -8832.612 0.745 1.000
Chain 1: 700 -8234.128 0.649 0.073
Chain 1: 800 -8162.806 0.569 0.073
Chain 1: 900 -7874.418 0.510 0.048
Chain 1: 1000 -7820.185 0.460 0.048
Chain 1: 1100 -7737.403 0.361 0.045
Chain 1: 1200 -7804.883 0.129 0.037
Chain 1: 1300 -7693.285 0.028 0.030
Chain 1: 1400 -7671.002 0.024 0.015
Chain 1: 1500 -7631.730 0.021 0.011
Chain 1: 1600 -7583.812 0.017 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003432 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.32 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85374.789 1.000 1.000
Chain 1: 200 -13226.283 3.227 5.455
Chain 1: 300 -9682.366 2.274 1.000
Chain 1: 400 -10396.694 1.722 1.000
Chain 1: 500 -8585.524 1.420 0.366
Chain 1: 600 -8387.725 1.187 0.366
Chain 1: 700 -8525.389 1.020 0.211
Chain 1: 800 -8541.076 0.893 0.211
Chain 1: 900 -8456.515 0.795 0.069
Chain 1: 1000 -8266.308 0.718 0.069
Chain 1: 1100 -8569.500 0.621 0.035 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8324.174 0.079 0.029
Chain 1: 1300 -8273.834 0.043 0.024
Chain 1: 1400 -8446.804 0.038 0.023
Chain 1: 1500 -8287.097 0.019 0.020
Chain 1: 1600 -8397.133 0.017 0.019
Chain 1: 1700 -8479.877 0.017 0.019
Chain 1: 1800 -8090.132 0.021 0.020
Chain 1: 1900 -8192.871 0.022 0.020
Chain 1: 2000 -8162.715 0.020 0.019
Chain 1: 2100 -8291.381 0.018 0.016
Chain 1: 2200 -8077.911 0.018 0.016
Chain 1: 2300 -8221.531 0.019 0.017
Chain 1: 2400 -8235.753 0.017 0.016
Chain 1: 2500 -8202.813 0.015 0.013
Chain 1: 2600 -8204.065 0.014 0.013
Chain 1: 2700 -8111.416 0.014 0.013
Chain 1: 2800 -8085.751 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003398 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.98 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8388786.357 1.000 1.000
Chain 1: 200 -1581275.559 2.653 4.305
Chain 1: 300 -890717.792 2.027 1.000
Chain 1: 400 -458065.821 1.756 1.000
Chain 1: 500 -358404.349 1.461 0.945
Chain 1: 600 -233269.658 1.307 0.945
Chain 1: 700 -119217.020 1.257 0.945
Chain 1: 800 -86366.190 1.147 0.945
Chain 1: 900 -66649.520 1.052 0.775
Chain 1: 1000 -51398.013 0.977 0.775
Chain 1: 1100 -38835.178 0.909 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38002.146 0.481 0.380
Chain 1: 1300 -25924.092 0.450 0.380
Chain 1: 1400 -25637.391 0.357 0.323
Chain 1: 1500 -22217.093 0.344 0.323
Chain 1: 1600 -21430.867 0.294 0.297
Chain 1: 1700 -20301.031 0.204 0.296
Chain 1: 1800 -20244.101 0.166 0.154
Chain 1: 1900 -20569.802 0.138 0.056
Chain 1: 2000 -19080.225 0.117 0.056
Chain 1: 2100 -19318.312 0.085 0.037
Chain 1: 2200 -19544.935 0.084 0.037
Chain 1: 2300 -19162.171 0.040 0.020
Chain 1: 2400 -18934.446 0.040 0.020
Chain 1: 2500 -18736.699 0.026 0.016
Chain 1: 2600 -18367.020 0.024 0.016
Chain 1: 2700 -18324.084 0.019 0.012
Chain 1: 2800 -18041.267 0.020 0.016
Chain 1: 2900 -18322.387 0.020 0.015
Chain 1: 3000 -18308.490 0.012 0.012
Chain 1: 3100 -18393.416 0.011 0.012
Chain 1: 3200 -18084.334 0.012 0.015
Chain 1: 3300 -18288.899 0.011 0.012
Chain 1: 3400 -17764.330 0.013 0.015
Chain 1: 3500 -18375.472 0.015 0.016
Chain 1: 3600 -17683.198 0.017 0.016
Chain 1: 3700 -18069.267 0.019 0.017
Chain 1: 3800 -17030.562 0.023 0.021
Chain 1: 3900 -17026.827 0.022 0.021
Chain 1: 4000 -17144.076 0.022 0.021
Chain 1: 4100 -17057.925 0.022 0.021
Chain 1: 4200 -16874.559 0.022 0.021
Chain 1: 4300 -17012.637 0.022 0.021
Chain 1: 4400 -16969.744 0.019 0.011
Chain 1: 4500 -16872.413 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001456 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.56 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48736.523 1.000 1.000
Chain 1: 200 -19228.415 1.267 1.535
Chain 1: 300 -20057.140 0.859 1.000
Chain 1: 400 -23348.762 0.679 1.000
Chain 1: 500 -13803.398 0.682 0.692
Chain 1: 600 -14035.697 0.571 0.692
Chain 1: 700 -15211.196 0.500 0.141
Chain 1: 800 -15447.277 0.440 0.141
Chain 1: 900 -11619.002 0.427 0.141
Chain 1: 1000 -14834.810 0.406 0.217
Chain 1: 1100 -11699.298 0.333 0.217
Chain 1: 1200 -10713.272 0.189 0.141
Chain 1: 1300 -11152.235 0.189 0.141
Chain 1: 1400 -10646.761 0.179 0.092
Chain 1: 1500 -11704.103 0.119 0.090
Chain 1: 1600 -12234.144 0.122 0.090
Chain 1: 1700 -15785.536 0.137 0.092
Chain 1: 1800 -10928.609 0.180 0.217
Chain 1: 1900 -9745.494 0.159 0.121
Chain 1: 2000 -23313.455 0.195 0.121
Chain 1: 2100 -9399.749 0.317 0.121
Chain 1: 2200 -10664.542 0.319 0.121
Chain 1: 2300 -13982.914 0.339 0.225
Chain 1: 2400 -8985.016 0.390 0.237
Chain 1: 2500 -16946.483 0.428 0.444
Chain 1: 2600 -9580.356 0.500 0.470 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2700 -9070.035 0.484 0.470
Chain 1: 2800 -9227.147 0.441 0.470
Chain 1: 2900 -9651.139 0.433 0.470
Chain 1: 3000 -11580.433 0.391 0.237
Chain 1: 3100 -9809.305 0.262 0.181
Chain 1: 3200 -9308.814 0.255 0.181
Chain 1: 3300 -9017.978 0.235 0.167
Chain 1: 3400 -9738.749 0.186 0.074
Chain 1: 3500 -12651.059 0.162 0.074
Chain 1: 3600 -10335.329 0.108 0.074
Chain 1: 3700 -9424.188 0.112 0.097
Chain 1: 3800 -10518.934 0.121 0.104
Chain 1: 3900 -15403.267 0.148 0.167
Chain 1: 4000 -9312.463 0.197 0.181
Chain 1: 4100 -9521.222 0.181 0.104
Chain 1: 4200 -9409.984 0.177 0.104
Chain 1: 4300 -13044.245 0.201 0.224
Chain 1: 4400 -12166.547 0.201 0.224
Chain 1: 4500 -9582.508 0.205 0.224
Chain 1: 4600 -12622.410 0.207 0.241
Chain 1: 4700 -8583.544 0.244 0.270
Chain 1: 4800 -8773.218 0.236 0.270
Chain 1: 4900 -13770.081 0.240 0.270
Chain 1: 5000 -9048.296 0.227 0.270
Chain 1: 5100 -10281.861 0.237 0.270
Chain 1: 5200 -11351.927 0.245 0.270
Chain 1: 5300 -12333.292 0.225 0.241
Chain 1: 5400 -12308.303 0.218 0.241
Chain 1: 5500 -8249.539 0.241 0.241
Chain 1: 5600 -9068.138 0.225 0.120
Chain 1: 5700 -15488.827 0.220 0.120
Chain 1: 5800 -9927.745 0.274 0.363
Chain 1: 5900 -9341.516 0.244 0.120
Chain 1: 6000 -9145.783 0.194 0.094
Chain 1: 6100 -15223.549 0.222 0.094
Chain 1: 6200 -8938.883 0.283 0.399
Chain 1: 6300 -8572.491 0.279 0.399
Chain 1: 6400 -8603.408 0.279 0.399
Chain 1: 6500 -8357.189 0.233 0.090
Chain 1: 6600 -9441.191 0.235 0.115
Chain 1: 6700 -9387.709 0.194 0.063
Chain 1: 6800 -8415.635 0.150 0.063
Chain 1: 6900 -13035.416 0.179 0.115
Chain 1: 7000 -9142.392 0.219 0.116
Chain 1: 7100 -11251.556 0.198 0.116
Chain 1: 7200 -8370.053 0.162 0.116
Chain 1: 7300 -8671.999 0.162 0.116
Chain 1: 7400 -11003.815 0.182 0.187
Chain 1: 7500 -9898.665 0.191 0.187
Chain 1: 7600 -11696.442 0.195 0.187
Chain 1: 7700 -8898.421 0.225 0.212
Chain 1: 7800 -8880.610 0.214 0.212
Chain 1: 7900 -8194.326 0.187 0.187
Chain 1: 8000 -8604.771 0.149 0.154
Chain 1: 8100 -8442.703 0.132 0.112
Chain 1: 8200 -10699.280 0.119 0.112
Chain 1: 8300 -10532.319 0.117 0.112
Chain 1: 8400 -9998.672 0.101 0.084
Chain 1: 8500 -8569.856 0.107 0.084
Chain 1: 8600 -8043.244 0.098 0.065
Chain 1: 8700 -9893.562 0.085 0.065
Chain 1: 8800 -8268.432 0.105 0.084
Chain 1: 8900 -9068.651 0.105 0.088
Chain 1: 9000 -11717.178 0.123 0.167
Chain 1: 9100 -8302.280 0.162 0.187
Chain 1: 9200 -8566.344 0.144 0.167
Chain 1: 9300 -9405.069 0.151 0.167
Chain 1: 9400 -9991.804 0.152 0.167
Chain 1: 9500 -8326.574 0.155 0.187
Chain 1: 9600 -8324.613 0.149 0.187
Chain 1: 9700 -8011.732 0.134 0.089
Chain 1: 9800 -8127.020 0.116 0.088
Chain 1: 9900 -9605.777 0.122 0.089
Chain 1: 10000 -9160.946 0.105 0.059
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001543 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.43 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56885.888 1.000 1.000
Chain 1: 200 -17436.841 1.631 2.262
Chain 1: 300 -8711.879 1.421 1.002
Chain 1: 400 -8352.026 1.077 1.002
Chain 1: 500 -8221.741 0.865 1.000
Chain 1: 600 -8664.951 0.729 1.000
Chain 1: 700 -8739.628 0.626 0.051
Chain 1: 800 -8032.707 0.559 0.088
Chain 1: 900 -8031.925 0.497 0.051
Chain 1: 1000 -7822.342 0.450 0.051
Chain 1: 1100 -7537.995 0.354 0.043
Chain 1: 1200 -7678.053 0.129 0.038
Chain 1: 1300 -7531.252 0.031 0.027
Chain 1: 1400 -7594.927 0.027 0.019
Chain 1: 1500 -7569.283 0.026 0.019
Chain 1: 1600 -7713.897 0.023 0.019
Chain 1: 1700 -7467.354 0.025 0.019
Chain 1: 1800 -7559.314 0.018 0.019
Chain 1: 1900 -7543.679 0.018 0.019
Chain 1: 2000 -7597.792 0.016 0.018
Chain 1: 2100 -7544.770 0.013 0.012
Chain 1: 2200 -7660.389 0.013 0.012
Chain 1: 2300 -7554.671 0.012 0.012
Chain 1: 2400 -7608.963 0.012 0.012
Chain 1: 2500 -7518.068 0.013 0.012
Chain 1: 2600 -7487.874 0.011 0.012
Chain 1: 2700 -7499.911 0.008 0.007 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002975 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.75 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86979.583 1.000 1.000
Chain 1: 200 -13495.326 3.223 5.445
Chain 1: 300 -9847.183 2.272 1.000
Chain 1: 400 -10728.924 1.724 1.000
Chain 1: 500 -8808.005 1.423 0.370
Chain 1: 600 -8322.294 1.196 0.370
Chain 1: 700 -8386.231 1.026 0.218
Chain 1: 800 -8985.657 0.906 0.218
Chain 1: 900 -8682.454 0.809 0.082
Chain 1: 1000 -8495.578 0.731 0.082
Chain 1: 1100 -8661.425 0.632 0.067 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8411.168 0.091 0.058
Chain 1: 1300 -8514.593 0.055 0.035
Chain 1: 1400 -8529.906 0.047 0.030
Chain 1: 1500 -8405.530 0.027 0.022
Chain 1: 1600 -8522.542 0.022 0.019
Chain 1: 1700 -8607.924 0.022 0.019
Chain 1: 1800 -8193.947 0.021 0.019
Chain 1: 1900 -8290.022 0.019 0.015
Chain 1: 2000 -8263.474 0.017 0.014
Chain 1: 2100 -8386.197 0.016 0.014
Chain 1: 2200 -8206.247 0.015 0.014
Chain 1: 2300 -8284.990 0.015 0.014
Chain 1: 2400 -8354.709 0.016 0.014
Chain 1: 2500 -8300.172 0.015 0.012
Chain 1: 2600 -8299.687 0.014 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00347 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.7 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8412788.092 1.000 1.000
Chain 1: 200 -1585298.652 2.653 4.307
Chain 1: 300 -892139.258 2.028 1.000
Chain 1: 400 -458141.369 1.758 1.000
Chain 1: 500 -358573.522 1.462 0.947
Chain 1: 600 -233354.684 1.308 0.947
Chain 1: 700 -119422.279 1.257 0.947
Chain 1: 800 -86558.237 1.147 0.947
Chain 1: 900 -66863.942 1.053 0.777
Chain 1: 1000 -51629.741 0.977 0.777
Chain 1: 1100 -39080.612 0.909 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38251.517 0.480 0.380
Chain 1: 1300 -26188.912 0.449 0.380
Chain 1: 1400 -25904.249 0.355 0.321
Chain 1: 1500 -22486.546 0.343 0.321
Chain 1: 1600 -21701.100 0.293 0.295
Chain 1: 1700 -20573.348 0.203 0.295
Chain 1: 1800 -20516.817 0.165 0.152
Chain 1: 1900 -20842.921 0.137 0.055
Chain 1: 2000 -19353.086 0.115 0.055
Chain 1: 2100 -19591.592 0.084 0.036
Chain 1: 2200 -19818.096 0.083 0.036
Chain 1: 2300 -19435.245 0.039 0.020
Chain 1: 2400 -19207.359 0.039 0.020
Chain 1: 2500 -19009.296 0.025 0.016
Chain 1: 2600 -18639.624 0.024 0.016
Chain 1: 2700 -18596.571 0.018 0.012
Chain 1: 2800 -18313.473 0.020 0.015
Chain 1: 2900 -18594.696 0.020 0.015
Chain 1: 3000 -18580.879 0.012 0.012
Chain 1: 3100 -18665.894 0.011 0.012
Chain 1: 3200 -18356.588 0.012 0.015
Chain 1: 3300 -18561.269 0.011 0.012
Chain 1: 3400 -18036.238 0.013 0.015
Chain 1: 3500 -18648.030 0.015 0.015
Chain 1: 3600 -17954.808 0.017 0.015
Chain 1: 3700 -18341.602 0.019 0.017
Chain 1: 3800 -17301.415 0.023 0.021
Chain 1: 3900 -17297.546 0.022 0.021
Chain 1: 4000 -17414.863 0.022 0.021
Chain 1: 4100 -17328.645 0.022 0.021
Chain 1: 4200 -17144.888 0.022 0.021
Chain 1: 4300 -17283.297 0.021 0.021
Chain 1: 4400 -17240.175 0.019 0.011
Chain 1: 4500 -17142.676 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001306 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.06 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48649.731 1.000 1.000
Chain 1: 200 -21080.044 1.154 1.308
Chain 1: 300 -17821.496 0.830 1.000
Chain 1: 400 -20016.633 0.650 1.000
Chain 1: 500 -11098.100 0.681 0.804
Chain 1: 600 -12186.900 0.582 0.804
Chain 1: 700 -15735.138 0.531 0.225
Chain 1: 800 -12776.700 0.494 0.232
Chain 1: 900 -10983.653 0.457 0.225
Chain 1: 1000 -11757.221 0.418 0.225
Chain 1: 1100 -19700.530 0.358 0.225
Chain 1: 1200 -16627.123 0.246 0.185
Chain 1: 1300 -12770.221 0.258 0.225
Chain 1: 1400 -9582.335 0.280 0.232
Chain 1: 1500 -10776.632 0.211 0.225
Chain 1: 1600 -9329.047 0.217 0.225
Chain 1: 1700 -12184.268 0.218 0.232
Chain 1: 1800 -12152.768 0.195 0.185
Chain 1: 1900 -11762.201 0.182 0.185
Chain 1: 2000 -10366.299 0.189 0.185
Chain 1: 2100 -10241.335 0.150 0.155
Chain 1: 2200 -12213.343 0.148 0.155
Chain 1: 2300 -9713.767 0.143 0.155
Chain 1: 2400 -9022.414 0.118 0.135
Chain 1: 2500 -13296.292 0.139 0.155
Chain 1: 2600 -9145.027 0.169 0.161
Chain 1: 2700 -13979.807 0.180 0.161
Chain 1: 2800 -10076.265 0.218 0.257
Chain 1: 2900 -10699.615 0.221 0.257
Chain 1: 3000 -9682.084 0.218 0.257
Chain 1: 3100 -9643.799 0.217 0.257
Chain 1: 3200 -10887.042 0.212 0.257
Chain 1: 3300 -14895.478 0.214 0.269
Chain 1: 3400 -11804.240 0.232 0.269
Chain 1: 3500 -8701.296 0.236 0.269
Chain 1: 3600 -9919.737 0.203 0.262
Chain 1: 3700 -9121.219 0.177 0.123
Chain 1: 3800 -15393.504 0.179 0.123
Chain 1: 3900 -8934.410 0.245 0.262
Chain 1: 4000 -8434.952 0.241 0.262
Chain 1: 4100 -8439.683 0.240 0.262
Chain 1: 4200 -8344.461 0.230 0.262
Chain 1: 4300 -9685.031 0.217 0.138
Chain 1: 4400 -12478.821 0.213 0.138
Chain 1: 4500 -9913.029 0.203 0.138
Chain 1: 4600 -13141.633 0.216 0.224
Chain 1: 4700 -8261.091 0.266 0.246
Chain 1: 4800 -8255.748 0.225 0.224
Chain 1: 4900 -8525.771 0.156 0.138
Chain 1: 5000 -10648.916 0.170 0.199
Chain 1: 5100 -8405.853 0.197 0.224
Chain 1: 5200 -8812.043 0.200 0.224
Chain 1: 5300 -9419.386 0.193 0.224
Chain 1: 5400 -8115.935 0.187 0.199
Chain 1: 5500 -8588.452 0.166 0.161
Chain 1: 5600 -14951.165 0.184 0.161
Chain 1: 5700 -9484.519 0.183 0.161
Chain 1: 5800 -8793.782 0.190 0.161
Chain 1: 5900 -13342.826 0.221 0.199
Chain 1: 6000 -8317.047 0.262 0.267
Chain 1: 6100 -8922.871 0.242 0.161
Chain 1: 6200 -9885.153 0.247 0.161
Chain 1: 6300 -12314.388 0.260 0.197
Chain 1: 6400 -9989.957 0.268 0.233
Chain 1: 6500 -8525.986 0.279 0.233
Chain 1: 6600 -11778.961 0.264 0.233
Chain 1: 6700 -13313.534 0.218 0.197
Chain 1: 6800 -12112.748 0.220 0.197
Chain 1: 6900 -12297.910 0.188 0.172
Chain 1: 7000 -8450.935 0.173 0.172
Chain 1: 7100 -8160.012 0.170 0.172
Chain 1: 7200 -8180.944 0.160 0.172
Chain 1: 7300 -9491.951 0.154 0.138
Chain 1: 7400 -8294.570 0.145 0.138
Chain 1: 7500 -8096.998 0.131 0.115
Chain 1: 7600 -8310.801 0.106 0.099
Chain 1: 7700 -8343.227 0.094 0.036
Chain 1: 7800 -8755.782 0.089 0.036
Chain 1: 7900 -8140.512 0.095 0.047
Chain 1: 8000 -8101.516 0.050 0.036
Chain 1: 8100 -8254.782 0.049 0.026
Chain 1: 8200 -8459.774 0.051 0.026
Chain 1: 8300 -8056.983 0.042 0.026
Chain 1: 8400 -8450.139 0.032 0.026
Chain 1: 8500 -8526.527 0.031 0.026
Chain 1: 8600 -7931.798 0.035 0.047
Chain 1: 8700 -9435.640 0.051 0.047
Chain 1: 8800 -9079.795 0.050 0.047
Chain 1: 8900 -8461.388 0.050 0.047
Chain 1: 9000 -8220.632 0.052 0.047
Chain 1: 9100 -9524.254 0.064 0.050
Chain 1: 9200 -8249.626 0.077 0.073
Chain 1: 9300 -11223.764 0.099 0.075
Chain 1: 9400 -10570.444 0.100 0.075
Chain 1: 9500 -7988.188 0.132 0.137
Chain 1: 9600 -8095.119 0.126 0.137
Chain 1: 9700 -11091.350 0.137 0.137
Chain 1: 9800 -8133.663 0.169 0.155
Chain 1: 9900 -8529.079 0.166 0.155
Chain 1: 10000 -9584.949 0.174 0.155
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001426 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.26 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -62916.745 1.000 1.000
Chain 1: 200 -17808.866 1.766 2.533
Chain 1: 300 -8612.194 1.534 1.068
Chain 1: 400 -8168.055 1.164 1.068
Chain 1: 500 -8245.612 0.933 1.000
Chain 1: 600 -8824.621 0.788 1.000
Chain 1: 700 -7835.150 0.694 0.126
Chain 1: 800 -8071.479 0.611 0.126
Chain 1: 900 -7932.604 0.545 0.066
Chain 1: 1000 -7848.639 0.491 0.066
Chain 1: 1100 -7707.312 0.393 0.054
Chain 1: 1200 -7581.448 0.142 0.029
Chain 1: 1300 -7726.531 0.037 0.019
Chain 1: 1400 -7702.871 0.032 0.018
Chain 1: 1500 -7612.471 0.032 0.018
Chain 1: 1600 -7537.683 0.026 0.018
Chain 1: 1700 -7519.419 0.014 0.017
Chain 1: 1800 -7526.988 0.011 0.012
Chain 1: 1900 -7613.297 0.010 0.011
Chain 1: 2000 -7611.326 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003273 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.73 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86403.563 1.000 1.000
Chain 1: 200 -13108.966 3.296 5.591
Chain 1: 300 -9582.747 2.320 1.000
Chain 1: 400 -10481.038 1.761 1.000
Chain 1: 500 -8473.252 1.456 0.368
Chain 1: 600 -8296.450 1.217 0.368
Chain 1: 700 -8459.794 1.046 0.237
Chain 1: 800 -8702.486 0.919 0.237
Chain 1: 900 -8432.453 0.820 0.086
Chain 1: 1000 -8166.914 0.741 0.086
Chain 1: 1100 -8398.810 0.644 0.033 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8130.774 0.088 0.033
Chain 1: 1300 -8166.695 0.052 0.032
Chain 1: 1400 -8270.433 0.045 0.028
Chain 1: 1500 -8201.329 0.022 0.028
Chain 1: 1600 -8207.861 0.020 0.028
Chain 1: 1700 -8144.905 0.019 0.028
Chain 1: 1800 -8025.527 0.017 0.015
Chain 1: 1900 -8141.456 0.016 0.014
Chain 1: 2000 -8101.452 0.013 0.013
Chain 1: 2100 -8239.239 0.012 0.013
Chain 1: 2200 -8025.699 0.011 0.013
Chain 1: 2300 -8167.054 0.012 0.014
Chain 1: 2400 -8175.109 0.011 0.014
Chain 1: 2500 -8144.443 0.011 0.014
Chain 1: 2600 -8139.277 0.011 0.014
Chain 1: 2700 -8049.844 0.011 0.014
Chain 1: 2800 -8030.542 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003359 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.59 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8435595.773 1.000 1.000
Chain 1: 200 -1591708.083 2.650 4.300
Chain 1: 300 -890916.478 2.029 1.000
Chain 1: 400 -456815.273 1.759 1.000
Chain 1: 500 -356796.463 1.463 0.950
Chain 1: 600 -231747.611 1.309 0.950
Chain 1: 700 -118376.360 1.259 0.950
Chain 1: 800 -85672.134 1.149 0.950
Chain 1: 900 -66100.916 1.055 0.787
Chain 1: 1000 -50962.768 0.979 0.787
Chain 1: 1100 -38502.514 0.911 0.540 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37680.184 0.483 0.382
Chain 1: 1300 -25714.736 0.451 0.382
Chain 1: 1400 -25437.120 0.357 0.324
Chain 1: 1500 -22045.246 0.345 0.324
Chain 1: 1600 -21266.625 0.294 0.297
Chain 1: 1700 -20150.769 0.204 0.296
Chain 1: 1800 -20096.768 0.166 0.154
Chain 1: 1900 -20422.275 0.138 0.055
Chain 1: 2000 -18939.983 0.116 0.055
Chain 1: 2100 -19177.986 0.085 0.037
Chain 1: 2200 -19403.103 0.084 0.037
Chain 1: 2300 -19021.650 0.040 0.020
Chain 1: 2400 -18794.107 0.040 0.020
Chain 1: 2500 -18595.735 0.026 0.016
Chain 1: 2600 -18227.066 0.024 0.016
Chain 1: 2700 -18184.350 0.019 0.012
Chain 1: 2800 -17901.365 0.020 0.016
Chain 1: 2900 -18182.172 0.020 0.015
Chain 1: 3000 -18168.508 0.012 0.012
Chain 1: 3100 -18253.372 0.011 0.012
Chain 1: 3200 -17944.616 0.012 0.015
Chain 1: 3300 -18148.884 0.011 0.012
Chain 1: 3400 -17624.671 0.013 0.015
Chain 1: 3500 -18235.126 0.015 0.016
Chain 1: 3600 -17543.636 0.017 0.016
Chain 1: 3700 -17929.025 0.019 0.017
Chain 1: 3800 -16891.482 0.023 0.021
Chain 1: 3900 -16887.644 0.022 0.021
Chain 1: 4000 -17004.991 0.023 0.021
Chain 1: 4100 -16918.879 0.023 0.021
Chain 1: 4200 -16735.723 0.022 0.021
Chain 1: 4300 -16873.742 0.022 0.021
Chain 1: 4400 -16831.067 0.019 0.011
Chain 1: 4500 -16733.643 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001256 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.56 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12350.934 1.000 1.000
Chain 1: 200 -9291.762 0.665 1.000
Chain 1: 300 -8054.807 0.494 0.329
Chain 1: 400 -8272.256 0.377 0.329
Chain 1: 500 -8145.030 0.305 0.154
Chain 1: 600 -7997.829 0.257 0.154
Chain 1: 700 -7937.052 0.222 0.026
Chain 1: 800 -7925.787 0.194 0.026
Chain 1: 900 -8012.160 0.174 0.018
Chain 1: 1000 -7984.018 0.157 0.018
Chain 1: 1100 -8036.784 0.057 0.016
Chain 1: 1200 -7940.053 0.026 0.012
Chain 1: 1300 -7907.590 0.011 0.011
Chain 1: 1400 -7931.040 0.008 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001383 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.83 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61892.378 1.000 1.000
Chain 1: 200 -17926.354 1.726 2.453
Chain 1: 300 -8865.306 1.492 1.022
Chain 1: 400 -9437.682 1.134 1.022
Chain 1: 500 -8740.068 0.923 1.000
Chain 1: 600 -8784.249 0.770 1.000
Chain 1: 700 -8235.821 0.670 0.080
Chain 1: 800 -8263.247 0.586 0.080
Chain 1: 900 -7960.301 0.525 0.067
Chain 1: 1000 -7625.808 0.477 0.067
Chain 1: 1100 -7626.825 0.377 0.061
Chain 1: 1200 -7636.184 0.132 0.044
Chain 1: 1300 -7687.514 0.031 0.038
Chain 1: 1400 -7928.829 0.028 0.030
Chain 1: 1500 -7613.495 0.024 0.030
Chain 1: 1600 -7773.223 0.025 0.030
Chain 1: 1700 -7503.012 0.022 0.030
Chain 1: 1800 -7568.688 0.023 0.030
Chain 1: 1900 -7593.611 0.019 0.021
Chain 1: 2000 -7631.901 0.015 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00314 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.4 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86238.468 1.000 1.000
Chain 1: 200 -13503.848 3.193 5.386
Chain 1: 300 -9885.307 2.251 1.000
Chain 1: 400 -10797.085 1.709 1.000
Chain 1: 500 -8854.285 1.411 0.366
Chain 1: 600 -8339.113 1.186 0.366
Chain 1: 700 -8428.284 1.018 0.219
Chain 1: 800 -9176.612 0.901 0.219
Chain 1: 900 -8744.398 0.807 0.084
Chain 1: 1000 -8367.511 0.730 0.084
Chain 1: 1100 -8731.772 0.635 0.082 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8383.281 0.100 0.062
Chain 1: 1300 -8596.732 0.066 0.049
Chain 1: 1400 -8603.023 0.058 0.045
Chain 1: 1500 -8461.345 0.037 0.042
Chain 1: 1600 -8573.327 0.033 0.042
Chain 1: 1700 -8660.134 0.032 0.042
Chain 1: 1800 -8253.336 0.029 0.042
Chain 1: 1900 -8349.521 0.025 0.025
Chain 1: 2000 -8321.814 0.021 0.017
Chain 1: 2100 -8442.560 0.019 0.014
Chain 1: 2200 -8261.744 0.017 0.014
Chain 1: 2300 -8388.894 0.016 0.014
Chain 1: 2400 -8399.399 0.016 0.014
Chain 1: 2500 -8361.430 0.014 0.013
Chain 1: 2600 -8360.257 0.013 0.012
Chain 1: 2700 -8275.107 0.013 0.012
Chain 1: 2800 -8239.974 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003221 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.21 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8403619.041 1.000 1.000
Chain 1: 200 -1584749.415 2.651 4.303
Chain 1: 300 -892145.973 2.026 1.000
Chain 1: 400 -458735.803 1.756 1.000
Chain 1: 500 -359099.309 1.460 0.945
Chain 1: 600 -233768.599 1.306 0.945
Chain 1: 700 -119581.326 1.256 0.945
Chain 1: 800 -86695.587 1.146 0.945
Chain 1: 900 -66962.420 1.052 0.776
Chain 1: 1000 -51700.949 0.976 0.776
Chain 1: 1100 -39130.287 0.908 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38296.039 0.480 0.379
Chain 1: 1300 -26211.787 0.449 0.379
Chain 1: 1400 -25925.212 0.355 0.321
Chain 1: 1500 -22502.918 0.343 0.321
Chain 1: 1600 -21715.964 0.293 0.295
Chain 1: 1700 -20585.374 0.203 0.295
Chain 1: 1800 -20528.219 0.165 0.152
Chain 1: 1900 -20854.152 0.137 0.055
Chain 1: 2000 -19363.349 0.115 0.055
Chain 1: 2100 -19601.750 0.084 0.036
Chain 1: 2200 -19828.559 0.083 0.036
Chain 1: 2300 -19445.505 0.039 0.020
Chain 1: 2400 -19217.619 0.039 0.020
Chain 1: 2500 -19019.808 0.025 0.016
Chain 1: 2600 -18650.039 0.024 0.016
Chain 1: 2700 -18606.912 0.018 0.012
Chain 1: 2800 -18323.990 0.020 0.015
Chain 1: 2900 -18605.181 0.020 0.015
Chain 1: 3000 -18591.313 0.012 0.012
Chain 1: 3100 -18676.322 0.011 0.012
Chain 1: 3200 -18367.052 0.012 0.015
Chain 1: 3300 -18571.691 0.011 0.012
Chain 1: 3400 -18046.828 0.013 0.015
Chain 1: 3500 -18658.479 0.015 0.015
Chain 1: 3600 -17965.427 0.017 0.015
Chain 1: 3700 -18352.103 0.019 0.017
Chain 1: 3800 -17312.287 0.023 0.021
Chain 1: 3900 -17308.453 0.021 0.021
Chain 1: 4000 -17425.728 0.022 0.021
Chain 1: 4100 -17339.588 0.022 0.021
Chain 1: 4200 -17155.855 0.022 0.021
Chain 1: 4300 -17294.196 0.021 0.021
Chain 1: 4400 -17251.107 0.019 0.011
Chain 1: 4500 -17153.671 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00131 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.1 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48436.886 1.000 1.000
Chain 1: 200 -14863.210 1.629 2.259
Chain 1: 300 -15572.379 1.101 1.000
Chain 1: 400 -11288.231 0.921 1.000
Chain 1: 500 -12118.972 0.750 0.380
Chain 1: 600 -14646.623 0.654 0.380
Chain 1: 700 -12001.302 0.592 0.220
Chain 1: 800 -14285.291 0.538 0.220
Chain 1: 900 -12430.590 0.495 0.173
Chain 1: 1000 -12849.041 0.449 0.173
Chain 1: 1100 -17372.430 0.375 0.173
Chain 1: 1200 -23556.728 0.175 0.173
Chain 1: 1300 -12644.979 0.257 0.220
Chain 1: 1400 -12672.805 0.219 0.173
Chain 1: 1500 -16110.339 0.234 0.213
Chain 1: 1600 -11720.161 0.254 0.220
Chain 1: 1700 -11496.881 0.234 0.213
Chain 1: 1800 -10016.026 0.233 0.213
Chain 1: 1900 -10437.526 0.222 0.213
Chain 1: 2000 -11537.533 0.228 0.213
Chain 1: 2100 -9423.482 0.224 0.213
Chain 1: 2200 -9281.983 0.200 0.148
Chain 1: 2300 -9169.058 0.115 0.095
Chain 1: 2400 -9358.267 0.116 0.095
Chain 1: 2500 -11686.838 0.115 0.095
Chain 1: 2600 -9028.731 0.107 0.095
Chain 1: 2700 -9743.068 0.112 0.095
Chain 1: 2800 -10901.585 0.108 0.095
Chain 1: 2900 -9121.566 0.124 0.106
Chain 1: 3000 -8774.949 0.118 0.106
Chain 1: 3100 -8999.927 0.098 0.073
Chain 1: 3200 -8635.371 0.101 0.073
Chain 1: 3300 -11293.069 0.123 0.106
Chain 1: 3400 -10725.236 0.126 0.106
Chain 1: 3500 -10087.670 0.113 0.073
Chain 1: 3600 -9106.817 0.094 0.073
Chain 1: 3700 -8649.263 0.092 0.063
Chain 1: 3800 -8575.207 0.082 0.053
Chain 1: 3900 -10349.646 0.080 0.053
Chain 1: 4000 -8950.265 0.092 0.063
Chain 1: 4100 -11225.337 0.109 0.108
Chain 1: 4200 -8567.852 0.136 0.156
Chain 1: 4300 -9458.508 0.122 0.108
Chain 1: 4400 -10014.702 0.122 0.108
Chain 1: 4500 -8694.946 0.131 0.152
Chain 1: 4600 -9976.171 0.133 0.152
Chain 1: 4700 -8493.540 0.145 0.156
Chain 1: 4800 -11720.649 0.172 0.171
Chain 1: 4900 -12515.400 0.161 0.156
Chain 1: 5000 -9397.394 0.179 0.175
Chain 1: 5100 -8659.368 0.167 0.152
Chain 1: 5200 -8878.002 0.138 0.128
Chain 1: 5300 -10666.123 0.146 0.152
Chain 1: 5400 -15410.262 0.171 0.168
Chain 1: 5500 -9784.649 0.213 0.175
Chain 1: 5600 -8924.522 0.210 0.175
Chain 1: 5700 -11822.506 0.217 0.245
Chain 1: 5800 -8827.091 0.224 0.245
Chain 1: 5900 -11090.014 0.238 0.245
Chain 1: 6000 -9258.508 0.224 0.204
Chain 1: 6100 -11828.407 0.238 0.217
Chain 1: 6200 -8089.112 0.281 0.245
Chain 1: 6300 -9995.336 0.284 0.245
Chain 1: 6400 -8810.166 0.266 0.217
Chain 1: 6500 -8562.307 0.212 0.204
Chain 1: 6600 -10195.406 0.218 0.204
Chain 1: 6700 -8233.651 0.217 0.204
Chain 1: 6800 -9286.326 0.195 0.198
Chain 1: 6900 -11813.996 0.196 0.198
Chain 1: 7000 -12265.085 0.180 0.191
Chain 1: 7100 -13888.137 0.170 0.160
Chain 1: 7200 -8329.576 0.190 0.160
Chain 1: 7300 -8139.559 0.173 0.135
Chain 1: 7400 -11610.720 0.190 0.160
Chain 1: 7500 -11133.369 0.191 0.160
Chain 1: 7600 -8641.238 0.204 0.214
Chain 1: 7700 -8945.093 0.184 0.117
Chain 1: 7800 -8595.687 0.176 0.117
Chain 1: 7900 -8037.136 0.162 0.069
Chain 1: 8000 -11647.374 0.189 0.117
Chain 1: 8100 -8019.903 0.223 0.288
Chain 1: 8200 -8647.644 0.163 0.073
Chain 1: 8300 -8103.298 0.168 0.073
Chain 1: 8400 -10795.225 0.163 0.073
Chain 1: 8500 -11334.517 0.163 0.073
Chain 1: 8600 -8151.386 0.173 0.073
Chain 1: 8700 -8013.656 0.172 0.073
Chain 1: 8800 -8884.822 0.177 0.098
Chain 1: 8900 -8270.742 0.178 0.098
Chain 1: 9000 -10013.086 0.164 0.098
Chain 1: 9100 -8149.792 0.142 0.098
Chain 1: 9200 -9526.437 0.149 0.145
Chain 1: 9300 -9486.906 0.143 0.145
Chain 1: 9400 -9335.417 0.120 0.098
Chain 1: 9500 -8119.455 0.130 0.145
Chain 1: 9600 -8318.339 0.093 0.098
Chain 1: 9700 -10476.176 0.112 0.145
Chain 1: 9800 -10311.638 0.104 0.145
Chain 1: 9900 -8213.888 0.122 0.150
Chain 1: 10000 -8217.968 0.105 0.145
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00143 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.3 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -45802.495 1.000 1.000
Chain 1: 200 -15274.252 1.499 1.999
Chain 1: 300 -8550.256 1.262 1.000
Chain 1: 400 -8530.689 0.947 1.000
Chain 1: 500 -8262.794 0.764 0.786
Chain 1: 600 -8691.551 0.645 0.786
Chain 1: 700 -7927.399 0.567 0.096
Chain 1: 800 -8010.427 0.497 0.096
Chain 1: 900 -8149.328 0.444 0.049
Chain 1: 1000 -7700.432 0.405 0.058
Chain 1: 1100 -7664.804 0.306 0.049
Chain 1: 1200 -7519.674 0.108 0.032
Chain 1: 1300 -7666.937 0.031 0.019
Chain 1: 1400 -7739.592 0.032 0.019
Chain 1: 1500 -7553.072 0.031 0.019
Chain 1: 1600 -7709.074 0.028 0.019
Chain 1: 1700 -7439.733 0.022 0.019
Chain 1: 1800 -7539.512 0.022 0.019
Chain 1: 1900 -7561.697 0.021 0.019
Chain 1: 2000 -7534.037 0.015 0.019
Chain 1: 2100 -7585.556 0.016 0.019
Chain 1: 2200 -7610.925 0.014 0.013
Chain 1: 2300 -7516.028 0.013 0.013
Chain 1: 2400 -7570.544 0.013 0.013
Chain 1: 2500 -7410.140 0.013 0.013
Chain 1: 2600 -7477.560 0.012 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003092 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.92 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86150.487 1.000 1.000
Chain 1: 200 -13255.340 3.250 5.499
Chain 1: 300 -9639.204 2.291 1.000
Chain 1: 400 -10336.804 1.735 1.000
Chain 1: 500 -8602.553 1.429 0.375
Chain 1: 600 -8139.824 1.200 0.375
Chain 1: 700 -8440.410 1.034 0.202
Chain 1: 800 -9019.965 0.913 0.202
Chain 1: 900 -8423.043 0.819 0.071
Chain 1: 1000 -8191.671 0.740 0.071
Chain 1: 1100 -8324.270 0.642 0.067 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8122.250 0.094 0.064
Chain 1: 1300 -8327.367 0.059 0.057
Chain 1: 1400 -8329.607 0.052 0.036
Chain 1: 1500 -8224.150 0.033 0.028
Chain 1: 1600 -8327.123 0.029 0.025
Chain 1: 1700 -8415.231 0.026 0.025
Chain 1: 1800 -8007.288 0.025 0.025
Chain 1: 1900 -8103.928 0.019 0.016
Chain 1: 2000 -8076.090 0.017 0.013
Chain 1: 2100 -8196.898 0.017 0.013
Chain 1: 2200 -8014.387 0.016 0.013
Chain 1: 2300 -8143.502 0.016 0.013
Chain 1: 2400 -8153.590 0.016 0.013
Chain 1: 2500 -8115.811 0.015 0.012
Chain 1: 2600 -8114.559 0.014 0.012
Chain 1: 2700 -8029.459 0.014 0.012
Chain 1: 2800 -7994.279 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003419 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.19 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8386132.733 1.000 1.000
Chain 1: 200 -1583558.541 2.648 4.296
Chain 1: 300 -891319.954 2.024 1.000
Chain 1: 400 -457121.137 1.756 1.000
Chain 1: 500 -357589.937 1.460 0.950
Chain 1: 600 -232717.563 1.306 0.950
Chain 1: 700 -119048.403 1.256 0.950
Chain 1: 800 -86195.492 1.147 0.950
Chain 1: 900 -66555.366 1.052 0.777
Chain 1: 1000 -51344.860 0.976 0.777
Chain 1: 1100 -38809.915 0.909 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37986.496 0.481 0.381
Chain 1: 1300 -25940.866 0.450 0.381
Chain 1: 1400 -25657.995 0.356 0.323
Chain 1: 1500 -22243.348 0.344 0.323
Chain 1: 1600 -21458.543 0.294 0.296
Chain 1: 1700 -20333.005 0.204 0.295
Chain 1: 1800 -20277.070 0.166 0.154
Chain 1: 1900 -20603.055 0.138 0.055
Chain 1: 2000 -19114.329 0.116 0.055
Chain 1: 2100 -19352.944 0.085 0.037
Chain 1: 2200 -19579.023 0.084 0.037
Chain 1: 2300 -19196.574 0.040 0.020
Chain 1: 2400 -18968.755 0.040 0.020
Chain 1: 2500 -18770.469 0.025 0.016
Chain 1: 2600 -18401.110 0.024 0.016
Chain 1: 2700 -18358.212 0.019 0.012
Chain 1: 2800 -18075.002 0.020 0.016
Chain 1: 2900 -18356.180 0.020 0.015
Chain 1: 3000 -18342.498 0.012 0.012
Chain 1: 3100 -18427.418 0.011 0.012
Chain 1: 3200 -18118.259 0.012 0.015
Chain 1: 3300 -18322.861 0.011 0.012
Chain 1: 3400 -17797.934 0.013 0.015
Chain 1: 3500 -18409.490 0.015 0.016
Chain 1: 3600 -17716.660 0.017 0.016
Chain 1: 3700 -18103.089 0.019 0.017
Chain 1: 3800 -17063.416 0.023 0.021
Chain 1: 3900 -17059.560 0.022 0.021
Chain 1: 4000 -17176.901 0.022 0.021
Chain 1: 4100 -17090.623 0.022 0.021
Chain 1: 4200 -16907.070 0.022 0.021
Chain 1: 4300 -17045.388 0.022 0.021
Chain 1: 4400 -17002.363 0.019 0.011
Chain 1: 4500 -16904.875 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001162 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 11.62 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12237.843 1.000 1.000
Chain 1: 200 -9078.553 0.674 1.000
Chain 1: 300 -7913.618 0.498 0.348
Chain 1: 400 -8108.293 0.380 0.348
Chain 1: 500 -8145.983 0.305 0.147
Chain 1: 600 -7987.452 0.257 0.147
Chain 1: 700 -7794.156 0.224 0.025
Chain 1: 800 -7799.036 0.196 0.025
Chain 1: 900 -7759.497 0.175 0.024
Chain 1: 1000 -7806.603 0.158 0.024
Chain 1: 1100 -8026.693 0.061 0.024
Chain 1: 1200 -7801.055 0.029 0.024
Chain 1: 1300 -7742.890 0.015 0.020
Chain 1: 1400 -7772.350 0.013 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.0015 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56746.849 1.000 1.000
Chain 1: 200 -17329.290 1.637 2.275
Chain 1: 300 -8641.133 1.427 1.005
Chain 1: 400 -8421.846 1.077 1.005
Chain 1: 500 -8773.519 0.869 1.000
Chain 1: 600 -8475.977 0.730 1.000
Chain 1: 700 -7753.566 0.639 0.093
Chain 1: 800 -7974.884 0.563 0.093
Chain 1: 900 -7940.080 0.501 0.040
Chain 1: 1000 -8099.155 0.453 0.040
Chain 1: 1100 -7590.837 0.359 0.040
Chain 1: 1200 -7593.480 0.132 0.035
Chain 1: 1300 -7640.884 0.032 0.028
Chain 1: 1400 -7796.773 0.031 0.028
Chain 1: 1500 -7620.846 0.030 0.023
Chain 1: 1600 -7772.445 0.028 0.020
Chain 1: 1700 -7452.780 0.023 0.020
Chain 1: 1800 -7564.420 0.022 0.020
Chain 1: 1900 -7624.789 0.022 0.020
Chain 1: 2000 -7621.867 0.020 0.020
Chain 1: 2100 -7579.556 0.014 0.015
Chain 1: 2200 -7689.942 0.015 0.015
Chain 1: 2300 -7590.724 0.016 0.015
Chain 1: 2400 -7629.995 0.015 0.014
Chain 1: 2500 -7569.145 0.013 0.013
Chain 1: 2600 -7509.210 0.012 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002743 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 27.43 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85337.842 1.000 1.000
Chain 1: 200 -13388.912 3.187 5.374
Chain 1: 300 -9765.424 2.248 1.000
Chain 1: 400 -10624.877 1.706 1.000
Chain 1: 500 -8581.885 1.413 0.371
Chain 1: 600 -8207.557 1.185 0.371
Chain 1: 700 -8672.942 1.023 0.238
Chain 1: 800 -8798.975 0.897 0.238
Chain 1: 900 -8599.054 0.800 0.081
Chain 1: 1000 -8234.768 0.724 0.081
Chain 1: 1100 -8575.329 0.628 0.054 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8319.170 0.094 0.046
Chain 1: 1300 -8449.431 0.059 0.044
Chain 1: 1400 -8487.585 0.051 0.040
Chain 1: 1500 -8320.387 0.029 0.031
Chain 1: 1600 -8439.831 0.026 0.023
Chain 1: 1700 -8520.633 0.022 0.020
Chain 1: 1800 -8107.898 0.025 0.023
Chain 1: 1900 -8203.682 0.024 0.020
Chain 1: 2000 -8176.962 0.020 0.015
Chain 1: 2100 -8299.652 0.018 0.015
Chain 1: 2200 -8119.658 0.017 0.015
Chain 1: 2300 -8198.784 0.016 0.014
Chain 1: 2400 -8268.426 0.016 0.014
Chain 1: 2500 -8213.790 0.015 0.012
Chain 1: 2600 -8213.178 0.014 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003532 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.32 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8364924.017 1.000 1.000
Chain 1: 200 -1579127.500 2.649 4.297
Chain 1: 300 -890821.386 2.023 1.000
Chain 1: 400 -457995.555 1.754 1.000
Chain 1: 500 -358722.965 1.458 0.945
Chain 1: 600 -233702.959 1.304 0.945
Chain 1: 700 -119540.515 1.255 0.945
Chain 1: 800 -86655.462 1.145 0.945
Chain 1: 900 -66917.838 1.051 0.773
Chain 1: 1000 -51650.709 0.975 0.773
Chain 1: 1100 -39065.166 0.907 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38236.081 0.480 0.379
Chain 1: 1300 -26126.329 0.449 0.379
Chain 1: 1400 -25839.886 0.356 0.322
Chain 1: 1500 -22409.979 0.343 0.322
Chain 1: 1600 -21621.492 0.293 0.296
Chain 1: 1700 -20487.095 0.203 0.295
Chain 1: 1800 -20429.519 0.166 0.153
Chain 1: 1900 -20755.610 0.138 0.055
Chain 1: 2000 -19262.766 0.116 0.055
Chain 1: 2100 -19501.286 0.085 0.036
Chain 1: 2200 -19728.407 0.084 0.036
Chain 1: 2300 -19345.066 0.040 0.020
Chain 1: 2400 -19117.070 0.040 0.020
Chain 1: 2500 -18919.445 0.025 0.016
Chain 1: 2600 -18549.285 0.024 0.016
Chain 1: 2700 -18506.224 0.018 0.012
Chain 1: 2800 -18223.184 0.020 0.016
Chain 1: 2900 -18504.593 0.020 0.015
Chain 1: 3000 -18490.682 0.012 0.012
Chain 1: 3100 -18575.637 0.011 0.012
Chain 1: 3200 -18266.294 0.012 0.015
Chain 1: 3300 -18471.093 0.011 0.012
Chain 1: 3400 -17946.027 0.013 0.015
Chain 1: 3500 -18557.971 0.015 0.016
Chain 1: 3600 -17864.683 0.017 0.016
Chain 1: 3700 -18251.468 0.019 0.017
Chain 1: 3800 -17211.243 0.023 0.021
Chain 1: 3900 -17207.468 0.022 0.021
Chain 1: 4000 -17324.714 0.022 0.021
Chain 1: 4100 -17238.434 0.022 0.021
Chain 1: 4200 -17054.762 0.022 0.021
Chain 1: 4300 -17193.064 0.021 0.021
Chain 1: 4400 -17149.890 0.019 0.011
Chain 1: 4500 -17052.493 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001437 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.37 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12683.616 1.000 1.000
Chain 1: 200 -9572.323 0.663 1.000
Chain 1: 300 -8179.246 0.498 0.325
Chain 1: 400 -8285.553 0.377 0.325
Chain 1: 500 -8254.799 0.302 0.170
Chain 1: 600 -8101.957 0.255 0.170
Chain 1: 700 -8083.011 0.219 0.019
Chain 1: 800 -8036.873 0.192 0.019
Chain 1: 900 -8002.409 0.171 0.013
Chain 1: 1000 -8078.959 0.155 0.013
Chain 1: 1100 -8068.219 0.055 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001433 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.33 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -46220.201 1.000 1.000
Chain 1: 200 -15715.730 1.471 1.941
Chain 1: 300 -8820.046 1.241 1.000
Chain 1: 400 -8774.663 0.932 1.000
Chain 1: 500 -8369.218 0.755 0.782
Chain 1: 600 -8733.356 0.636 0.782
Chain 1: 700 -7970.711 0.559 0.096
Chain 1: 800 -7738.264 0.493 0.096
Chain 1: 900 -8093.185 0.443 0.048
Chain 1: 1000 -7964.431 0.400 0.048
Chain 1: 1100 -7954.385 0.301 0.044
Chain 1: 1200 -7842.567 0.108 0.042
Chain 1: 1300 -7846.000 0.030 0.030
Chain 1: 1400 -7893.521 0.030 0.030
Chain 1: 1500 -7677.396 0.028 0.028
Chain 1: 1600 -7740.078 0.024 0.016
Chain 1: 1700 -7629.492 0.016 0.014
Chain 1: 1800 -7683.633 0.014 0.014
Chain 1: 1900 -7703.374 0.010 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003219 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.19 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86324.364 1.000 1.000
Chain 1: 200 -13660.662 3.160 5.319
Chain 1: 300 -10013.734 2.228 1.000
Chain 1: 400 -10722.439 1.687 1.000
Chain 1: 500 -8995.719 1.388 0.364
Chain 1: 600 -8434.855 1.168 0.364
Chain 1: 700 -8570.489 1.003 0.192
Chain 1: 800 -9487.864 0.890 0.192
Chain 1: 900 -8813.488 0.800 0.097
Chain 1: 1000 -8571.907 0.723 0.097
Chain 1: 1100 -8838.239 0.626 0.077 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8471.489 0.098 0.066
Chain 1: 1300 -8596.015 0.063 0.066
Chain 1: 1400 -8682.222 0.057 0.043
Chain 1: 1500 -8572.029 0.039 0.030
Chain 1: 1600 -8676.473 0.034 0.028
Chain 1: 1700 -8764.978 0.033 0.028
Chain 1: 1800 -8347.611 0.029 0.028
Chain 1: 1900 -8445.618 0.022 0.014
Chain 1: 2000 -8419.277 0.020 0.013
Chain 1: 2100 -8542.844 0.018 0.013
Chain 1: 2200 -8359.657 0.016 0.013
Chain 1: 2300 -8440.122 0.016 0.012
Chain 1: 2400 -8509.769 0.015 0.012
Chain 1: 2500 -8455.606 0.015 0.012
Chain 1: 2600 -8455.656 0.014 0.010
Chain 1: 2700 -8372.880 0.014 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003249 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.49 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8376683.139 1.000 1.000
Chain 1: 200 -1577882.727 2.654 4.309
Chain 1: 300 -890838.029 2.027 1.000
Chain 1: 400 -457919.859 1.756 1.000
Chain 1: 500 -358917.851 1.460 0.945
Chain 1: 600 -233796.907 1.306 0.945
Chain 1: 700 -119775.303 1.255 0.945
Chain 1: 800 -86888.751 1.146 0.945
Chain 1: 900 -67158.690 1.051 0.771
Chain 1: 1000 -51895.907 0.975 0.771
Chain 1: 1100 -39317.716 0.907 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38486.452 0.479 0.378
Chain 1: 1300 -26388.244 0.447 0.378
Chain 1: 1400 -26100.682 0.354 0.320
Chain 1: 1500 -22673.751 0.342 0.320
Chain 1: 1600 -21885.903 0.292 0.294
Chain 1: 1700 -20753.408 0.202 0.294
Chain 1: 1800 -20696.050 0.164 0.151
Chain 1: 1900 -21022.177 0.137 0.055
Chain 1: 2000 -19529.996 0.115 0.055
Chain 1: 2100 -19768.520 0.084 0.036
Chain 1: 2200 -19995.495 0.083 0.036
Chain 1: 2300 -19612.244 0.039 0.020
Chain 1: 2400 -19384.283 0.039 0.020
Chain 1: 2500 -19186.448 0.025 0.016
Chain 1: 2600 -18816.443 0.023 0.016
Chain 1: 2700 -18773.378 0.018 0.012
Chain 1: 2800 -18490.321 0.019 0.015
Chain 1: 2900 -18771.631 0.019 0.015
Chain 1: 3000 -18757.739 0.012 0.012
Chain 1: 3100 -18842.762 0.011 0.012
Chain 1: 3200 -18533.389 0.012 0.015
Chain 1: 3300 -18738.160 0.011 0.012
Chain 1: 3400 -18213.062 0.012 0.015
Chain 1: 3500 -18825.037 0.015 0.015
Chain 1: 3600 -18131.618 0.017 0.015
Chain 1: 3700 -18518.552 0.018 0.017
Chain 1: 3800 -17478.126 0.023 0.021
Chain 1: 3900 -17474.316 0.021 0.021
Chain 1: 4000 -17591.570 0.022 0.021
Chain 1: 4100 -17505.328 0.022 0.021
Chain 1: 4200 -17321.574 0.021 0.021
Chain 1: 4300 -17459.969 0.021 0.021
Chain 1: 4400 -17416.791 0.018 0.011
Chain 1: 4500 -17319.325 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001583 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.83 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12662.972 1.000 1.000
Chain 1: 200 -9590.488 0.660 1.000
Chain 1: 300 -8297.858 0.492 0.320
Chain 1: 400 -8437.352 0.373 0.320
Chain 1: 500 -8368.360 0.300 0.156
Chain 1: 600 -8211.338 0.253 0.156
Chain 1: 700 -8140.330 0.218 0.019
Chain 1: 800 -8100.580 0.192 0.019
Chain 1: 900 -8062.559 0.171 0.017
Chain 1: 1000 -8236.482 0.156 0.019
Chain 1: 1100 -8245.938 0.056 0.017
Chain 1: 1200 -8127.823 0.025 0.015
Chain 1: 1300 -8099.874 0.010 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001464 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.64 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58305.827 1.000 1.000
Chain 1: 200 -18027.627 1.617 2.234
Chain 1: 300 -8812.521 1.427 1.046
Chain 1: 400 -8154.098 1.090 1.046
Chain 1: 500 -8820.528 0.887 1.000
Chain 1: 600 -8742.392 0.741 1.000
Chain 1: 700 -7986.552 0.649 0.095
Chain 1: 800 -8078.658 0.569 0.095
Chain 1: 900 -7827.879 0.509 0.081
Chain 1: 1000 -7720.550 0.460 0.081
Chain 1: 1100 -7858.093 0.361 0.076
Chain 1: 1200 -7575.816 0.142 0.037
Chain 1: 1300 -7555.209 0.037 0.032
Chain 1: 1400 -8140.282 0.037 0.032
Chain 1: 1500 -7606.374 0.036 0.032
Chain 1: 1600 -7781.172 0.037 0.032
Chain 1: 1700 -7515.966 0.031 0.032
Chain 1: 1800 -7654.508 0.032 0.032
Chain 1: 1900 -7673.978 0.029 0.022
Chain 1: 2000 -7686.417 0.028 0.022
Chain 1: 2100 -7506.511 0.029 0.024
Chain 1: 2200 -7797.490 0.029 0.024
Chain 1: 2300 -7528.368 0.032 0.035
Chain 1: 2400 -7550.587 0.025 0.024
Chain 1: 2500 -7584.989 0.018 0.022
Chain 1: 2600 -7541.961 0.017 0.018
Chain 1: 2700 -7542.453 0.013 0.006 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002772 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 27.72 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86795.293 1.000 1.000
Chain 1: 200 -13796.141 3.146 5.291
Chain 1: 300 -10139.803 2.217 1.000
Chain 1: 400 -11069.637 1.684 1.000
Chain 1: 500 -9121.242 1.390 0.361
Chain 1: 600 -9036.753 1.160 0.361
Chain 1: 700 -8698.468 1.000 0.214
Chain 1: 800 -8894.249 0.877 0.214
Chain 1: 900 -8915.848 0.780 0.084
Chain 1: 1000 -8590.778 0.706 0.084
Chain 1: 1100 -8934.963 0.610 0.039 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8550.782 0.085 0.039
Chain 1: 1300 -8833.841 0.052 0.039
Chain 1: 1400 -8812.582 0.044 0.038
Chain 1: 1500 -8670.054 0.024 0.032
Chain 1: 1600 -8791.939 0.025 0.032
Chain 1: 1700 -8869.827 0.022 0.022
Chain 1: 1800 -8445.375 0.025 0.032
Chain 1: 1900 -8546.588 0.026 0.032
Chain 1: 2000 -8521.141 0.022 0.016
Chain 1: 2100 -8646.675 0.020 0.015
Chain 1: 2200 -8449.608 0.018 0.015
Chain 1: 2300 -8541.400 0.016 0.014
Chain 1: 2400 -8610.193 0.016 0.014
Chain 1: 2500 -8556.446 0.015 0.012
Chain 1: 2600 -8557.816 0.014 0.011
Chain 1: 2700 -8474.526 0.014 0.011
Chain 1: 2800 -8434.403 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003164 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.64 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8416980.466 1.000 1.000
Chain 1: 200 -1586550.662 2.653 4.305
Chain 1: 300 -890821.338 2.029 1.000
Chain 1: 400 -458098.882 1.758 1.000
Chain 1: 500 -358416.395 1.462 0.945
Chain 1: 600 -233410.549 1.307 0.945
Chain 1: 700 -119569.530 1.257 0.945
Chain 1: 800 -86783.552 1.147 0.945
Chain 1: 900 -67114.468 1.052 0.781
Chain 1: 1000 -51902.169 0.976 0.781
Chain 1: 1100 -39373.938 0.908 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38548.675 0.479 0.378
Chain 1: 1300 -26497.864 0.447 0.378
Chain 1: 1400 -26216.489 0.353 0.318
Chain 1: 1500 -22802.795 0.341 0.318
Chain 1: 1600 -22018.996 0.291 0.293
Chain 1: 1700 -20891.721 0.201 0.293
Chain 1: 1800 -20835.620 0.163 0.150
Chain 1: 1900 -21161.931 0.136 0.054
Chain 1: 2000 -19672.391 0.114 0.054
Chain 1: 2100 -19910.751 0.083 0.036
Chain 1: 2200 -20137.509 0.082 0.036
Chain 1: 2300 -19754.402 0.039 0.019
Chain 1: 2400 -19526.426 0.039 0.019
Chain 1: 2500 -19328.544 0.025 0.015
Chain 1: 2600 -18958.583 0.023 0.015
Chain 1: 2700 -18915.402 0.018 0.012
Chain 1: 2800 -18632.313 0.019 0.015
Chain 1: 2900 -18913.610 0.019 0.015
Chain 1: 3000 -18899.730 0.012 0.012
Chain 1: 3100 -18984.783 0.011 0.012
Chain 1: 3200 -18675.348 0.011 0.015
Chain 1: 3300 -18880.126 0.011 0.012
Chain 1: 3400 -18354.937 0.012 0.015
Chain 1: 3500 -18967.028 0.015 0.015
Chain 1: 3600 -18273.364 0.016 0.015
Chain 1: 3700 -18660.492 0.018 0.017
Chain 1: 3800 -17619.698 0.023 0.021
Chain 1: 3900 -17615.813 0.021 0.021
Chain 1: 4000 -17733.119 0.022 0.021
Chain 1: 4100 -17646.919 0.022 0.021
Chain 1: 4200 -17462.968 0.021 0.021
Chain 1: 4300 -17601.476 0.021 0.021
Chain 1: 4400 -17558.219 0.018 0.011
Chain 1: 4500 -17460.721 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001252 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.52 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49741.638 1.000 1.000
Chain 1: 200 -17928.826 1.387 1.774
Chain 1: 300 -18079.906 0.928 1.000
Chain 1: 400 -19489.795 0.714 1.000
Chain 1: 500 -12768.953 0.676 0.526
Chain 1: 600 -23172.326 0.638 0.526
Chain 1: 700 -17637.001 0.592 0.449
Chain 1: 800 -16225.412 0.529 0.449
Chain 1: 900 -14374.187 0.484 0.314
Chain 1: 1000 -12176.863 0.454 0.314
Chain 1: 1100 -23550.884 0.402 0.314
Chain 1: 1200 -14660.946 0.286 0.314
Chain 1: 1300 -15090.611 0.288 0.314
Chain 1: 1400 -14264.539 0.286 0.314
Chain 1: 1500 -12097.617 0.251 0.180
Chain 1: 1600 -12203.270 0.207 0.179
Chain 1: 1700 -10437.400 0.193 0.169
Chain 1: 1800 -10596.330 0.186 0.169
Chain 1: 1900 -11263.727 0.179 0.169
Chain 1: 2000 -11303.014 0.161 0.059
Chain 1: 2100 -11380.977 0.113 0.058
Chain 1: 2200 -10672.790 0.059 0.058
Chain 1: 2300 -10277.831 0.060 0.058
Chain 1: 2400 -10318.839 0.055 0.038
Chain 1: 2500 -17377.489 0.078 0.038
Chain 1: 2600 -10396.965 0.144 0.059
Chain 1: 2700 -11417.476 0.136 0.059
Chain 1: 2800 -10656.450 0.142 0.066
Chain 1: 2900 -10373.308 0.138 0.066
Chain 1: 3000 -12334.322 0.154 0.071
Chain 1: 3100 -9436.030 0.184 0.089
Chain 1: 3200 -9795.979 0.181 0.089
Chain 1: 3300 -9676.389 0.178 0.089
Chain 1: 3400 -10388.635 0.185 0.089
Chain 1: 3500 -10518.427 0.146 0.071
Chain 1: 3600 -10701.775 0.080 0.069
Chain 1: 3700 -9715.202 0.081 0.069
Chain 1: 3800 -12995.824 0.099 0.069
Chain 1: 3900 -10695.391 0.118 0.102
Chain 1: 4000 -15398.142 0.133 0.102
Chain 1: 4100 -9655.908 0.162 0.102
Chain 1: 4200 -9566.727 0.159 0.102
Chain 1: 4300 -10626.845 0.168 0.102
Chain 1: 4400 -9977.087 0.167 0.102
Chain 1: 4500 -14752.739 0.198 0.215
Chain 1: 4600 -9577.309 0.251 0.252
Chain 1: 4700 -12235.467 0.262 0.252
Chain 1: 4800 -10611.844 0.252 0.217
Chain 1: 4900 -9659.899 0.241 0.217
Chain 1: 5000 -13204.273 0.237 0.217
Chain 1: 5100 -9444.318 0.217 0.217
Chain 1: 5200 -14132.946 0.250 0.268
Chain 1: 5300 -14694.715 0.243 0.268
Chain 1: 5400 -10661.800 0.275 0.324
Chain 1: 5500 -9595.363 0.254 0.268
Chain 1: 5600 -15894.384 0.239 0.268
Chain 1: 5700 -9318.900 0.288 0.332
Chain 1: 5800 -9138.651 0.275 0.332
Chain 1: 5900 -9443.145 0.268 0.332
Chain 1: 6000 -10123.082 0.248 0.332
Chain 1: 6100 -14203.744 0.237 0.287
Chain 1: 6200 -9097.774 0.260 0.287
Chain 1: 6300 -9316.860 0.258 0.287
Chain 1: 6400 -9877.853 0.226 0.111
Chain 1: 6500 -9555.529 0.218 0.067
Chain 1: 6600 -12316.368 0.201 0.067
Chain 1: 6700 -9054.174 0.167 0.067
Chain 1: 6800 -9132.199 0.165 0.067
Chain 1: 6900 -13597.885 0.195 0.224
Chain 1: 7000 -9142.824 0.237 0.287
Chain 1: 7100 -9159.685 0.209 0.224
Chain 1: 7200 -10068.972 0.161 0.090
Chain 1: 7300 -12465.404 0.178 0.192
Chain 1: 7400 -8998.435 0.211 0.224
Chain 1: 7500 -12371.424 0.235 0.273
Chain 1: 7600 -9257.256 0.246 0.328
Chain 1: 7700 -12269.997 0.235 0.273
Chain 1: 7800 -8830.579 0.273 0.328
Chain 1: 7900 -8827.118 0.240 0.273
Chain 1: 8000 -8727.376 0.193 0.246
Chain 1: 8100 -9046.700 0.196 0.246
Chain 1: 8200 -11038.083 0.205 0.246
Chain 1: 8300 -9127.129 0.207 0.246
Chain 1: 8400 -9466.066 0.172 0.209
Chain 1: 8500 -8811.049 0.152 0.180
Chain 1: 8600 -10831.938 0.137 0.180
Chain 1: 8700 -12679.561 0.127 0.146
Chain 1: 8800 -11230.523 0.101 0.129
Chain 1: 8900 -10797.642 0.105 0.129
Chain 1: 9000 -11129.684 0.107 0.129
Chain 1: 9100 -8671.822 0.131 0.146
Chain 1: 9200 -8974.794 0.117 0.129
Chain 1: 9300 -9614.416 0.103 0.074
Chain 1: 9400 -8642.945 0.110 0.112
Chain 1: 9500 -11716.363 0.129 0.129
Chain 1: 9600 -8925.671 0.142 0.129
Chain 1: 9700 -8606.381 0.131 0.112
Chain 1: 9800 -8855.535 0.121 0.067
Chain 1: 9900 -8801.721 0.117 0.067
Chain 1: 10000 -9029.027 0.117 0.067
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001524 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.24 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57905.350 1.000 1.000
Chain 1: 200 -18312.228 1.581 2.162
Chain 1: 300 -9177.376 1.386 1.000
Chain 1: 400 -8257.775 1.067 1.000
Chain 1: 500 -8584.510 0.861 0.995
Chain 1: 600 -8513.471 0.719 0.995
Chain 1: 700 -7595.899 0.634 0.121
Chain 1: 800 -8532.629 0.568 0.121
Chain 1: 900 -7854.496 0.515 0.111
Chain 1: 1000 -8177.636 0.467 0.111
Chain 1: 1100 -7769.025 0.372 0.110
Chain 1: 1200 -7686.467 0.157 0.086
Chain 1: 1300 -7791.333 0.059 0.053
Chain 1: 1400 -7728.704 0.049 0.040
Chain 1: 1500 -7683.257 0.046 0.040
Chain 1: 1600 -7743.305 0.046 0.040
Chain 1: 1700 -7723.815 0.034 0.013
Chain 1: 1800 -7701.172 0.023 0.011
Chain 1: 1900 -7616.390 0.015 0.011
Chain 1: 2000 -7835.668 0.014 0.011
Chain 1: 2100 -7679.959 0.011 0.011
Chain 1: 2200 -7877.154 0.013 0.011
Chain 1: 2300 -7657.596 0.014 0.011
Chain 1: 2400 -7658.611 0.013 0.011
Chain 1: 2500 -7670.190 0.013 0.011
Chain 1: 2600 -7593.081 0.013 0.011
Chain 1: 2700 -7594.169 0.013 0.011
Chain 1: 2800 -7559.485 0.013 0.011
Chain 1: 2900 -7431.326 0.014 0.017
Chain 1: 3000 -7614.282 0.013 0.017
Chain 1: 3100 -7587.112 0.012 0.010
Chain 1: 3200 -7791.942 0.012 0.010
Chain 1: 3300 -7471.877 0.013 0.010
Chain 1: 3400 -7730.065 0.016 0.017
Chain 1: 3500 -7522.586 0.019 0.024
Chain 1: 3600 -7551.624 0.018 0.024
Chain 1: 3700 -7522.756 0.019 0.024
Chain 1: 3800 -7478.854 0.019 0.024
Chain 1: 3900 -7462.461 0.017 0.024
Chain 1: 4000 -7452.298 0.015 0.006 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002537 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.37 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86661.743 1.000 1.000
Chain 1: 200 -14345.846 3.020 5.041
Chain 1: 300 -10565.200 2.133 1.000
Chain 1: 400 -12372.766 1.636 1.000
Chain 1: 500 -9113.489 1.380 0.358
Chain 1: 600 -8866.235 1.155 0.358
Chain 1: 700 -9495.021 1.000 0.358
Chain 1: 800 -10182.420 0.883 0.358
Chain 1: 900 -9362.822 0.795 0.146
Chain 1: 1000 -9356.324 0.715 0.146
Chain 1: 1100 -9294.388 0.616 0.088 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8838.043 0.117 0.068
Chain 1: 1300 -9088.947 0.084 0.066
Chain 1: 1400 -9112.460 0.070 0.052
Chain 1: 1500 -9023.505 0.035 0.028
Chain 1: 1600 -9100.330 0.033 0.028
Chain 1: 1700 -9182.287 0.027 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00267 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 26.7 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8387580.984 1.000 1.000
Chain 1: 200 -1585580.683 2.645 4.290
Chain 1: 300 -892933.562 2.022 1.000
Chain 1: 400 -459501.457 1.752 1.000
Chain 1: 500 -359811.033 1.457 0.943
Chain 1: 600 -234514.966 1.303 0.943
Chain 1: 700 -120407.983 1.253 0.943
Chain 1: 800 -87565.436 1.143 0.943
Chain 1: 900 -67852.441 1.048 0.776
Chain 1: 1000 -52616.756 0.972 0.776
Chain 1: 1100 -40054.154 0.904 0.534 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39233.587 0.477 0.375
Chain 1: 1300 -27130.177 0.444 0.375
Chain 1: 1400 -26848.098 0.351 0.314
Chain 1: 1500 -23419.388 0.337 0.314
Chain 1: 1600 -22632.727 0.288 0.291
Chain 1: 1700 -21498.080 0.198 0.290
Chain 1: 1800 -21440.928 0.161 0.146
Chain 1: 1900 -21767.878 0.133 0.053
Chain 1: 2000 -20273.333 0.112 0.053
Chain 1: 2100 -20511.993 0.081 0.035
Chain 1: 2200 -20739.784 0.080 0.035
Chain 1: 2300 -20355.603 0.038 0.019
Chain 1: 2400 -20127.327 0.038 0.019
Chain 1: 2500 -19929.697 0.024 0.015
Chain 1: 2600 -19558.737 0.023 0.015
Chain 1: 2700 -19515.378 0.018 0.012
Chain 1: 2800 -19232.017 0.019 0.015
Chain 1: 2900 -19513.727 0.019 0.014
Chain 1: 3000 -19499.783 0.011 0.012
Chain 1: 3100 -19584.917 0.011 0.011
Chain 1: 3200 -19274.987 0.011 0.014
Chain 1: 3300 -19480.211 0.010 0.011
Chain 1: 3400 -18954.163 0.012 0.014
Chain 1: 3500 -19567.607 0.014 0.015
Chain 1: 3600 -18872.289 0.016 0.015
Chain 1: 3700 -19260.606 0.018 0.016
Chain 1: 3800 -18217.288 0.022 0.020
Chain 1: 3900 -18213.412 0.021 0.020
Chain 1: 4000 -18330.668 0.021 0.020
Chain 1: 4100 -18244.297 0.021 0.020
Chain 1: 4200 -18059.890 0.021 0.020
Chain 1: 4300 -18198.703 0.020 0.020
Chain 1: 4400 -18154.975 0.018 0.010
Chain 1: 4500 -18057.464 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001441 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.41 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12753.761 1.000 1.000
Chain 1: 200 -9544.981 0.668 1.000
Chain 1: 300 -8336.727 0.494 0.336
Chain 1: 400 -8504.995 0.375 0.336
Chain 1: 500 -8402.178 0.303 0.145
Chain 1: 600 -8235.530 0.256 0.145
Chain 1: 700 -8117.690 0.221 0.020
Chain 1: 800 -8120.379 0.194 0.020
Chain 1: 900 -8106.206 0.172 0.020
Chain 1: 1000 -8193.510 0.156 0.020
Chain 1: 1100 -8171.634 0.056 0.015
Chain 1: 1200 -8145.701 0.023 0.012
Chain 1: 1300 -8093.671 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001412 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.12 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57158.592 1.000 1.000
Chain 1: 200 -17886.087 1.598 2.196
Chain 1: 300 -8958.918 1.397 1.000
Chain 1: 400 -8257.970 1.069 1.000
Chain 1: 500 -8513.357 0.861 0.996
Chain 1: 600 -8812.525 0.723 0.996
Chain 1: 700 -8694.113 0.622 0.085
Chain 1: 800 -8238.310 0.551 0.085
Chain 1: 900 -8196.552 0.491 0.055
Chain 1: 1000 -7883.487 0.445 0.055
Chain 1: 1100 -7917.855 0.346 0.040
Chain 1: 1200 -8045.227 0.128 0.034
Chain 1: 1300 -7964.806 0.029 0.030
Chain 1: 1400 -8096.676 0.022 0.016
Chain 1: 1500 -7670.801 0.025 0.016
Chain 1: 1600 -7823.102 0.024 0.016
Chain 1: 1700 -7677.516 0.024 0.019
Chain 1: 1800 -7763.052 0.020 0.016
Chain 1: 1900 -7680.279 0.020 0.016
Chain 1: 2000 -7720.346 0.017 0.016
Chain 1: 2100 -7673.061 0.017 0.016
Chain 1: 2200 -7820.650 0.017 0.016
Chain 1: 2300 -7677.207 0.018 0.019
Chain 1: 2400 -7705.441 0.017 0.019
Chain 1: 2500 -7720.364 0.011 0.011
Chain 1: 2600 -7617.413 0.011 0.011
Chain 1: 2700 -7618.536 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002744 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 27.44 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86324.192 1.000 1.000
Chain 1: 200 -13944.483 3.095 5.191
Chain 1: 300 -10222.018 2.185 1.000
Chain 1: 400 -11581.554 1.668 1.000
Chain 1: 500 -9232.575 1.385 0.364
Chain 1: 600 -8651.862 1.166 0.364
Chain 1: 700 -8911.992 1.003 0.254
Chain 1: 800 -9411.061 0.884 0.254
Chain 1: 900 -8945.443 0.792 0.117
Chain 1: 1000 -8932.546 0.713 0.117
Chain 1: 1100 -8960.621 0.613 0.067 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8595.578 0.098 0.053
Chain 1: 1300 -8869.114 0.065 0.052
Chain 1: 1400 -8854.953 0.054 0.042
Chain 1: 1500 -8725.526 0.030 0.031
Chain 1: 1600 -8840.288 0.024 0.029
Chain 1: 1700 -8899.887 0.022 0.015
Chain 1: 1800 -8462.157 0.022 0.015
Chain 1: 1900 -8566.074 0.018 0.013
Chain 1: 2000 -8544.410 0.018 0.013
Chain 1: 2100 -8520.913 0.018 0.013
Chain 1: 2200 -8484.611 0.014 0.012
Chain 1: 2300 -8619.362 0.013 0.012
Chain 1: 2400 -8464.256 0.014 0.013
Chain 1: 2500 -8535.467 0.014 0.012
Chain 1: 2600 -8448.291 0.013 0.010
Chain 1: 2700 -8485.571 0.013 0.010
Chain 1: 2800 -8443.639 0.008 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003046 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.46 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8386207.188 1.000 1.000
Chain 1: 200 -1581962.442 2.651 4.301
Chain 1: 300 -891571.767 2.025 1.000
Chain 1: 400 -458630.713 1.755 1.000
Chain 1: 500 -359333.178 1.459 0.944
Chain 1: 600 -234103.229 1.305 0.944
Chain 1: 700 -120004.463 1.255 0.944
Chain 1: 800 -87158.580 1.145 0.944
Chain 1: 900 -67436.479 1.050 0.774
Chain 1: 1000 -52188.570 0.974 0.774
Chain 1: 1100 -39622.544 0.906 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38796.903 0.478 0.377
Chain 1: 1300 -26698.461 0.446 0.377
Chain 1: 1400 -26414.710 0.353 0.317
Chain 1: 1500 -22988.171 0.340 0.317
Chain 1: 1600 -22201.408 0.290 0.292
Chain 1: 1700 -21067.932 0.200 0.292
Chain 1: 1800 -21010.797 0.163 0.149
Chain 1: 1900 -21337.375 0.135 0.054
Chain 1: 2000 -19844.144 0.113 0.054
Chain 1: 2100 -20082.662 0.083 0.035
Chain 1: 2200 -20310.141 0.082 0.035
Chain 1: 2300 -19926.342 0.038 0.019
Chain 1: 2400 -19698.192 0.039 0.019
Chain 1: 2500 -19500.528 0.025 0.015
Chain 1: 2600 -19129.921 0.023 0.015
Chain 1: 2700 -19086.672 0.018 0.012
Chain 1: 2800 -18803.469 0.019 0.015
Chain 1: 2900 -19084.997 0.019 0.015
Chain 1: 3000 -19071.054 0.012 0.012
Chain 1: 3100 -19156.140 0.011 0.012
Chain 1: 3200 -18846.451 0.011 0.015
Chain 1: 3300 -19051.474 0.011 0.012
Chain 1: 3400 -18525.846 0.012 0.015
Chain 1: 3500 -19138.671 0.014 0.015
Chain 1: 3600 -18444.129 0.016 0.015
Chain 1: 3700 -18831.867 0.018 0.016
Chain 1: 3800 -17789.774 0.022 0.021
Chain 1: 3900 -17785.920 0.021 0.021
Chain 1: 4000 -17903.169 0.022 0.021
Chain 1: 4100 -17816.875 0.022 0.021
Chain 1: 4200 -17632.714 0.021 0.021
Chain 1: 4300 -17771.348 0.021 0.021
Chain 1: 4400 -17727.834 0.018 0.010
Chain 1: 4500 -17630.353 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001302 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.02 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12728.997 1.000 1.000
Chain 1: 200 -9520.715 0.668 1.000
Chain 1: 300 -8306.566 0.494 0.337
Chain 1: 400 -8477.185 0.376 0.337
Chain 1: 500 -8430.023 0.302 0.146
Chain 1: 600 -8236.783 0.255 0.146
Chain 1: 700 -8165.143 0.220 0.023
Chain 1: 800 -8191.580 0.193 0.023
Chain 1: 900 -8232.867 0.172 0.020
Chain 1: 1000 -8173.206 0.156 0.020
Chain 1: 1100 -8256.189 0.057 0.010
Chain 1: 1200 -8179.220 0.024 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002458 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 24.58 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57266.176 1.000 1.000
Chain 1: 200 -17771.874 1.611 2.222
Chain 1: 300 -8941.523 1.403 1.000
Chain 1: 400 -8234.334 1.074 1.000
Chain 1: 500 -8796.165 0.872 0.988
Chain 1: 600 -8835.979 0.727 0.988
Chain 1: 700 -8522.502 0.629 0.086
Chain 1: 800 -8234.319 0.554 0.086
Chain 1: 900 -8059.075 0.495 0.064
Chain 1: 1000 -8029.081 0.446 0.064
Chain 1: 1100 -7737.327 0.350 0.038
Chain 1: 1200 -7766.414 0.128 0.037
Chain 1: 1300 -7866.695 0.031 0.035
Chain 1: 1400 -7902.717 0.022 0.022
Chain 1: 1500 -7623.696 0.020 0.022
Chain 1: 1600 -7772.250 0.021 0.022
Chain 1: 1700 -7760.968 0.018 0.019
Chain 1: 1800 -7710.491 0.015 0.013
Chain 1: 1900 -7644.151 0.013 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003251 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.51 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86545.678 1.000 1.000
Chain 1: 200 -13855.060 3.123 5.247
Chain 1: 300 -10187.311 2.202 1.000
Chain 1: 400 -11272.499 1.676 1.000
Chain 1: 500 -9171.739 1.386 0.360
Chain 1: 600 -8583.436 1.167 0.360
Chain 1: 700 -9147.437 1.009 0.229
Chain 1: 800 -9495.240 0.887 0.229
Chain 1: 900 -8991.708 0.795 0.096
Chain 1: 1000 -8952.181 0.716 0.096
Chain 1: 1100 -8809.186 0.618 0.069 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8576.871 0.096 0.062
Chain 1: 1300 -8862.426 0.063 0.056
Chain 1: 1400 -8818.888 0.054 0.037
Chain 1: 1500 -8708.193 0.032 0.032
Chain 1: 1600 -8815.663 0.026 0.027
Chain 1: 1700 -8892.185 0.021 0.016
Chain 1: 1800 -8461.865 0.023 0.016
Chain 1: 1900 -8565.651 0.018 0.013
Chain 1: 2000 -8540.893 0.018 0.013
Chain 1: 2100 -8675.175 0.018 0.013
Chain 1: 2200 -8469.704 0.018 0.013
Chain 1: 2300 -8565.191 0.016 0.012
Chain 1: 2400 -8629.459 0.016 0.012
Chain 1: 2500 -8574.289 0.015 0.012
Chain 1: 2600 -8578.378 0.014 0.011
Chain 1: 2700 -8493.671 0.014 0.011
Chain 1: 2800 -8450.517 0.010 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002728 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 27.28 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8413954.481 1.000 1.000
Chain 1: 200 -1588537.453 2.648 4.297
Chain 1: 300 -892484.432 2.026 1.000
Chain 1: 400 -458572.932 1.756 1.000
Chain 1: 500 -358395.410 1.460 0.946
Chain 1: 600 -233227.333 1.306 0.946
Chain 1: 700 -119490.249 1.256 0.946
Chain 1: 800 -86715.944 1.146 0.946
Chain 1: 900 -67082.197 1.051 0.780
Chain 1: 1000 -51903.397 0.975 0.780
Chain 1: 1100 -39401.869 0.907 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38583.374 0.480 0.378
Chain 1: 1300 -26557.537 0.447 0.378
Chain 1: 1400 -26279.899 0.353 0.317
Chain 1: 1500 -22870.943 0.340 0.317
Chain 1: 1600 -22089.185 0.290 0.293
Chain 1: 1700 -20964.703 0.200 0.292
Chain 1: 1800 -20909.518 0.163 0.149
Chain 1: 1900 -21235.817 0.135 0.054
Chain 1: 2000 -19747.509 0.113 0.054
Chain 1: 2100 -19985.966 0.083 0.035
Chain 1: 2200 -20212.353 0.082 0.035
Chain 1: 2300 -19829.538 0.038 0.019
Chain 1: 2400 -19601.551 0.039 0.019
Chain 1: 2500 -19403.470 0.025 0.015
Chain 1: 2600 -19033.566 0.023 0.015
Chain 1: 2700 -18990.521 0.018 0.012
Chain 1: 2800 -18707.212 0.019 0.015
Chain 1: 2900 -18988.516 0.019 0.015
Chain 1: 3000 -18974.772 0.012 0.012
Chain 1: 3100 -19059.772 0.011 0.012
Chain 1: 3200 -18750.345 0.011 0.015
Chain 1: 3300 -18955.155 0.011 0.012
Chain 1: 3400 -18429.823 0.012 0.015
Chain 1: 3500 -19042.065 0.014 0.015
Chain 1: 3600 -18348.254 0.016 0.015
Chain 1: 3700 -18735.351 0.018 0.017
Chain 1: 3800 -17694.342 0.023 0.021
Chain 1: 3900 -17690.443 0.021 0.021
Chain 1: 4000 -17807.761 0.022 0.021
Chain 1: 4100 -17721.467 0.022 0.021
Chain 1: 4200 -17537.567 0.021 0.021
Chain 1: 4300 -17676.089 0.021 0.021
Chain 1: 4400 -17632.770 0.018 0.010
Chain 1: 4500 -17535.262 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001377 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.77 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49982.824 1.000 1.000
Chain 1: 200 -17059.946 1.465 1.930
Chain 1: 300 -18682.299 1.006 1.000
Chain 1: 400 -15248.982 0.810 1.000
Chain 1: 500 -19816.998 0.694 0.231
Chain 1: 600 -18426.129 0.591 0.231
Chain 1: 700 -13019.188 0.566 0.231
Chain 1: 800 -12161.041 0.504 0.231
Chain 1: 900 -12520.314 0.451 0.225
Chain 1: 1000 -24926.445 0.456 0.231
Chain 1: 1100 -21666.174 0.371 0.225
Chain 1: 1200 -12550.726 0.251 0.225
Chain 1: 1300 -13285.921 0.248 0.225
Chain 1: 1400 -11863.612 0.237 0.150
Chain 1: 1500 -11398.166 0.218 0.120
Chain 1: 1600 -10738.629 0.217 0.120
Chain 1: 1700 -13104.637 0.193 0.120
Chain 1: 1800 -13534.047 0.189 0.120
Chain 1: 1900 -11916.603 0.200 0.136
Chain 1: 2000 -18321.048 0.185 0.136
Chain 1: 2100 -11363.116 0.231 0.136
Chain 1: 2200 -12390.733 0.167 0.120
Chain 1: 2300 -10891.075 0.175 0.136
Chain 1: 2400 -9726.643 0.175 0.136
Chain 1: 2500 -10496.293 0.178 0.136
Chain 1: 2600 -10481.592 0.172 0.136
Chain 1: 2700 -11329.699 0.162 0.120
Chain 1: 2800 -11371.591 0.159 0.120
Chain 1: 2900 -11581.322 0.147 0.083
Chain 1: 3000 -9681.257 0.132 0.083
Chain 1: 3100 -10460.241 0.078 0.075
Chain 1: 3200 -10359.320 0.071 0.074
Chain 1: 3300 -16808.799 0.096 0.074
Chain 1: 3400 -9999.277 0.152 0.074
Chain 1: 3500 -11875.492 0.160 0.075
Chain 1: 3600 -16697.447 0.189 0.158
Chain 1: 3700 -17077.722 0.184 0.158
Chain 1: 3800 -10406.736 0.247 0.196
Chain 1: 3900 -12169.726 0.260 0.196
Chain 1: 4000 -10503.800 0.256 0.159
Chain 1: 4100 -10395.595 0.250 0.159
Chain 1: 4200 -10944.770 0.254 0.159
Chain 1: 4300 -10795.209 0.217 0.158
Chain 1: 4400 -10416.764 0.152 0.145
Chain 1: 4500 -10010.907 0.141 0.050
Chain 1: 4600 -10334.786 0.115 0.041
Chain 1: 4700 -9354.552 0.123 0.050
Chain 1: 4800 -9749.477 0.063 0.041
Chain 1: 4900 -9837.104 0.050 0.041
Chain 1: 5000 -9889.956 0.034 0.036
Chain 1: 5100 -9745.952 0.035 0.036
Chain 1: 5200 -11485.724 0.045 0.036
Chain 1: 5300 -10326.056 0.055 0.041
Chain 1: 5400 -9249.516 0.063 0.041
Chain 1: 5500 -9531.704 0.062 0.041
Chain 1: 5600 -9543.747 0.059 0.041
Chain 1: 5700 -9936.120 0.052 0.039
Chain 1: 5800 -9628.546 0.051 0.032
Chain 1: 5900 -13073.939 0.077 0.039
Chain 1: 6000 -9475.241 0.114 0.112
Chain 1: 6100 -15871.571 0.153 0.116
Chain 1: 6200 -15021.289 0.143 0.112
Chain 1: 6300 -9424.883 0.192 0.116
Chain 1: 6400 -9216.113 0.182 0.057
Chain 1: 6500 -13776.569 0.212 0.264
Chain 1: 6600 -10014.717 0.250 0.331
Chain 1: 6700 -14037.753 0.274 0.331
Chain 1: 6800 -9972.569 0.312 0.376
Chain 1: 6900 -9452.653 0.291 0.376
Chain 1: 7000 -9622.268 0.255 0.331
Chain 1: 7100 -11063.102 0.228 0.287
Chain 1: 7200 -9099.091 0.244 0.287
Chain 1: 7300 -11157.749 0.203 0.216
Chain 1: 7400 -8709.690 0.229 0.281
Chain 1: 7500 -11568.533 0.220 0.247
Chain 1: 7600 -10402.290 0.194 0.216
Chain 1: 7700 -9017.148 0.180 0.185
Chain 1: 7800 -13306.856 0.172 0.185
Chain 1: 7900 -12572.605 0.172 0.185
Chain 1: 8000 -8944.234 0.211 0.216
Chain 1: 8100 -14091.382 0.235 0.247
Chain 1: 8200 -9149.984 0.267 0.281
Chain 1: 8300 -9179.025 0.249 0.281
Chain 1: 8400 -11475.220 0.241 0.247
Chain 1: 8500 -9179.696 0.241 0.250
Chain 1: 8600 -9662.233 0.235 0.250
Chain 1: 8700 -9140.832 0.225 0.250
Chain 1: 8800 -8694.490 0.198 0.200
Chain 1: 8900 -9548.189 0.201 0.200
Chain 1: 9000 -11220.851 0.176 0.149
Chain 1: 9100 -9083.103 0.163 0.149
Chain 1: 9200 -8837.210 0.111 0.089
Chain 1: 9300 -8988.000 0.113 0.089
Chain 1: 9400 -8812.198 0.095 0.057
Chain 1: 9500 -11612.543 0.094 0.057
Chain 1: 9600 -8999.774 0.118 0.089
Chain 1: 9700 -8832.973 0.114 0.089
Chain 1: 9800 -11096.466 0.129 0.149
Chain 1: 9900 -8851.293 0.146 0.204
Chain 1: 10000 -9420.791 0.137 0.204
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001374 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.74 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58016.562 1.000 1.000
Chain 1: 200 -18628.230 1.557 2.114
Chain 1: 300 -9406.656 1.365 1.000
Chain 1: 400 -8384.394 1.054 1.000
Chain 1: 500 -8991.632 0.857 0.980
Chain 1: 600 -9734.021 0.727 0.980
Chain 1: 700 -8400.043 0.646 0.159
Chain 1: 800 -8609.479 0.568 0.159
Chain 1: 900 -8164.271 0.511 0.122
Chain 1: 1000 -7904.025 0.463 0.122
Chain 1: 1100 -7885.008 0.363 0.076
Chain 1: 1200 -8054.476 0.154 0.068
Chain 1: 1300 -7944.886 0.057 0.055
Chain 1: 1400 -7988.098 0.046 0.033
Chain 1: 1500 -7859.833 0.041 0.024
Chain 1: 1600 -7948.250 0.034 0.021
Chain 1: 1700 -7701.656 0.021 0.021
Chain 1: 1800 -7744.838 0.020 0.016
Chain 1: 1900 -7770.887 0.014 0.014
Chain 1: 2000 -7923.511 0.013 0.014
Chain 1: 2100 -7803.875 0.014 0.015
Chain 1: 2200 -8214.422 0.017 0.015
Chain 1: 2300 -7781.853 0.021 0.016
Chain 1: 2400 -7751.678 0.021 0.016
Chain 1: 2500 -7883.314 0.021 0.017
Chain 1: 2600 -7716.508 0.022 0.019
Chain 1: 2700 -7715.666 0.019 0.017
Chain 1: 2800 -7789.398 0.020 0.017
Chain 1: 2900 -7540.067 0.023 0.019
Chain 1: 3000 -7711.850 0.023 0.022
Chain 1: 3100 -7688.763 0.022 0.022
Chain 1: 3200 -7975.184 0.020 0.022
Chain 1: 3300 -7590.709 0.020 0.022
Chain 1: 3400 -7808.737 0.022 0.022
Chain 1: 3500 -7601.570 0.023 0.027
Chain 1: 3600 -7653.922 0.022 0.027
Chain 1: 3700 -7561.041 0.023 0.027
Chain 1: 3800 -7581.463 0.022 0.027
Chain 1: 3900 -7586.574 0.019 0.022
Chain 1: 4000 -7552.893 0.017 0.012
Chain 1: 4100 -7565.421 0.017 0.012
Chain 1: 4200 -7746.515 0.016 0.012
Chain 1: 4300 -7548.114 0.013 0.012
Chain 1: 4400 -7603.243 0.011 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004983 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 49.83 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87813.602 1.000 1.000
Chain 1: 200 -14648.132 2.997 4.995
Chain 1: 300 -10733.436 2.120 1.000
Chain 1: 400 -13126.940 1.635 1.000
Chain 1: 500 -9140.799 1.396 0.436
Chain 1: 600 -9733.241 1.173 0.436
Chain 1: 700 -8798.665 1.021 0.365
Chain 1: 800 -9302.453 0.900 0.365
Chain 1: 900 -9343.927 0.800 0.182
Chain 1: 1000 -8950.395 0.725 0.182
Chain 1: 1100 -9219.545 0.628 0.106 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -9013.189 0.130 0.061
Chain 1: 1300 -9215.826 0.096 0.054
Chain 1: 1400 -9307.817 0.079 0.044
Chain 1: 1500 -9103.391 0.038 0.029
Chain 1: 1600 -9176.502 0.032 0.023
Chain 1: 1700 -9247.104 0.022 0.022
Chain 1: 1800 -8755.739 0.023 0.022
Chain 1: 1900 -8877.936 0.024 0.022
Chain 1: 2000 -8883.444 0.019 0.022
Chain 1: 2100 -9030.369 0.018 0.016
Chain 1: 2200 -8745.017 0.019 0.016
Chain 1: 2300 -8831.070 0.018 0.014
Chain 1: 2400 -8925.036 0.018 0.014
Chain 1: 2500 -8821.880 0.017 0.012
Chain 1: 2600 -8872.875 0.016 0.012
Chain 1: 2700 -8779.209 0.017 0.012
Chain 1: 2800 -8747.628 0.012 0.011
Chain 1: 2900 -8833.373 0.011 0.011
Chain 1: 3000 -8764.332 0.012 0.011
Chain 1: 3100 -8719.189 0.011 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003065 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.65 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8392860.474 1.000 1.000
Chain 1: 200 -1580637.054 2.655 4.310
Chain 1: 300 -892280.589 2.027 1.000
Chain 1: 400 -459820.811 1.755 1.000
Chain 1: 500 -360518.445 1.459 0.940
Chain 1: 600 -235246.922 1.305 0.940
Chain 1: 700 -120989.361 1.253 0.940
Chain 1: 800 -88079.728 1.143 0.940
Chain 1: 900 -68318.530 1.049 0.771
Chain 1: 1000 -53044.686 0.972 0.771
Chain 1: 1100 -40445.877 0.904 0.533 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39623.889 0.475 0.374
Chain 1: 1300 -27480.066 0.442 0.374
Chain 1: 1400 -27195.351 0.349 0.311
Chain 1: 1500 -23755.958 0.336 0.311
Chain 1: 1600 -22966.773 0.286 0.289
Chain 1: 1700 -21827.219 0.197 0.288
Chain 1: 1800 -21769.103 0.160 0.145
Chain 1: 1900 -22096.617 0.132 0.052
Chain 1: 2000 -20598.306 0.111 0.052
Chain 1: 2100 -20837.248 0.081 0.034
Chain 1: 2200 -21065.770 0.080 0.034
Chain 1: 2300 -20680.766 0.037 0.019
Chain 1: 2400 -20452.259 0.037 0.019
Chain 1: 2500 -20254.643 0.024 0.015
Chain 1: 2600 -19883.044 0.022 0.015
Chain 1: 2700 -19839.445 0.017 0.011
Chain 1: 2800 -19555.896 0.018 0.014
Chain 1: 2900 -19837.929 0.018 0.014
Chain 1: 3000 -19823.836 0.011 0.011
Chain 1: 3100 -19909.114 0.010 0.011
Chain 1: 3200 -19598.750 0.011 0.014
Chain 1: 3300 -19804.291 0.010 0.011
Chain 1: 3400 -19277.538 0.012 0.014
Chain 1: 3500 -19892.059 0.014 0.014
Chain 1: 3600 -19195.290 0.016 0.014
Chain 1: 3700 -19584.800 0.017 0.016
Chain 1: 3800 -18539.196 0.022 0.020
Chain 1: 3900 -18535.253 0.020 0.020
Chain 1: 4000 -18652.507 0.021 0.020
Chain 1: 4100 -18566.056 0.021 0.020
Chain 1: 4200 -18381.105 0.020 0.020
Chain 1: 4300 -18520.300 0.020 0.020
Chain 1: 4400 -18476.201 0.017 0.010
Chain 1: 4500 -18378.570 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001277 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.77 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13987.024 1.000 1.000
Chain 1: 200 -10616.719 0.659 1.000
Chain 1: 300 -9336.128 0.485 0.317
Chain 1: 400 -8957.274 0.374 0.317
Chain 1: 500 -8895.709 0.301 0.137
Chain 1: 600 -8581.290 0.257 0.137
Chain 1: 700 -8733.415 0.223 0.042
Chain 1: 800 -8587.949 0.197 0.042
Chain 1: 900 -8588.526 0.175 0.037
Chain 1: 1000 -8604.701 0.158 0.037
Chain 1: 1100 -8603.827 0.058 0.017
Chain 1: 1200 -8526.086 0.027 0.017
Chain 1: 1300 -8472.292 0.014 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001366 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.66 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -64702.744 1.000 1.000
Chain 1: 200 -19391.424 1.668 2.337
Chain 1: 300 -9397.792 1.467 1.063
Chain 1: 400 -9104.566 1.108 1.063
Chain 1: 500 -7784.053 0.920 1.000
Chain 1: 600 -8495.419 0.781 1.000
Chain 1: 700 -8786.701 0.674 0.170
Chain 1: 800 -8647.185 0.592 0.170
Chain 1: 900 -8641.555 0.526 0.084
Chain 1: 1000 -7838.124 0.484 0.103
Chain 1: 1100 -7864.251 0.384 0.084
Chain 1: 1200 -7922.534 0.151 0.033
Chain 1: 1300 -7913.342 0.045 0.032
Chain 1: 1400 -7985.499 0.043 0.016
Chain 1: 1500 -7605.628 0.031 0.016
Chain 1: 1600 -7626.964 0.023 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003325 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.25 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87066.713 1.000 1.000
Chain 1: 200 -14574.216 2.987 4.974
Chain 1: 300 -10740.702 2.110 1.000
Chain 1: 400 -12842.443 1.624 1.000
Chain 1: 500 -9326.567 1.374 0.377
Chain 1: 600 -9546.849 1.149 0.377
Chain 1: 700 -8965.495 0.994 0.357
Chain 1: 800 -9496.943 0.877 0.357
Chain 1: 900 -9311.826 0.782 0.164
Chain 1: 1000 -9293.834 0.704 0.164
Chain 1: 1100 -9494.131 0.606 0.065 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -9093.745 0.113 0.056
Chain 1: 1300 -9318.758 0.080 0.044
Chain 1: 1400 -9261.754 0.064 0.024
Chain 1: 1500 -9189.066 0.027 0.023
Chain 1: 1600 -9264.484 0.025 0.021
Chain 1: 1700 -9326.225 0.020 0.020
Chain 1: 1800 -8877.643 0.019 0.020
Chain 1: 1900 -8973.738 0.018 0.011
Chain 1: 2000 -8995.294 0.018 0.011
Chain 1: 2100 -9090.521 0.017 0.010
Chain 1: 2200 -8854.182 0.015 0.010
Chain 1: 2300 -9027.659 0.015 0.010
Chain 1: 2400 -8882.590 0.016 0.011
Chain 1: 2500 -8944.579 0.016 0.011
Chain 1: 2600 -8851.750 0.016 0.011
Chain 1: 2700 -8887.032 0.016 0.011
Chain 1: 2800 -8843.491 0.011 0.010
Chain 1: 2900 -8953.523 0.011 0.010
Chain 1: 3000 -8861.625 0.012 0.010
Chain 1: 3100 -8829.166 0.011 0.010
Chain 1: 3200 -8798.560 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004109 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 41.09 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8422054.838 1.000 1.000
Chain 1: 200 -1587606.573 2.652 4.305
Chain 1: 300 -892274.504 2.028 1.000
Chain 1: 400 -459155.924 1.757 1.000
Chain 1: 500 -359109.594 1.461 0.943
Chain 1: 600 -233968.004 1.307 0.943
Chain 1: 700 -120282.777 1.255 0.943
Chain 1: 800 -87492.082 1.145 0.943
Chain 1: 900 -67857.274 1.050 0.779
Chain 1: 1000 -52687.083 0.974 0.779
Chain 1: 1100 -40180.436 0.905 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39369.372 0.477 0.375
Chain 1: 1300 -27327.510 0.443 0.375
Chain 1: 1400 -27051.457 0.349 0.311
Chain 1: 1500 -23638.093 0.336 0.311
Chain 1: 1600 -22856.095 0.286 0.289
Chain 1: 1700 -21728.894 0.197 0.288
Chain 1: 1800 -21673.654 0.159 0.144
Chain 1: 1900 -22000.765 0.132 0.052
Chain 1: 2000 -20509.685 0.110 0.052
Chain 1: 2100 -20748.195 0.080 0.034
Chain 1: 2200 -20975.397 0.079 0.034
Chain 1: 2300 -20591.748 0.037 0.019
Chain 1: 2400 -20363.488 0.037 0.019
Chain 1: 2500 -20165.435 0.024 0.015
Chain 1: 2600 -19794.617 0.022 0.015
Chain 1: 2700 -19751.339 0.017 0.011
Chain 1: 2800 -19467.645 0.019 0.015
Chain 1: 2900 -19749.428 0.018 0.014
Chain 1: 3000 -19735.565 0.011 0.011
Chain 1: 3100 -19820.684 0.011 0.011
Chain 1: 3200 -19510.711 0.011 0.014
Chain 1: 3300 -19715.982 0.010 0.011
Chain 1: 3400 -19189.644 0.012 0.014
Chain 1: 3500 -19803.333 0.014 0.015
Chain 1: 3600 -19107.724 0.016 0.015
Chain 1: 3700 -19496.185 0.017 0.016
Chain 1: 3800 -18452.260 0.022 0.020
Chain 1: 3900 -18448.323 0.020 0.020
Chain 1: 4000 -18565.658 0.021 0.020
Chain 1: 4100 -18479.177 0.021 0.020
Chain 1: 4200 -18294.680 0.020 0.020
Chain 1: 4300 -18433.621 0.020 0.020
Chain 1: 4400 -18389.797 0.018 0.010
Chain 1: 4500 -18292.222 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001309 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.09 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48937.176 1.000 1.000
Chain 1: 200 -22187.787 1.103 1.206
Chain 1: 300 -20731.101 0.759 1.000
Chain 1: 400 -17798.659 0.610 1.000
Chain 1: 500 -12010.672 0.585 0.482
Chain 1: 600 -14836.833 0.519 0.482
Chain 1: 700 -11907.443 0.480 0.246
Chain 1: 800 -14100.862 0.439 0.246
Chain 1: 900 -10693.998 0.426 0.246
Chain 1: 1000 -24475.873 0.440 0.319
Chain 1: 1100 -34583.788 0.369 0.292
Chain 1: 1200 -16706.656 0.355 0.292
Chain 1: 1300 -11706.221 0.391 0.319
Chain 1: 1400 -20646.587 0.418 0.427
Chain 1: 1500 -10684.555 0.463 0.427
Chain 1: 1600 -24272.703 0.500 0.433
Chain 1: 1700 -20231.490 0.495 0.433
Chain 1: 1800 -9372.817 0.595 0.560 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1900 -10993.122 0.578 0.560 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2000 -9692.677 0.535 0.433 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2100 -9533.923 0.508 0.433 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2200 -13080.231 0.428 0.427
Chain 1: 2300 -15805.510 0.403 0.271
Chain 1: 2400 -9618.786 0.424 0.271
Chain 1: 2500 -10252.837 0.336 0.200
Chain 1: 2600 -9006.173 0.294 0.172
Chain 1: 2700 -10197.591 0.286 0.147
Chain 1: 2800 -9371.774 0.179 0.138
Chain 1: 2900 -9766.726 0.168 0.134
Chain 1: 3000 -13431.372 0.182 0.138
Chain 1: 3100 -9745.691 0.218 0.172
Chain 1: 3200 -9053.920 0.199 0.138
Chain 1: 3300 -9290.994 0.184 0.117
Chain 1: 3400 -9606.439 0.123 0.088
Chain 1: 3500 -9356.305 0.120 0.088
Chain 1: 3600 -10165.554 0.114 0.080
Chain 1: 3700 -9664.237 0.107 0.076
Chain 1: 3800 -9457.092 0.101 0.052
Chain 1: 3900 -10028.635 0.102 0.057
Chain 1: 4000 -10132.801 0.076 0.052
Chain 1: 4100 -10228.318 0.039 0.033
Chain 1: 4200 -9756.980 0.036 0.033
Chain 1: 4300 -9586.028 0.036 0.033
Chain 1: 4400 -10489.220 0.041 0.048
Chain 1: 4500 -8508.106 0.062 0.052
Chain 1: 4600 -12533.304 0.086 0.052
Chain 1: 4700 -8642.606 0.125 0.057
Chain 1: 4800 -8607.694 0.124 0.057
Chain 1: 4900 -9227.992 0.125 0.067
Chain 1: 5000 -16305.140 0.167 0.086
Chain 1: 5100 -9409.324 0.239 0.233
Chain 1: 5200 -9800.856 0.239 0.233
Chain 1: 5300 -12909.470 0.261 0.241
Chain 1: 5400 -9334.474 0.291 0.321
Chain 1: 5500 -13035.569 0.296 0.321
Chain 1: 5600 -8518.686 0.317 0.383
Chain 1: 5700 -8516.226 0.272 0.284
Chain 1: 5800 -8461.414 0.272 0.284
Chain 1: 5900 -8782.635 0.269 0.284
Chain 1: 6000 -9658.482 0.234 0.241
Chain 1: 6100 -8860.321 0.170 0.091
Chain 1: 6200 -8398.221 0.172 0.091
Chain 1: 6300 -8567.443 0.150 0.090
Chain 1: 6400 -12375.363 0.142 0.090
Chain 1: 6500 -13958.311 0.125 0.090
Chain 1: 6600 -12723.608 0.082 0.090
Chain 1: 6700 -8424.729 0.133 0.091
Chain 1: 6800 -8893.638 0.137 0.091
Chain 1: 6900 -10055.295 0.145 0.097
Chain 1: 7000 -10008.769 0.137 0.097
Chain 1: 7100 -8420.046 0.146 0.113
Chain 1: 7200 -9116.730 0.149 0.113
Chain 1: 7300 -8471.948 0.154 0.113
Chain 1: 7400 -8248.025 0.126 0.097
Chain 1: 7500 -8379.629 0.116 0.076
Chain 1: 7600 -9518.417 0.119 0.076
Chain 1: 7700 -8419.324 0.081 0.076
Chain 1: 7800 -11957.607 0.105 0.116
Chain 1: 7900 -9627.696 0.118 0.120
Chain 1: 8000 -8242.388 0.134 0.131
Chain 1: 8100 -8195.018 0.116 0.120
Chain 1: 8200 -9654.052 0.123 0.131
Chain 1: 8300 -9435.402 0.118 0.131
Chain 1: 8400 -8123.138 0.131 0.151
Chain 1: 8500 -8347.771 0.132 0.151
Chain 1: 8600 -10895.658 0.144 0.162
Chain 1: 8700 -10573.505 0.134 0.162
Chain 1: 8800 -8667.424 0.126 0.162
Chain 1: 8900 -9928.222 0.115 0.151
Chain 1: 9000 -9667.045 0.101 0.127
Chain 1: 9100 -8252.984 0.117 0.151
Chain 1: 9200 -8852.387 0.109 0.127
Chain 1: 9300 -8279.273 0.113 0.127
Chain 1: 9400 -8391.779 0.099 0.069
Chain 1: 9500 -8019.423 0.101 0.069
Chain 1: 9600 -9117.831 0.089 0.069
Chain 1: 9700 -10119.892 0.096 0.099
Chain 1: 9800 -10196.700 0.075 0.069
Chain 1: 9900 -8330.833 0.085 0.069
Chain 1: 10000 -8994.553 0.089 0.074
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001449 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.49 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -46502.211 1.000 1.000
Chain 1: 200 -15603.140 1.490 1.980
Chain 1: 300 -8705.216 1.258 1.000
Chain 1: 400 -8680.130 0.944 1.000
Chain 1: 500 -8239.879 0.766 0.792
Chain 1: 600 -8064.426 0.642 0.792
Chain 1: 700 -8434.341 0.556 0.053
Chain 1: 800 -8603.844 0.489 0.053
Chain 1: 900 -7958.278 0.444 0.053
Chain 1: 1000 -7822.166 0.401 0.053
Chain 1: 1100 -7737.336 0.302 0.044
Chain 1: 1200 -7606.241 0.106 0.022
Chain 1: 1300 -7750.177 0.029 0.020
Chain 1: 1400 -7913.793 0.030 0.021
Chain 1: 1500 -7611.110 0.029 0.021
Chain 1: 1600 -7717.180 0.028 0.020
Chain 1: 1700 -7544.734 0.026 0.020
Chain 1: 1800 -7607.568 0.025 0.019
Chain 1: 1900 -7608.884 0.017 0.017
Chain 1: 2000 -7719.501 0.017 0.017
Chain 1: 2100 -7661.145 0.016 0.017
Chain 1: 2200 -7711.886 0.015 0.014
Chain 1: 2300 -7623.988 0.015 0.014
Chain 1: 2400 -7675.973 0.013 0.012
Chain 1: 2500 -7554.202 0.011 0.012
Chain 1: 2600 -7541.583 0.010 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00362 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 36.2 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86341.264 1.000 1.000
Chain 1: 200 -13484.259 3.202 5.403
Chain 1: 300 -9874.040 2.256 1.000
Chain 1: 400 -10645.973 1.710 1.000
Chain 1: 500 -8846.818 1.409 0.366
Chain 1: 600 -8366.390 1.184 0.366
Chain 1: 700 -8454.479 1.016 0.203
Chain 1: 800 -9217.073 0.899 0.203
Chain 1: 900 -8684.691 0.806 0.083
Chain 1: 1000 -8502.447 0.728 0.083
Chain 1: 1100 -8701.952 0.630 0.073 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8348.192 0.094 0.061
Chain 1: 1300 -8567.024 0.060 0.057
Chain 1: 1400 -8569.985 0.053 0.042
Chain 1: 1500 -8458.894 0.034 0.026
Chain 1: 1600 -8563.106 0.029 0.023
Chain 1: 1700 -8651.611 0.029 0.023
Chain 1: 1800 -8244.026 0.026 0.023
Chain 1: 1900 -8340.876 0.021 0.021
Chain 1: 2000 -8312.994 0.019 0.013
Chain 1: 2100 -8433.527 0.018 0.013
Chain 1: 2200 -8245.034 0.016 0.013
Chain 1: 2300 -8380.754 0.015 0.013
Chain 1: 2400 -8388.269 0.015 0.013
Chain 1: 2500 -8354.073 0.015 0.012
Chain 1: 2600 -8352.050 0.013 0.012
Chain 1: 2700 -8266.225 0.013 0.012
Chain 1: 2800 -8231.359 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003379 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.79 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8401792.667 1.000 1.000
Chain 1: 200 -1586635.364 2.648 4.295
Chain 1: 300 -891640.756 2.025 1.000
Chain 1: 400 -458090.787 1.755 1.000
Chain 1: 500 -358148.195 1.460 0.946
Chain 1: 600 -233060.447 1.306 0.946
Chain 1: 700 -119230.664 1.256 0.946
Chain 1: 800 -86416.385 1.146 0.946
Chain 1: 900 -66755.729 1.052 0.779
Chain 1: 1000 -51550.681 0.976 0.779
Chain 1: 1100 -39027.108 0.908 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38200.920 0.481 0.380
Chain 1: 1300 -26165.338 0.449 0.380
Chain 1: 1400 -25883.746 0.355 0.321
Chain 1: 1500 -22473.276 0.343 0.321
Chain 1: 1600 -21689.873 0.293 0.295
Chain 1: 1700 -20565.155 0.203 0.295
Chain 1: 1800 -20509.441 0.165 0.152
Chain 1: 1900 -20835.403 0.137 0.055
Chain 1: 2000 -19347.405 0.115 0.055
Chain 1: 2100 -19585.836 0.084 0.036
Chain 1: 2200 -19812.021 0.083 0.036
Chain 1: 2300 -19429.496 0.039 0.020
Chain 1: 2400 -19201.641 0.039 0.020
Chain 1: 2500 -19003.541 0.025 0.016
Chain 1: 2600 -18634.064 0.024 0.016
Chain 1: 2700 -18591.059 0.018 0.012
Chain 1: 2800 -18307.966 0.020 0.015
Chain 1: 2900 -18589.114 0.020 0.015
Chain 1: 3000 -18575.347 0.012 0.012
Chain 1: 3100 -18660.320 0.011 0.012
Chain 1: 3200 -18351.131 0.012 0.015
Chain 1: 3300 -18555.721 0.011 0.012
Chain 1: 3400 -18030.860 0.013 0.015
Chain 1: 3500 -18642.408 0.015 0.015
Chain 1: 3600 -17949.481 0.017 0.015
Chain 1: 3700 -18335.995 0.019 0.017
Chain 1: 3800 -17296.314 0.023 0.021
Chain 1: 3900 -17292.443 0.022 0.021
Chain 1: 4000 -17409.774 0.022 0.021
Chain 1: 4100 -17323.581 0.022 0.021
Chain 1: 4200 -17139.911 0.022 0.021
Chain 1: 4300 -17278.256 0.021 0.021
Chain 1: 4400 -17235.197 0.019 0.011
Chain 1: 4500 -17137.719 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001296 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.96 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12838.615 1.000 1.000
Chain 1: 200 -9563.478 0.671 1.000
Chain 1: 300 -8443.430 0.492 0.342
Chain 1: 400 -8571.408 0.373 0.342
Chain 1: 500 -8520.021 0.299 0.133
Chain 1: 600 -8343.339 0.253 0.133
Chain 1: 700 -8258.918 0.218 0.021
Chain 1: 800 -8270.616 0.191 0.021
Chain 1: 900 -8198.751 0.171 0.015
Chain 1: 1000 -8380.388 0.156 0.021
Chain 1: 1100 -8400.076 0.056 0.015
Chain 1: 1200 -8275.238 0.023 0.015
Chain 1: 1300 -8244.853 0.011 0.010
Chain 1: 1400 -8255.171 0.009 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002827 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 28.27 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57596.368 1.000 1.000
Chain 1: 200 -17881.655 1.610 2.221
Chain 1: 300 -8943.654 1.407 1.000
Chain 1: 400 -8226.766 1.077 1.000
Chain 1: 500 -9119.290 0.881 0.999
Chain 1: 600 -8547.122 0.745 0.999
Chain 1: 700 -8929.022 0.645 0.098
Chain 1: 800 -8189.379 0.576 0.098
Chain 1: 900 -8013.616 0.514 0.090
Chain 1: 1000 -7865.628 0.465 0.090
Chain 1: 1100 -7687.406 0.367 0.087
Chain 1: 1200 -7970.428 0.148 0.067
Chain 1: 1300 -7856.344 0.050 0.043
Chain 1: 1400 -7908.964 0.042 0.036
Chain 1: 1500 -7614.808 0.036 0.036
Chain 1: 1600 -7761.725 0.031 0.023
Chain 1: 1700 -7730.703 0.027 0.022
Chain 1: 1800 -7752.814 0.019 0.019
Chain 1: 1900 -7620.025 0.018 0.019
Chain 1: 2000 -7735.955 0.018 0.017
Chain 1: 2100 -7501.413 0.018 0.017
Chain 1: 2200 -7755.431 0.018 0.017
Chain 1: 2300 -7607.137 0.019 0.019
Chain 1: 2400 -7677.900 0.019 0.019
Chain 1: 2500 -7621.832 0.016 0.017
Chain 1: 2600 -7542.634 0.015 0.015
Chain 1: 2700 -7528.688 0.015 0.015
Chain 1: 2800 -7510.811 0.015 0.015
Chain 1: 2900 -7400.171 0.014 0.015
Chain 1: 3000 -7547.871 0.015 0.015
Chain 1: 3100 -7547.559 0.012 0.011
Chain 1: 3200 -7768.232 0.011 0.011
Chain 1: 3300 -7471.079 0.013 0.011
Chain 1: 3400 -7714.083 0.016 0.015
Chain 1: 3500 -7460.414 0.018 0.020
Chain 1: 3600 -7526.873 0.018 0.020
Chain 1: 3700 -7475.458 0.019 0.020
Chain 1: 3800 -7476.102 0.018 0.020
Chain 1: 3900 -7435.338 0.017 0.020
Chain 1: 4000 -7427.441 0.016 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003594 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.94 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86593.087 1.000 1.000
Chain 1: 200 -13933.741 3.107 5.215
Chain 1: 300 -10289.877 2.190 1.000
Chain 1: 400 -11271.598 1.664 1.000
Chain 1: 500 -9079.144 1.379 0.354
Chain 1: 600 -8701.366 1.157 0.354
Chain 1: 700 -8805.147 0.993 0.241
Chain 1: 800 -9374.945 0.877 0.241
Chain 1: 900 -9132.548 0.782 0.087
Chain 1: 1000 -8745.987 0.708 0.087
Chain 1: 1100 -9093.954 0.612 0.061 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8693.458 0.095 0.046
Chain 1: 1300 -8917.370 0.062 0.044
Chain 1: 1400 -8934.435 0.054 0.043
Chain 1: 1500 -8829.784 0.031 0.038
Chain 1: 1600 -8938.161 0.028 0.027
Chain 1: 1700 -9018.063 0.028 0.027
Chain 1: 1800 -8593.535 0.026 0.027
Chain 1: 1900 -8695.008 0.025 0.025
Chain 1: 2000 -8669.612 0.021 0.012
Chain 1: 2100 -8795.537 0.018 0.012
Chain 1: 2200 -8597.080 0.016 0.012
Chain 1: 2300 -8689.919 0.015 0.012
Chain 1: 2400 -8758.507 0.015 0.012
Chain 1: 2500 -8704.809 0.015 0.012
Chain 1: 2600 -8706.459 0.014 0.011
Chain 1: 2700 -8623.009 0.014 0.011
Chain 1: 2800 -8582.519 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.0035 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8417568.224 1.000 1.000
Chain 1: 200 -1585334.467 2.655 4.310
Chain 1: 300 -890134.584 2.030 1.000
Chain 1: 400 -457882.159 1.759 1.000
Chain 1: 500 -357942.676 1.463 0.944
Chain 1: 600 -232946.770 1.308 0.944
Chain 1: 700 -119381.866 1.257 0.944
Chain 1: 800 -86711.345 1.147 0.944
Chain 1: 900 -67095.353 1.052 0.781
Chain 1: 1000 -51934.512 0.976 0.781
Chain 1: 1100 -39455.719 0.908 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38636.340 0.479 0.377
Chain 1: 1300 -26626.117 0.446 0.377
Chain 1: 1400 -26349.982 0.353 0.316
Chain 1: 1500 -22947.037 0.340 0.316
Chain 1: 1600 -22167.371 0.289 0.292
Chain 1: 1700 -21044.261 0.200 0.292
Chain 1: 1800 -20989.577 0.162 0.148
Chain 1: 1900 -21315.852 0.135 0.053
Chain 1: 2000 -19828.854 0.113 0.053
Chain 1: 2100 -20066.937 0.082 0.035
Chain 1: 2200 -20293.400 0.081 0.035
Chain 1: 2300 -19910.592 0.038 0.019
Chain 1: 2400 -19682.630 0.038 0.019
Chain 1: 2500 -19484.761 0.025 0.015
Chain 1: 2600 -19114.685 0.023 0.015
Chain 1: 2700 -19071.628 0.018 0.012
Chain 1: 2800 -18788.473 0.019 0.015
Chain 1: 2900 -19069.725 0.019 0.015
Chain 1: 3000 -19055.851 0.012 0.012
Chain 1: 3100 -19140.889 0.011 0.012
Chain 1: 3200 -18831.474 0.011 0.015
Chain 1: 3300 -19036.293 0.011 0.012
Chain 1: 3400 -18511.071 0.012 0.015
Chain 1: 3500 -19123.177 0.014 0.015
Chain 1: 3600 -18429.496 0.016 0.015
Chain 1: 3700 -18816.521 0.018 0.016
Chain 1: 3800 -17775.766 0.022 0.021
Chain 1: 3900 -17771.908 0.021 0.021
Chain 1: 4000 -17889.197 0.022 0.021
Chain 1: 4100 -17802.945 0.022 0.021
Chain 1: 4200 -17619.090 0.021 0.021
Chain 1: 4300 -17757.541 0.021 0.021
Chain 1: 4400 -17714.239 0.018 0.010
Chain 1: 4500 -17616.770 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002305 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 23.05 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49381.017 1.000 1.000
Chain 1: 200 -26585.495 0.929 1.000
Chain 1: 300 -21157.269 0.705 0.857
Chain 1: 400 -28020.652 0.590 0.857
Chain 1: 500 -23838.518 0.507 0.257
Chain 1: 600 -11233.041 0.609 0.857
Chain 1: 700 -16279.073 0.567 0.310
Chain 1: 800 -15090.845 0.506 0.310
Chain 1: 900 -14814.441 0.452 0.257
Chain 1: 1000 -12552.299 0.424 0.257
Chain 1: 1100 -12190.574 0.327 0.245
Chain 1: 1200 -11936.020 0.244 0.180
Chain 1: 1300 -12187.565 0.220 0.175
Chain 1: 1400 -11262.006 0.204 0.082
Chain 1: 1500 -10992.728 0.189 0.079
Chain 1: 1600 -9840.932 0.088 0.079
Chain 1: 1700 -11159.305 0.069 0.079
Chain 1: 1800 -10657.901 0.066 0.047
Chain 1: 1900 -11364.758 0.070 0.062
Chain 1: 2000 -18518.029 0.091 0.062
Chain 1: 2100 -11336.724 0.151 0.082
Chain 1: 2200 -9837.711 0.164 0.117
Chain 1: 2300 -9599.909 0.165 0.117
Chain 1: 2400 -9373.913 0.159 0.117
Chain 1: 2500 -16596.708 0.200 0.118
Chain 1: 2600 -9615.581 0.261 0.152
Chain 1: 2700 -19159.270 0.299 0.386
Chain 1: 2800 -9044.839 0.406 0.435
Chain 1: 2900 -10207.771 0.411 0.435
Chain 1: 3000 -10052.800 0.374 0.435
Chain 1: 3100 -9576.375 0.316 0.152
Chain 1: 3200 -15115.971 0.337 0.366
Chain 1: 3300 -14446.020 0.339 0.366
Chain 1: 3400 -9897.868 0.383 0.435
Chain 1: 3500 -9165.465 0.347 0.366
Chain 1: 3600 -12627.682 0.302 0.274
Chain 1: 3700 -8820.239 0.296 0.274
Chain 1: 3800 -15057.150 0.225 0.274
Chain 1: 3900 -8831.773 0.284 0.366
Chain 1: 4000 -10732.095 0.300 0.366
Chain 1: 4100 -9266.413 0.311 0.366
Chain 1: 4200 -9253.378 0.275 0.274
Chain 1: 4300 -11016.873 0.286 0.274
Chain 1: 4400 -9500.046 0.256 0.177
Chain 1: 4500 -9414.698 0.249 0.177
Chain 1: 4600 -13125.409 0.250 0.177
Chain 1: 4700 -13511.146 0.210 0.160
Chain 1: 4800 -8806.081 0.222 0.160
Chain 1: 4900 -9080.385 0.154 0.160
Chain 1: 5000 -11663.962 0.159 0.160
Chain 1: 5100 -12249.620 0.148 0.160
Chain 1: 5200 -9892.557 0.171 0.160
Chain 1: 5300 -11127.310 0.166 0.160
Chain 1: 5400 -9720.677 0.165 0.145
Chain 1: 5500 -9270.550 0.169 0.145
Chain 1: 5600 -10010.792 0.148 0.111
Chain 1: 5700 -12627.726 0.166 0.145
Chain 1: 5800 -9527.149 0.145 0.145
Chain 1: 5900 -9314.458 0.144 0.145
Chain 1: 6000 -8629.744 0.130 0.111
Chain 1: 6100 -11142.392 0.148 0.145
Chain 1: 6200 -8395.920 0.157 0.145
Chain 1: 6300 -11737.312 0.174 0.207
Chain 1: 6400 -11301.175 0.163 0.207
Chain 1: 6500 -9480.069 0.178 0.207
Chain 1: 6600 -9691.660 0.172 0.207
Chain 1: 6700 -14611.684 0.185 0.226
Chain 1: 6800 -13528.393 0.161 0.192
Chain 1: 6900 -12087.622 0.171 0.192
Chain 1: 7000 -9953.316 0.184 0.214
Chain 1: 7100 -8971.180 0.172 0.192
Chain 1: 7200 -8957.932 0.140 0.119
Chain 1: 7300 -10216.165 0.124 0.119
Chain 1: 7400 -11696.944 0.133 0.123
Chain 1: 7500 -9071.185 0.142 0.123
Chain 1: 7600 -8585.296 0.146 0.123
Chain 1: 7700 -8692.041 0.113 0.119
Chain 1: 7800 -15444.861 0.149 0.123
Chain 1: 7900 -9263.685 0.204 0.127
Chain 1: 8000 -9395.584 0.184 0.123
Chain 1: 8100 -8780.062 0.180 0.123
Chain 1: 8200 -8338.745 0.185 0.123
Chain 1: 8300 -8273.900 0.173 0.070
Chain 1: 8400 -9030.138 0.169 0.070
Chain 1: 8500 -8316.863 0.149 0.070
Chain 1: 8600 -9170.312 0.152 0.084
Chain 1: 8700 -9867.515 0.158 0.084
Chain 1: 8800 -8935.044 0.125 0.084
Chain 1: 8900 -9456.519 0.064 0.071
Chain 1: 9000 -8399.480 0.075 0.084
Chain 1: 9100 -8835.508 0.073 0.084
Chain 1: 9200 -8426.408 0.072 0.084
Chain 1: 9300 -12164.181 0.102 0.086
Chain 1: 9400 -11892.498 0.096 0.086
Chain 1: 9500 -9992.851 0.107 0.093
Chain 1: 9600 -8509.558 0.115 0.104
Chain 1: 9700 -8177.349 0.112 0.104
Chain 1: 9800 -11434.729 0.130 0.126
Chain 1: 9900 -10102.208 0.138 0.132
Chain 1: 10000 -8440.015 0.145 0.174
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001584 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.84 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -59342.657 1.000 1.000
Chain 1: 200 -17940.687 1.654 2.308
Chain 1: 300 -8960.734 1.437 1.002
Chain 1: 400 -8426.048 1.093 1.002
Chain 1: 500 -8323.752 0.877 1.000
Chain 1: 600 -9487.120 0.751 1.000
Chain 1: 700 -8107.533 0.668 0.170
Chain 1: 800 -7808.274 0.590 0.170
Chain 1: 900 -7979.742 0.526 0.123
Chain 1: 1000 -7872.809 0.475 0.123
Chain 1: 1100 -7827.756 0.376 0.063
Chain 1: 1200 -7611.291 0.148 0.038
Chain 1: 1300 -7724.693 0.049 0.028
Chain 1: 1400 -7789.361 0.044 0.021
Chain 1: 1500 -7595.293 0.045 0.026
Chain 1: 1600 -7737.903 0.034 0.021
Chain 1: 1700 -7615.377 0.019 0.018
Chain 1: 1800 -7671.332 0.016 0.016
Chain 1: 1900 -7751.840 0.015 0.015
Chain 1: 2000 -7684.121 0.014 0.015
Chain 1: 2100 -7604.039 0.015 0.015
Chain 1: 2200 -7733.964 0.014 0.015
Chain 1: 2300 -7577.367 0.014 0.016
Chain 1: 2400 -7547.312 0.014 0.016
Chain 1: 2500 -7539.584 0.011 0.011
Chain 1: 2600 -7534.559 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004057 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 40.57 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85768.293 1.000 1.000
Chain 1: 200 -13814.700 3.104 5.208
Chain 1: 300 -10090.752 2.193 1.000
Chain 1: 400 -11604.706 1.677 1.000
Chain 1: 500 -8926.154 1.402 0.369
Chain 1: 600 -9294.558 1.175 0.369
Chain 1: 700 -8724.304 1.016 0.300
Chain 1: 800 -8656.032 0.890 0.300
Chain 1: 900 -8843.602 0.794 0.130
Chain 1: 1000 -8663.101 0.716 0.130
Chain 1: 1100 -8857.801 0.618 0.065 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8405.519 0.103 0.054
Chain 1: 1300 -8707.812 0.070 0.040
Chain 1: 1400 -8578.359 0.058 0.035
Chain 1: 1500 -8636.007 0.029 0.022
Chain 1: 1600 -8684.756 0.025 0.021
Chain 1: 1700 -8743.141 0.019 0.021
Chain 1: 1800 -8318.874 0.024 0.021
Chain 1: 1900 -8416.369 0.023 0.021
Chain 1: 2000 -8403.221 0.021 0.015
Chain 1: 2100 -8524.993 0.020 0.014
Chain 1: 2200 -8316.330 0.017 0.014
Chain 1: 2300 -8410.909 0.015 0.012
Chain 1: 2400 -8478.065 0.014 0.011
Chain 1: 2500 -8426.583 0.014 0.011
Chain 1: 2600 -8439.460 0.014 0.011
Chain 1: 2700 -8347.502 0.014 0.011
Chain 1: 2800 -8295.592 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003879 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 38.79 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8388754.643 1.000 1.000
Chain 1: 200 -1580088.641 2.655 4.309
Chain 1: 300 -890412.052 2.028 1.000
Chain 1: 400 -457683.107 1.757 1.000
Chain 1: 500 -358499.816 1.461 0.945
Chain 1: 600 -233465.598 1.307 0.945
Chain 1: 700 -119669.828 1.256 0.945
Chain 1: 800 -86875.455 1.146 0.945
Chain 1: 900 -67201.537 1.051 0.775
Chain 1: 1000 -51993.251 0.975 0.775
Chain 1: 1100 -39456.251 0.907 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38636.155 0.478 0.377
Chain 1: 1300 -26562.541 0.446 0.377
Chain 1: 1400 -26281.504 0.353 0.318
Chain 1: 1500 -22860.799 0.340 0.318
Chain 1: 1600 -22076.095 0.290 0.293
Chain 1: 1700 -20945.457 0.201 0.293
Chain 1: 1800 -20889.120 0.163 0.150
Chain 1: 1900 -21215.829 0.135 0.054
Chain 1: 2000 -19723.828 0.114 0.054
Chain 1: 2100 -19962.270 0.083 0.036
Chain 1: 2200 -20189.557 0.082 0.036
Chain 1: 2300 -19805.931 0.039 0.019
Chain 1: 2400 -19577.819 0.039 0.019
Chain 1: 2500 -19380.012 0.025 0.015
Chain 1: 2600 -19009.444 0.023 0.015
Chain 1: 2700 -18966.254 0.018 0.012
Chain 1: 2800 -18682.944 0.019 0.015
Chain 1: 2900 -18964.463 0.019 0.015
Chain 1: 3000 -18950.577 0.012 0.012
Chain 1: 3100 -19035.656 0.011 0.012
Chain 1: 3200 -18725.946 0.011 0.015
Chain 1: 3300 -18930.999 0.011 0.012
Chain 1: 3400 -18405.264 0.012 0.015
Chain 1: 3500 -19018.177 0.015 0.015
Chain 1: 3600 -18323.550 0.016 0.015
Chain 1: 3700 -18711.337 0.018 0.017
Chain 1: 3800 -17669.045 0.023 0.021
Chain 1: 3900 -17665.193 0.021 0.021
Chain 1: 4000 -17782.449 0.022 0.021
Chain 1: 4100 -17696.123 0.022 0.021
Chain 1: 4200 -17511.975 0.021 0.021
Chain 1: 4300 -17650.629 0.021 0.021
Chain 1: 4400 -17607.089 0.018 0.011
Chain 1: 4500 -17509.613 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001275 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.75 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12890.777 1.000 1.000
Chain 1: 200 -9840.131 0.655 1.000
Chain 1: 300 -8352.375 0.496 0.310
Chain 1: 400 -8584.118 0.379 0.310
Chain 1: 500 -8453.786 0.306 0.178
Chain 1: 600 -8309.907 0.258 0.178
Chain 1: 700 -8221.747 0.223 0.027
Chain 1: 800 -8169.183 0.196 0.027
Chain 1: 900 -8189.592 0.174 0.017
Chain 1: 1000 -8290.881 0.158 0.017
Chain 1: 1100 -8316.159 0.058 0.015
Chain 1: 1200 -8221.405 0.028 0.012
Chain 1: 1300 -8143.940 0.012 0.012
Chain 1: 1400 -8157.392 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001513 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.13 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -59110.403 1.000 1.000
Chain 1: 200 -18216.915 1.622 2.245
Chain 1: 300 -9160.933 1.411 1.000
Chain 1: 400 -8589.469 1.075 1.000
Chain 1: 500 -8945.175 0.868 0.989
Chain 1: 600 -8685.949 0.728 0.989
Chain 1: 700 -8036.745 0.636 0.081
Chain 1: 800 -8563.002 0.564 0.081
Chain 1: 900 -8228.481 0.506 0.067
Chain 1: 1000 -8101.605 0.457 0.067
Chain 1: 1100 -7925.469 0.359 0.061
Chain 1: 1200 -7916.372 0.135 0.041
Chain 1: 1300 -7953.936 0.036 0.040
Chain 1: 1400 -7921.256 0.030 0.030
Chain 1: 1500 -7680.049 0.029 0.030
Chain 1: 1600 -7940.761 0.030 0.031
Chain 1: 1700 -7716.529 0.024 0.029
Chain 1: 1800 -7756.065 0.019 0.022
Chain 1: 1900 -7822.140 0.015 0.016
Chain 1: 2000 -7895.947 0.015 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003875 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 38.75 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86445.670 1.000 1.000
Chain 1: 200 -14106.131 3.064 5.128
Chain 1: 300 -10351.589 2.164 1.000
Chain 1: 400 -12022.154 1.657 1.000
Chain 1: 500 -8954.833 1.394 0.363
Chain 1: 600 -9090.319 1.165 0.363
Chain 1: 700 -9493.595 1.004 0.343
Chain 1: 800 -8589.067 0.892 0.343
Chain 1: 900 -8730.558 0.795 0.139
Chain 1: 1000 -8965.050 0.718 0.139
Chain 1: 1100 -9127.221 0.620 0.105 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8639.902 0.112 0.056
Chain 1: 1300 -8998.640 0.080 0.042
Chain 1: 1400 -8750.499 0.069 0.040
Chain 1: 1500 -8826.153 0.036 0.028
Chain 1: 1600 -8929.067 0.035 0.028
Chain 1: 1700 -8985.820 0.032 0.026
Chain 1: 1800 -8536.163 0.026 0.026
Chain 1: 1900 -8644.168 0.026 0.026
Chain 1: 2000 -8640.422 0.023 0.018
Chain 1: 2100 -8795.076 0.023 0.018
Chain 1: 2200 -8541.458 0.021 0.018
Chain 1: 2300 -8716.553 0.019 0.018
Chain 1: 2400 -8539.359 0.018 0.018
Chain 1: 2500 -8617.306 0.018 0.018
Chain 1: 2600 -8650.044 0.017 0.018
Chain 1: 2700 -8570.100 0.018 0.018
Chain 1: 2800 -8520.866 0.013 0.012
Chain 1: 2900 -8630.343 0.013 0.013
Chain 1: 3000 -8560.988 0.014 0.013
Chain 1: 3100 -8505.534 0.013 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003215 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.15 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8386024.956 1.000 1.000
Chain 1: 200 -1577671.844 2.658 4.315
Chain 1: 300 -891441.740 2.028 1.000
Chain 1: 400 -458880.189 1.757 1.000
Chain 1: 500 -359865.444 1.461 0.943
Chain 1: 600 -234676.323 1.306 0.943
Chain 1: 700 -120432.451 1.255 0.943
Chain 1: 800 -87517.106 1.145 0.943
Chain 1: 900 -67745.501 1.050 0.770
Chain 1: 1000 -52463.740 0.974 0.770
Chain 1: 1100 -39860.666 0.906 0.533 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39033.399 0.477 0.376
Chain 1: 1300 -26894.498 0.445 0.376
Chain 1: 1400 -26607.284 0.352 0.316
Chain 1: 1500 -23169.787 0.339 0.316
Chain 1: 1600 -22380.173 0.289 0.292
Chain 1: 1700 -21241.728 0.200 0.291
Chain 1: 1800 -21183.550 0.162 0.148
Chain 1: 1900 -21510.328 0.135 0.054
Chain 1: 2000 -20014.143 0.113 0.054
Chain 1: 2100 -20252.822 0.083 0.035
Chain 1: 2200 -20480.781 0.081 0.035
Chain 1: 2300 -20096.526 0.038 0.019
Chain 1: 2400 -19868.281 0.038 0.019
Chain 1: 2500 -19670.710 0.025 0.015
Chain 1: 2600 -19299.721 0.023 0.015
Chain 1: 2700 -19256.412 0.018 0.012
Chain 1: 2800 -18973.138 0.019 0.015
Chain 1: 2900 -19254.862 0.019 0.015
Chain 1: 3000 -19240.819 0.012 0.012
Chain 1: 3100 -19325.955 0.011 0.011
Chain 1: 3200 -19016.064 0.011 0.015
Chain 1: 3300 -19221.274 0.010 0.011
Chain 1: 3400 -18695.353 0.012 0.015
Chain 1: 3500 -19308.602 0.014 0.015
Chain 1: 3600 -18613.563 0.016 0.015
Chain 1: 3700 -19001.733 0.018 0.016
Chain 1: 3800 -17958.801 0.022 0.020
Chain 1: 3900 -17954.969 0.021 0.020
Chain 1: 4000 -18072.188 0.021 0.020
Chain 1: 4100 -17985.821 0.021 0.020
Chain 1: 4200 -17801.543 0.021 0.020
Chain 1: 4300 -17940.264 0.021 0.020
Chain 1: 4400 -17896.626 0.018 0.010
Chain 1: 4500 -17799.132 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001534 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.34 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49749.136 1.000 1.000
Chain 1: 200 -18483.543 1.346 1.692
Chain 1: 300 -33020.862 1.044 1.000
Chain 1: 400 -22950.117 0.893 1.000
Chain 1: 500 -24535.763 0.727 0.440
Chain 1: 600 -21197.288 0.632 0.440
Chain 1: 700 -13820.283 0.618 0.440
Chain 1: 800 -16072.554 0.558 0.440
Chain 1: 900 -14482.143 0.508 0.439
Chain 1: 1000 -12955.675 0.469 0.439
Chain 1: 1100 -18513.521 0.399 0.300
Chain 1: 1200 -14198.780 0.261 0.300
Chain 1: 1300 -13100.412 0.225 0.157
Chain 1: 1400 -11172.492 0.198 0.157
Chain 1: 1500 -10747.874 0.196 0.157
Chain 1: 1600 -23505.340 0.234 0.173
Chain 1: 1700 -11247.482 0.290 0.173
Chain 1: 1800 -10452.889 0.284 0.173
Chain 1: 1900 -12303.137 0.288 0.173
Chain 1: 2000 -11966.954 0.279 0.173
Chain 1: 2100 -10959.768 0.258 0.150
Chain 1: 2200 -10156.478 0.235 0.092
Chain 1: 2300 -9988.139 0.229 0.092
Chain 1: 2400 -10700.516 0.218 0.079
Chain 1: 2500 -9831.258 0.223 0.088
Chain 1: 2600 -13368.507 0.195 0.088
Chain 1: 2700 -11757.521 0.100 0.088
Chain 1: 2800 -10161.617 0.108 0.092
Chain 1: 2900 -9708.711 0.098 0.088
Chain 1: 3000 -9346.603 0.099 0.088
Chain 1: 3100 -11376.891 0.107 0.088
Chain 1: 3200 -9924.736 0.114 0.137
Chain 1: 3300 -18850.340 0.160 0.146
Chain 1: 3400 -9322.617 0.255 0.157
Chain 1: 3500 -10729.474 0.260 0.157
Chain 1: 3600 -10244.616 0.238 0.146
Chain 1: 3700 -9435.628 0.233 0.146
Chain 1: 3800 -9097.428 0.221 0.131
Chain 1: 3900 -13861.843 0.250 0.146
Chain 1: 4000 -10128.603 0.283 0.178
Chain 1: 4100 -9963.028 0.267 0.146
Chain 1: 4200 -12939.507 0.276 0.230
Chain 1: 4300 -9884.818 0.259 0.230
Chain 1: 4400 -18404.573 0.203 0.230
Chain 1: 4500 -9485.441 0.284 0.309
Chain 1: 4600 -9201.638 0.282 0.309
Chain 1: 4700 -9363.179 0.276 0.309
Chain 1: 4800 -9147.103 0.274 0.309
Chain 1: 4900 -9045.624 0.241 0.230
Chain 1: 5000 -14894.985 0.243 0.230
Chain 1: 5100 -11656.590 0.270 0.278
Chain 1: 5200 -10173.455 0.261 0.278
Chain 1: 5300 -9571.193 0.237 0.146
Chain 1: 5400 -11284.903 0.205 0.146
Chain 1: 5500 -9365.884 0.132 0.146
Chain 1: 5600 -9966.731 0.135 0.146
Chain 1: 5700 -9674.050 0.136 0.146
Chain 1: 5800 -8991.401 0.141 0.146
Chain 1: 5900 -10118.482 0.151 0.146
Chain 1: 6000 -9427.483 0.119 0.111
Chain 1: 6100 -9105.322 0.095 0.076
Chain 1: 6200 -8828.298 0.084 0.073
Chain 1: 6300 -9022.139 0.080 0.073
Chain 1: 6400 -11741.829 0.088 0.073
Chain 1: 6500 -9164.210 0.095 0.073
Chain 1: 6600 -10795.596 0.104 0.076
Chain 1: 6700 -13170.192 0.119 0.111
Chain 1: 6800 -8852.706 0.160 0.151
Chain 1: 6900 -11839.665 0.175 0.180
Chain 1: 7000 -8768.789 0.202 0.232
Chain 1: 7100 -11417.827 0.222 0.232
Chain 1: 7200 -13698.311 0.235 0.232
Chain 1: 7300 -8909.645 0.287 0.252
Chain 1: 7400 -8924.873 0.264 0.252
Chain 1: 7500 -8706.517 0.238 0.232
Chain 1: 7600 -9074.295 0.227 0.232
Chain 1: 7700 -11581.049 0.231 0.232
Chain 1: 7800 -9231.616 0.208 0.232
Chain 1: 7900 -9229.010 0.182 0.216
Chain 1: 8000 -8653.755 0.154 0.166
Chain 1: 8100 -8877.240 0.133 0.066
Chain 1: 8200 -8970.891 0.118 0.041
Chain 1: 8300 -12389.163 0.092 0.041
Chain 1: 8400 -9249.775 0.125 0.066
Chain 1: 8500 -9015.748 0.126 0.066
Chain 1: 8600 -12880.294 0.151 0.216
Chain 1: 8700 -11295.491 0.144 0.140
Chain 1: 8800 -10471.601 0.126 0.079
Chain 1: 8900 -9666.232 0.135 0.083
Chain 1: 9000 -8954.064 0.136 0.083
Chain 1: 9100 -9309.773 0.137 0.083
Chain 1: 9200 -9387.895 0.137 0.083
Chain 1: 9300 -8820.018 0.116 0.080
Chain 1: 9400 -12839.029 0.113 0.080
Chain 1: 9500 -9225.088 0.150 0.083
Chain 1: 9600 -8832.421 0.124 0.080
Chain 1: 9700 -8533.060 0.114 0.079
Chain 1: 9800 -8883.218 0.110 0.064
Chain 1: 9900 -10572.631 0.117 0.064
Chain 1: 10000 -8527.763 0.133 0.064
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001651 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 16.51 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -59528.887 1.000 1.000
Chain 1: 200 -18428.949 1.615 2.230
Chain 1: 300 -9237.937 1.408 1.000
Chain 1: 400 -8404.609 1.081 1.000
Chain 1: 500 -8405.003 0.865 0.995
Chain 1: 600 -9266.447 0.736 0.995
Chain 1: 700 -8404.996 0.646 0.102
Chain 1: 800 -7982.697 0.572 0.102
Chain 1: 900 -8278.379 0.512 0.099
Chain 1: 1000 -7719.334 0.468 0.099
Chain 1: 1100 -7863.910 0.370 0.093
Chain 1: 1200 -7953.549 0.148 0.072
Chain 1: 1300 -7680.386 0.052 0.053
Chain 1: 1400 -7988.032 0.046 0.039
Chain 1: 1500 -7729.785 0.049 0.039
Chain 1: 1600 -7846.662 0.042 0.036
Chain 1: 1700 -7679.524 0.033 0.036
Chain 1: 1800 -7782.195 0.030 0.033
Chain 1: 1900 -7601.686 0.028 0.024
Chain 1: 2000 -7751.725 0.023 0.022
Chain 1: 2100 -7612.530 0.023 0.022
Chain 1: 2200 -7935.537 0.026 0.024
Chain 1: 2300 -7676.756 0.026 0.024
Chain 1: 2400 -7779.423 0.023 0.022
Chain 1: 2500 -7736.329 0.020 0.019
Chain 1: 2600 -7639.613 0.020 0.019
Chain 1: 2700 -7659.853 0.018 0.018
Chain 1: 2800 -7728.682 0.018 0.018
Chain 1: 2900 -7481.955 0.019 0.018
Chain 1: 3000 -7642.865 0.019 0.018
Chain 1: 3100 -7640.368 0.017 0.013
Chain 1: 3200 -7858.826 0.016 0.013
Chain 1: 3300 -7582.253 0.016 0.013
Chain 1: 3400 -7827.702 0.018 0.021
Chain 1: 3500 -7550.282 0.021 0.028
Chain 1: 3600 -7615.285 0.021 0.028
Chain 1: 3700 -7563.962 0.021 0.028
Chain 1: 3800 -7564.581 0.020 0.028
Chain 1: 3900 -7518.423 0.018 0.021
Chain 1: 4000 -7513.358 0.015 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003033 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.33 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86304.301 1.000 1.000
Chain 1: 200 -14147.130 3.050 5.100
Chain 1: 300 -10401.405 2.154 1.000
Chain 1: 400 -12131.558 1.651 1.000
Chain 1: 500 -9039.821 1.389 0.360
Chain 1: 600 -8967.394 1.159 0.360
Chain 1: 700 -9397.318 1.000 0.342
Chain 1: 800 -9790.446 0.880 0.342
Chain 1: 900 -9157.990 0.790 0.143
Chain 1: 1000 -8738.356 0.716 0.143
Chain 1: 1100 -8953.477 0.618 0.069 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8725.985 0.111 0.048
Chain 1: 1300 -9008.386 0.078 0.046
Chain 1: 1400 -8832.445 0.065 0.040
Chain 1: 1500 -8891.122 0.032 0.031
Chain 1: 1600 -8997.122 0.032 0.031
Chain 1: 1700 -9052.099 0.028 0.026
Chain 1: 1800 -8604.575 0.029 0.026
Chain 1: 1900 -8712.902 0.024 0.024
Chain 1: 2000 -8689.755 0.019 0.020
Chain 1: 2100 -8830.769 0.018 0.016
Chain 1: 2200 -8607.508 0.018 0.016
Chain 1: 2300 -8717.056 0.017 0.013
Chain 1: 2400 -8773.106 0.015 0.012
Chain 1: 2500 -8721.278 0.015 0.012
Chain 1: 2600 -8736.257 0.014 0.012
Chain 1: 2700 -8643.091 0.015 0.012
Chain 1: 2800 -8588.624 0.010 0.011
Chain 1: 2900 -8689.798 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004367 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 43.67 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8427704.871 1.000 1.000
Chain 1: 200 -1585498.829 2.658 4.315
Chain 1: 300 -891348.965 2.031 1.000
Chain 1: 400 -458275.983 1.760 1.000
Chain 1: 500 -358459.012 1.464 0.945
Chain 1: 600 -233371.934 1.309 0.945
Chain 1: 700 -119737.985 1.258 0.945
Chain 1: 800 -87017.277 1.147 0.945
Chain 1: 900 -67386.851 1.052 0.779
Chain 1: 1000 -52221.016 0.976 0.779
Chain 1: 1100 -39725.521 0.908 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38909.780 0.478 0.376
Chain 1: 1300 -26875.087 0.445 0.376
Chain 1: 1400 -26598.739 0.351 0.315
Chain 1: 1500 -23188.684 0.338 0.315
Chain 1: 1600 -22407.455 0.288 0.291
Chain 1: 1700 -21281.144 0.199 0.290
Chain 1: 1800 -21225.935 0.161 0.147
Chain 1: 1900 -21552.711 0.134 0.053
Chain 1: 2000 -20063.019 0.112 0.053
Chain 1: 2100 -20301.252 0.082 0.035
Chain 1: 2200 -20528.302 0.081 0.035
Chain 1: 2300 -20144.900 0.038 0.019
Chain 1: 2400 -19916.780 0.038 0.019
Chain 1: 2500 -19718.892 0.024 0.015
Chain 1: 2600 -19348.305 0.023 0.015
Chain 1: 2700 -19305.112 0.018 0.012
Chain 1: 2800 -19021.720 0.019 0.015
Chain 1: 2900 -19303.237 0.019 0.015
Chain 1: 3000 -19289.368 0.011 0.012
Chain 1: 3100 -19374.454 0.011 0.011
Chain 1: 3200 -19064.695 0.011 0.015
Chain 1: 3300 -19269.784 0.010 0.011
Chain 1: 3400 -18743.940 0.012 0.015
Chain 1: 3500 -19356.951 0.014 0.015
Chain 1: 3600 -18662.161 0.016 0.015
Chain 1: 3700 -19050.033 0.018 0.016
Chain 1: 3800 -18007.476 0.022 0.020
Chain 1: 3900 -18003.583 0.021 0.020
Chain 1: 4000 -18120.881 0.021 0.020
Chain 1: 4100 -18034.536 0.021 0.020
Chain 1: 4200 -17850.302 0.021 0.020
Chain 1: 4300 -17989.017 0.020 0.020
Chain 1: 4400 -17945.409 0.018 0.010
Chain 1: 4500 -17847.904 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001343 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.43 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49005.676 1.000 1.000
Chain 1: 200 -20112.965 1.218 1.437
Chain 1: 300 -18589.119 0.839 1.000
Chain 1: 400 -16800.714 0.656 1.000
Chain 1: 500 -17249.225 0.530 0.106
Chain 1: 600 -18350.931 0.452 0.106
Chain 1: 700 -15060.256 0.418 0.106
Chain 1: 800 -11308.816 0.408 0.219
Chain 1: 900 -11797.421 0.367 0.106
Chain 1: 1000 -12828.905 0.338 0.106
Chain 1: 1100 -16719.187 0.262 0.106
Chain 1: 1200 -10586.123 0.176 0.106
Chain 1: 1300 -12548.288 0.183 0.156
Chain 1: 1400 -10712.856 0.190 0.171
Chain 1: 1500 -13385.121 0.207 0.200
Chain 1: 1600 -11844.048 0.214 0.200
Chain 1: 1700 -17724.773 0.225 0.200
Chain 1: 1800 -10120.241 0.267 0.200
Chain 1: 1900 -10860.837 0.270 0.200
Chain 1: 2000 -18632.696 0.304 0.233
Chain 1: 2100 -10953.895 0.351 0.332
Chain 1: 2200 -10476.273 0.297 0.200
Chain 1: 2300 -11873.197 0.293 0.200
Chain 1: 2400 -9594.363 0.300 0.238
Chain 1: 2500 -16079.168 0.320 0.332
Chain 1: 2600 -9451.653 0.377 0.403
Chain 1: 2700 -10742.046 0.356 0.403
Chain 1: 2800 -16467.838 0.316 0.348
Chain 1: 2900 -9168.111 0.389 0.403
Chain 1: 3000 -16590.850 0.392 0.403
Chain 1: 3100 -10225.118 0.384 0.403
Chain 1: 3200 -14966.915 0.411 0.403
Chain 1: 3300 -9596.124 0.455 0.447
Chain 1: 3400 -12151.811 0.453 0.447
Chain 1: 3500 -9957.277 0.434 0.447
Chain 1: 3600 -9818.987 0.366 0.348
Chain 1: 3700 -9190.180 0.360 0.348
Chain 1: 3800 -11267.050 0.344 0.317
Chain 1: 3900 -9985.768 0.277 0.220
Chain 1: 4000 -16605.683 0.272 0.220
Chain 1: 4100 -11195.330 0.258 0.220
Chain 1: 4200 -9763.627 0.241 0.210
Chain 1: 4300 -9402.394 0.189 0.184
Chain 1: 4400 -8755.387 0.176 0.147
Chain 1: 4500 -9098.948 0.157 0.128
Chain 1: 4600 -14496.097 0.193 0.147
Chain 1: 4700 -13131.779 0.197 0.147
Chain 1: 4800 -9146.352 0.222 0.147
Chain 1: 4900 -8725.421 0.214 0.147
Chain 1: 5000 -10032.662 0.187 0.130
Chain 1: 5100 -8624.794 0.155 0.130
Chain 1: 5200 -11009.876 0.162 0.130
Chain 1: 5300 -8876.265 0.182 0.163
Chain 1: 5400 -13481.463 0.209 0.217
Chain 1: 5500 -12759.814 0.211 0.217
Chain 1: 5600 -12432.574 0.176 0.163
Chain 1: 5700 -14596.921 0.181 0.163
Chain 1: 5800 -11523.005 0.164 0.163
Chain 1: 5900 -10798.197 0.166 0.163
Chain 1: 6000 -12342.894 0.165 0.163
Chain 1: 6100 -8846.545 0.188 0.217
Chain 1: 6200 -8753.975 0.168 0.148
Chain 1: 6300 -12275.089 0.172 0.148
Chain 1: 6400 -13178.133 0.145 0.125
Chain 1: 6500 -9908.926 0.172 0.148
Chain 1: 6600 -8496.341 0.186 0.166
Chain 1: 6700 -9043.060 0.178 0.166
Chain 1: 6800 -9851.043 0.159 0.125
Chain 1: 6900 -9760.755 0.153 0.125
Chain 1: 7000 -8699.988 0.153 0.122
Chain 1: 7100 -8346.206 0.118 0.082
Chain 1: 7200 -8385.794 0.117 0.082
Chain 1: 7300 -8989.017 0.095 0.069
Chain 1: 7400 -8998.744 0.089 0.067
Chain 1: 7500 -9752.663 0.063 0.067
Chain 1: 7600 -8423.519 0.062 0.067
Chain 1: 7700 -12195.799 0.087 0.077
Chain 1: 7800 -10734.171 0.093 0.077
Chain 1: 7900 -13063.941 0.110 0.122
Chain 1: 8000 -9689.801 0.132 0.136
Chain 1: 8100 -9897.467 0.130 0.136
Chain 1: 8200 -12065.458 0.148 0.158
Chain 1: 8300 -8526.024 0.182 0.178
Chain 1: 8400 -11151.319 0.206 0.180
Chain 1: 8500 -8150.977 0.235 0.235
Chain 1: 8600 -8824.034 0.227 0.235
Chain 1: 8700 -10189.636 0.209 0.180
Chain 1: 8800 -8678.011 0.213 0.180
Chain 1: 8900 -8727.983 0.196 0.180
Chain 1: 9000 -8883.254 0.163 0.174
Chain 1: 9100 -10707.996 0.178 0.174
Chain 1: 9200 -10039.721 0.166 0.170
Chain 1: 9300 -10105.057 0.125 0.134
Chain 1: 9400 -9638.083 0.107 0.076
Chain 1: 9500 -8948.371 0.078 0.076
Chain 1: 9600 -10318.927 0.083 0.077
Chain 1: 9700 -8311.751 0.094 0.077
Chain 1: 9800 -10866.445 0.100 0.077
Chain 1: 9900 -11225.467 0.103 0.077
Chain 1: 10000 -8995.837 0.126 0.133
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001529 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.29 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58225.024 1.000 1.000
Chain 1: 200 -17854.051 1.631 2.261
Chain 1: 300 -8774.864 1.432 1.035
Chain 1: 400 -8076.463 1.096 1.035
Chain 1: 500 -8782.424 0.893 1.000
Chain 1: 600 -8827.640 0.745 1.000
Chain 1: 700 -7947.168 0.654 0.111
Chain 1: 800 -8091.033 0.575 0.111
Chain 1: 900 -8168.058 0.512 0.086
Chain 1: 1000 -7947.259 0.463 0.086
Chain 1: 1100 -7712.324 0.366 0.080
Chain 1: 1200 -7583.604 0.142 0.030
Chain 1: 1300 -7789.996 0.041 0.028
Chain 1: 1400 -7925.816 0.034 0.026
Chain 1: 1500 -7583.430 0.031 0.026
Chain 1: 1600 -7607.934 0.031 0.026
Chain 1: 1700 -7561.005 0.020 0.018
Chain 1: 1800 -7612.630 0.019 0.017
Chain 1: 1900 -7611.245 0.018 0.017
Chain 1: 2000 -7665.135 0.016 0.017
Chain 1: 2100 -7534.772 0.015 0.017
Chain 1: 2200 -7690.268 0.015 0.017
Chain 1: 2300 -7561.455 0.014 0.017
Chain 1: 2400 -7665.599 0.014 0.014
Chain 1: 2500 -7655.842 0.009 0.007 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003621 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 36.21 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85987.632 1.000 1.000
Chain 1: 200 -13731.182 3.131 5.262
Chain 1: 300 -10054.683 2.209 1.000
Chain 1: 400 -11070.542 1.680 1.000
Chain 1: 500 -9038.348 1.389 0.366
Chain 1: 600 -8486.411 1.168 0.366
Chain 1: 700 -8678.230 1.005 0.225
Chain 1: 800 -9315.922 0.888 0.225
Chain 1: 900 -8895.881 0.794 0.092
Chain 1: 1000 -8739.475 0.717 0.092
Chain 1: 1100 -8832.812 0.618 0.068 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8345.381 0.097 0.065
Chain 1: 1300 -8704.357 0.065 0.058
Chain 1: 1400 -8672.120 0.056 0.047
Chain 1: 1500 -8597.338 0.034 0.041
Chain 1: 1600 -8699.971 0.029 0.022
Chain 1: 1700 -8768.897 0.028 0.018
Chain 1: 1800 -8337.898 0.026 0.018
Chain 1: 1900 -8441.955 0.022 0.012
Chain 1: 2000 -8417.191 0.021 0.012
Chain 1: 2100 -8550.328 0.021 0.012
Chain 1: 2200 -8345.627 0.018 0.012
Chain 1: 2300 -8440.813 0.015 0.012
Chain 1: 2400 -8505.642 0.015 0.012
Chain 1: 2500 -8450.680 0.015 0.012
Chain 1: 2600 -8454.720 0.014 0.011
Chain 1: 2700 -8370.071 0.014 0.011
Chain 1: 2800 -8327.021 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003432 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.32 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8365938.485 1.000 1.000
Chain 1: 200 -1579916.962 2.648 4.295
Chain 1: 300 -891262.168 2.023 1.000
Chain 1: 400 -458570.350 1.753 1.000
Chain 1: 500 -359290.438 1.458 0.944
Chain 1: 600 -234092.750 1.304 0.944
Chain 1: 700 -119908.903 1.254 0.944
Chain 1: 800 -87008.039 1.144 0.944
Chain 1: 900 -67270.281 1.050 0.773
Chain 1: 1000 -52005.645 0.974 0.773
Chain 1: 1100 -39422.899 0.906 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38593.609 0.479 0.378
Chain 1: 1300 -26485.162 0.447 0.378
Chain 1: 1400 -26199.417 0.354 0.319
Chain 1: 1500 -22769.243 0.341 0.319
Chain 1: 1600 -21980.956 0.291 0.294
Chain 1: 1700 -20846.657 0.201 0.293
Chain 1: 1800 -20789.086 0.164 0.151
Chain 1: 1900 -21115.490 0.136 0.054
Chain 1: 2000 -19621.790 0.114 0.054
Chain 1: 2100 -19860.505 0.084 0.036
Chain 1: 2200 -20087.839 0.083 0.036
Chain 1: 2300 -19704.145 0.039 0.019
Chain 1: 2400 -19476.034 0.039 0.019
Chain 1: 2500 -19278.329 0.025 0.015
Chain 1: 2600 -18908.069 0.023 0.015
Chain 1: 2700 -18864.818 0.018 0.012
Chain 1: 2800 -18581.710 0.019 0.015
Chain 1: 2900 -18863.121 0.019 0.015
Chain 1: 3000 -18849.273 0.012 0.012
Chain 1: 3100 -18934.336 0.011 0.012
Chain 1: 3200 -18624.794 0.012 0.015
Chain 1: 3300 -18829.644 0.011 0.012
Chain 1: 3400 -18304.298 0.012 0.015
Chain 1: 3500 -18916.762 0.015 0.015
Chain 1: 3600 -18222.640 0.016 0.015
Chain 1: 3700 -18610.108 0.018 0.017
Chain 1: 3800 -17568.704 0.023 0.021
Chain 1: 3900 -17564.839 0.021 0.021
Chain 1: 4000 -17682.105 0.022 0.021
Chain 1: 4100 -17595.881 0.022 0.021
Chain 1: 4200 -17411.819 0.021 0.021
Chain 1: 4300 -17550.404 0.021 0.021
Chain 1: 4400 -17507.032 0.018 0.011
Chain 1: 4500 -17409.540 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001201 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.01 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49353.853 1.000 1.000
Chain 1: 200 -17012.158 1.451 1.901
Chain 1: 300 -24731.516 1.071 1.000
Chain 1: 400 -14861.813 0.969 1.000
Chain 1: 500 -15162.199 0.779 0.664
Chain 1: 600 -17248.387 0.670 0.664
Chain 1: 700 -17768.500 0.578 0.312
Chain 1: 800 -12728.970 0.555 0.396
Chain 1: 900 -16503.296 0.519 0.312
Chain 1: 1000 -12704.051 0.497 0.312
Chain 1: 1100 -12377.747 0.400 0.299
Chain 1: 1200 -16978.872 0.237 0.271
Chain 1: 1300 -12248.321 0.244 0.271
Chain 1: 1400 -10760.070 0.192 0.229
Chain 1: 1500 -11952.980 0.200 0.229
Chain 1: 1600 -11623.903 0.190 0.229
Chain 1: 1700 -9993.935 0.204 0.229
Chain 1: 1800 -11129.052 0.174 0.163
Chain 1: 1900 -11534.933 0.155 0.138
Chain 1: 2000 -16225.023 0.154 0.138
Chain 1: 2100 -11050.831 0.198 0.163
Chain 1: 2200 -12714.710 0.184 0.138
Chain 1: 2300 -12360.756 0.148 0.131
Chain 1: 2400 -10446.958 0.153 0.131
Chain 1: 2500 -9490.062 0.153 0.131
Chain 1: 2600 -10952.788 0.163 0.134
Chain 1: 2700 -9812.585 0.159 0.131
Chain 1: 2800 -11680.584 0.165 0.134
Chain 1: 2900 -9690.619 0.182 0.160
Chain 1: 3000 -9812.165 0.154 0.134
Chain 1: 3100 -10229.562 0.111 0.131
Chain 1: 3200 -9415.748 0.107 0.116
Chain 1: 3300 -10110.158 0.111 0.116
Chain 1: 3400 -9556.583 0.098 0.101
Chain 1: 3500 -9471.654 0.089 0.086
Chain 1: 3600 -10547.397 0.086 0.086
Chain 1: 3700 -9062.564 0.091 0.086
Chain 1: 3800 -13931.685 0.110 0.086
Chain 1: 3900 -9886.576 0.130 0.086
Chain 1: 4000 -10775.899 0.137 0.086
Chain 1: 4100 -9675.471 0.144 0.102
Chain 1: 4200 -12841.684 0.160 0.114
Chain 1: 4300 -10876.786 0.171 0.164
Chain 1: 4400 -12753.202 0.180 0.164
Chain 1: 4500 -10013.554 0.207 0.181
Chain 1: 4600 -14796.721 0.229 0.247
Chain 1: 4700 -13841.276 0.220 0.247
Chain 1: 4800 -9291.333 0.234 0.247
Chain 1: 4900 -15783.770 0.234 0.247
Chain 1: 5000 -14157.334 0.237 0.247
Chain 1: 5100 -15269.403 0.233 0.247
Chain 1: 5200 -9686.486 0.266 0.274
Chain 1: 5300 -11064.400 0.260 0.274
Chain 1: 5400 -8875.443 0.270 0.274
Chain 1: 5500 -9617.232 0.251 0.247
Chain 1: 5600 -8865.957 0.227 0.125
Chain 1: 5700 -9722.074 0.229 0.125
Chain 1: 5800 -11101.076 0.192 0.124
Chain 1: 5900 -14617.957 0.175 0.124
Chain 1: 6000 -9486.865 0.218 0.125
Chain 1: 6100 -8838.854 0.218 0.125
Chain 1: 6200 -9568.693 0.168 0.124
Chain 1: 6300 -10259.799 0.162 0.088
Chain 1: 6400 -9578.559 0.144 0.085
Chain 1: 6500 -8975.962 0.143 0.085
Chain 1: 6600 -8904.861 0.136 0.076
Chain 1: 6700 -9098.309 0.129 0.073
Chain 1: 6800 -9260.838 0.118 0.071
Chain 1: 6900 -13387.568 0.125 0.071
Chain 1: 7000 -10497.316 0.099 0.071
Chain 1: 7100 -8608.699 0.113 0.071
Chain 1: 7200 -11828.938 0.133 0.071
Chain 1: 7300 -10751.312 0.136 0.100
Chain 1: 7400 -13607.648 0.150 0.210
Chain 1: 7500 -11276.663 0.164 0.210
Chain 1: 7600 -8879.687 0.190 0.219
Chain 1: 7700 -9726.992 0.197 0.219
Chain 1: 7800 -8454.883 0.210 0.219
Chain 1: 7900 -8599.157 0.181 0.210
Chain 1: 8000 -10237.178 0.169 0.207
Chain 1: 8100 -9184.609 0.159 0.160
Chain 1: 8200 -8788.598 0.136 0.150
Chain 1: 8300 -9451.184 0.133 0.150
Chain 1: 8400 -9147.292 0.115 0.115
Chain 1: 8500 -8699.974 0.100 0.087
Chain 1: 8600 -9429.137 0.081 0.077
Chain 1: 8700 -8645.921 0.081 0.077
Chain 1: 8800 -11142.823 0.088 0.077
Chain 1: 8900 -9717.655 0.101 0.091
Chain 1: 9000 -10181.528 0.090 0.077
Chain 1: 9100 -9179.786 0.089 0.077
Chain 1: 9200 -9141.328 0.085 0.077
Chain 1: 9300 -9468.660 0.082 0.077
Chain 1: 9400 -11373.263 0.095 0.091
Chain 1: 9500 -9380.466 0.111 0.109
Chain 1: 9600 -11898.856 0.125 0.147
Chain 1: 9700 -12736.576 0.122 0.147
Chain 1: 9800 -11229.099 0.113 0.134
Chain 1: 9900 -8635.943 0.129 0.134
Chain 1: 10000 -8708.980 0.125 0.134
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001451 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.51 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57504.271 1.000 1.000
Chain 1: 200 -18061.087 1.592 2.184
Chain 1: 300 -9015.895 1.396 1.003
Chain 1: 400 -8214.760 1.071 1.003
Chain 1: 500 -8571.041 0.865 1.000
Chain 1: 600 -7945.965 0.734 1.000
Chain 1: 700 -7900.367 0.630 0.098
Chain 1: 800 -8468.833 0.560 0.098
Chain 1: 900 -8420.754 0.498 0.079
Chain 1: 1000 -8041.412 0.453 0.079
Chain 1: 1100 -7785.955 0.356 0.067
Chain 1: 1200 -7796.623 0.138 0.047
Chain 1: 1300 -7780.802 0.038 0.042
Chain 1: 1400 -7698.087 0.029 0.033
Chain 1: 1500 -7568.552 0.027 0.017
Chain 1: 1600 -7822.420 0.022 0.017
Chain 1: 1700 -7637.260 0.024 0.024
Chain 1: 1800 -7643.849 0.017 0.017
Chain 1: 1900 -7605.230 0.017 0.017
Chain 1: 2000 -7635.767 0.013 0.011
Chain 1: 2100 -7601.559 0.010 0.005 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002952 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.52 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86544.364 1.000 1.000
Chain 1: 200 -14074.153 3.075 5.149
Chain 1: 300 -10319.018 2.171 1.000
Chain 1: 400 -11664.319 1.657 1.000
Chain 1: 500 -9339.403 1.375 0.364
Chain 1: 600 -9127.635 1.150 0.364
Chain 1: 700 -9214.046 0.987 0.249
Chain 1: 800 -8577.620 0.873 0.249
Chain 1: 900 -8606.225 0.776 0.115
Chain 1: 1000 -9368.308 0.707 0.115
Chain 1: 1100 -8821.895 0.613 0.081 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -9230.266 0.103 0.074
Chain 1: 1300 -8634.074 0.073 0.069
Chain 1: 1400 -8811.494 0.064 0.062
Chain 1: 1500 -8703.292 0.040 0.044
Chain 1: 1600 -8708.421 0.038 0.044
Chain 1: 1700 -8590.441 0.038 0.044
Chain 1: 1800 -8645.802 0.031 0.020
Chain 1: 1900 -8525.395 0.032 0.020
Chain 1: 2000 -8596.362 0.025 0.014
Chain 1: 2100 -8582.634 0.019 0.014
Chain 1: 2200 -8537.675 0.015 0.012
Chain 1: 2300 -8696.304 0.010 0.012
Chain 1: 2400 -8515.003 0.010 0.012
Chain 1: 2500 -8589.696 0.010 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003194 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.94 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8362315.821 1.000 1.000
Chain 1: 200 -1577996.343 2.650 4.299
Chain 1: 300 -889782.151 2.024 1.000
Chain 1: 400 -457793.355 1.754 1.000
Chain 1: 500 -358699.072 1.459 0.944
Chain 1: 600 -233913.705 1.304 0.944
Chain 1: 700 -120036.236 1.254 0.944
Chain 1: 800 -87222.269 1.144 0.944
Chain 1: 900 -67537.659 1.049 0.773
Chain 1: 1000 -52309.632 0.973 0.773
Chain 1: 1100 -39753.693 0.905 0.533 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38934.600 0.477 0.376
Chain 1: 1300 -26841.764 0.445 0.376
Chain 1: 1400 -26560.214 0.352 0.316
Chain 1: 1500 -23133.888 0.339 0.316
Chain 1: 1600 -22347.657 0.289 0.291
Chain 1: 1700 -21214.423 0.199 0.291
Chain 1: 1800 -21157.611 0.162 0.148
Chain 1: 1900 -21484.464 0.134 0.053
Chain 1: 2000 -19990.744 0.113 0.053
Chain 1: 2100 -20229.472 0.082 0.035
Chain 1: 2200 -20456.960 0.081 0.035
Chain 1: 2300 -20073.069 0.038 0.019
Chain 1: 2400 -19844.825 0.038 0.019
Chain 1: 2500 -19647.079 0.024 0.015
Chain 1: 2600 -19276.357 0.023 0.015
Chain 1: 2700 -19233.078 0.018 0.012
Chain 1: 2800 -18949.703 0.019 0.015
Chain 1: 2900 -19231.388 0.019 0.015
Chain 1: 3000 -19217.452 0.012 0.012
Chain 1: 3100 -19302.542 0.011 0.012
Chain 1: 3200 -18992.745 0.011 0.015
Chain 1: 3300 -19197.881 0.010 0.012
Chain 1: 3400 -18671.986 0.012 0.015
Chain 1: 3500 -19285.166 0.014 0.015
Chain 1: 3600 -18590.194 0.016 0.015
Chain 1: 3700 -18978.255 0.018 0.016
Chain 1: 3800 -17935.423 0.022 0.020
Chain 1: 3900 -17931.541 0.021 0.020
Chain 1: 4000 -18048.820 0.021 0.020
Chain 1: 4100 -17962.437 0.021 0.020
Chain 1: 4200 -17778.149 0.021 0.020
Chain 1: 4300 -17916.907 0.021 0.020
Chain 1: 4400 -17873.278 0.018 0.010
Chain 1: 4500 -17775.743 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001354 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.54 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12361.863 1.000 1.000
Chain 1: 200 -9299.107 0.665 1.000
Chain 1: 300 -8033.654 0.496 0.329
Chain 1: 400 -8184.518 0.376 0.329
Chain 1: 500 -8117.032 0.303 0.158
Chain 1: 600 -8016.514 0.254 0.158
Chain 1: 700 -7927.925 0.220 0.018
Chain 1: 800 -7966.268 0.193 0.018
Chain 1: 900 -8090.198 0.173 0.015
Chain 1: 1000 -8019.220 0.157 0.015
Chain 1: 1100 -8023.730 0.057 0.013
Chain 1: 1200 -7965.431 0.024 0.011
Chain 1: 1300 -7905.722 0.009 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001764 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 17.64 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -63404.045 1.000 1.000
Chain 1: 200 -18173.978 1.744 2.489
Chain 1: 300 -8772.656 1.520 1.072
Chain 1: 400 -9213.677 1.152 1.072
Chain 1: 500 -8492.868 0.939 1.000
Chain 1: 600 -9235.565 0.796 1.000
Chain 1: 700 -8135.251 0.701 0.135
Chain 1: 800 -8103.659 0.614 0.135
Chain 1: 900 -8090.217 0.546 0.085
Chain 1: 1000 -7773.179 0.496 0.085
Chain 1: 1100 -7690.508 0.397 0.080
Chain 1: 1200 -7608.313 0.149 0.048
Chain 1: 1300 -7775.500 0.044 0.041
Chain 1: 1400 -7675.338 0.040 0.022
Chain 1: 1500 -7611.350 0.033 0.013
Chain 1: 1600 -7654.385 0.025 0.011
Chain 1: 1700 -7571.356 0.013 0.011
Chain 1: 1800 -7611.381 0.013 0.011
Chain 1: 1900 -7606.155 0.013 0.011
Chain 1: 2000 -7579.551 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003819 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 38.19 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86566.581 1.000 1.000
Chain 1: 200 -13460.231 3.216 5.431
Chain 1: 300 -9859.276 2.266 1.000
Chain 1: 400 -10704.354 1.719 1.000
Chain 1: 500 -8817.808 1.418 0.365
Chain 1: 600 -8682.826 1.184 0.365
Chain 1: 700 -8622.364 1.016 0.214
Chain 1: 800 -9318.872 0.898 0.214
Chain 1: 900 -8644.299 0.807 0.079
Chain 1: 1000 -8517.977 0.728 0.079
Chain 1: 1100 -8745.647 0.631 0.078 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8306.694 0.093 0.075
Chain 1: 1300 -8588.531 0.059 0.053
Chain 1: 1400 -8580.735 0.052 0.033
Chain 1: 1500 -8452.965 0.032 0.026
Chain 1: 1600 -8558.716 0.031 0.026
Chain 1: 1700 -8645.821 0.032 0.026
Chain 1: 1800 -8244.608 0.029 0.026
Chain 1: 1900 -8343.821 0.023 0.015
Chain 1: 2000 -8315.207 0.021 0.015
Chain 1: 2100 -8435.023 0.020 0.014
Chain 1: 2200 -8225.730 0.017 0.014
Chain 1: 2300 -8376.206 0.016 0.014
Chain 1: 2400 -8255.886 0.017 0.015
Chain 1: 2500 -8319.348 0.017 0.014
Chain 1: 2600 -8341.333 0.016 0.014
Chain 1: 2700 -8260.199 0.016 0.014
Chain 1: 2800 -8233.837 0.011 0.012
Chain 1: 2900 -8289.227 0.011 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00387 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 38.7 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8423959.157 1.000 1.000
Chain 1: 200 -1585879.762 2.656 4.312
Chain 1: 300 -890361.620 2.031 1.000
Chain 1: 400 -457486.963 1.760 1.000
Chain 1: 500 -357560.901 1.464 0.946
Chain 1: 600 -232556.363 1.309 0.946
Chain 1: 700 -118936.250 1.259 0.946
Chain 1: 800 -86212.778 1.149 0.946
Chain 1: 900 -66589.591 1.054 0.781
Chain 1: 1000 -51418.584 0.978 0.781
Chain 1: 1100 -38931.202 0.910 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38109.731 0.481 0.380
Chain 1: 1300 -26106.784 0.449 0.380
Chain 1: 1400 -25828.648 0.355 0.321
Chain 1: 1500 -22427.274 0.343 0.321
Chain 1: 1600 -21647.249 0.293 0.295
Chain 1: 1700 -20525.836 0.202 0.295
Chain 1: 1800 -20471.148 0.165 0.152
Chain 1: 1900 -20796.995 0.137 0.055
Chain 1: 2000 -19311.440 0.115 0.055
Chain 1: 2100 -19549.582 0.084 0.036
Chain 1: 2200 -19775.472 0.083 0.036
Chain 1: 2300 -19393.251 0.039 0.020
Chain 1: 2400 -19165.475 0.039 0.020
Chain 1: 2500 -18967.409 0.025 0.016
Chain 1: 2600 -18597.942 0.024 0.016
Chain 1: 2700 -18555.024 0.018 0.012
Chain 1: 2800 -18271.949 0.020 0.015
Chain 1: 2900 -18553.055 0.020 0.015
Chain 1: 3000 -18539.214 0.012 0.012
Chain 1: 3100 -18624.195 0.011 0.012
Chain 1: 3200 -18315.054 0.012 0.015
Chain 1: 3300 -18519.652 0.011 0.012
Chain 1: 3400 -17994.913 0.013 0.015
Chain 1: 3500 -18606.235 0.015 0.015
Chain 1: 3600 -17913.579 0.017 0.015
Chain 1: 3700 -18299.858 0.019 0.017
Chain 1: 3800 -17260.620 0.023 0.021
Chain 1: 3900 -17256.769 0.022 0.021
Chain 1: 4000 -17374.082 0.022 0.021
Chain 1: 4100 -17287.895 0.022 0.021
Chain 1: 4200 -17104.364 0.022 0.021
Chain 1: 4300 -17242.607 0.021 0.021
Chain 1: 4400 -17199.612 0.019 0.011
Chain 1: 4500 -17102.159 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001184 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 11.84 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12165.772 1.000 1.000
Chain 1: 200 -9144.524 0.665 1.000
Chain 1: 300 -8133.607 0.485 0.330
Chain 1: 400 -8181.877 0.365 0.330
Chain 1: 500 -8222.269 0.293 0.124
Chain 1: 600 -8118.832 0.246 0.124
Chain 1: 700 -7882.495 0.215 0.030
Chain 1: 800 -7911.754 0.189 0.030
Chain 1: 900 -7859.769 0.169 0.013
Chain 1: 1000 -7940.227 0.153 0.013
Chain 1: 1100 -7994.262 0.054 0.010
Chain 1: 1200 -7899.880 0.022 0.010
Chain 1: 1300 -7927.793 0.010 0.007 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001449 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.49 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -55708.786 1.000 1.000
Chain 1: 200 -16960.817 1.642 2.285
Chain 1: 300 -8603.430 1.419 1.000
Chain 1: 400 -9164.423 1.079 1.000
Chain 1: 500 -8626.322 0.876 0.971
Chain 1: 600 -8839.105 0.734 0.971
Chain 1: 700 -7824.771 0.648 0.130
Chain 1: 800 -8302.588 0.574 0.130
Chain 1: 900 -7989.969 0.514 0.062
Chain 1: 1000 -7772.233 0.466 0.062
Chain 1: 1100 -7730.547 0.366 0.061
Chain 1: 1200 -7657.972 0.139 0.058
Chain 1: 1300 -7745.214 0.043 0.039
Chain 1: 1400 -7886.260 0.038 0.028
Chain 1: 1500 -7625.508 0.036 0.028
Chain 1: 1600 -7635.380 0.033 0.028
Chain 1: 1700 -7544.751 0.022 0.018
Chain 1: 1800 -7606.614 0.017 0.012
Chain 1: 1900 -7658.971 0.013 0.011
Chain 1: 2000 -7642.506 0.011 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003316 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.16 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85833.198 1.000 1.000
Chain 1: 200 -13276.827 3.232 5.465
Chain 1: 300 -9751.871 2.275 1.000
Chain 1: 400 -10550.978 1.726 1.000
Chain 1: 500 -8671.363 1.424 0.361
Chain 1: 600 -8537.400 1.189 0.361
Chain 1: 700 -8538.634 1.019 0.217
Chain 1: 800 -8816.907 0.896 0.217
Chain 1: 900 -8635.742 0.799 0.076
Chain 1: 1000 -8370.795 0.722 0.076
Chain 1: 1100 -8664.567 0.625 0.034 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8235.387 0.084 0.034
Chain 1: 1300 -8518.515 0.051 0.033
Chain 1: 1400 -8497.106 0.044 0.032
Chain 1: 1500 -8395.046 0.023 0.032
Chain 1: 1600 -8488.831 0.023 0.032
Chain 1: 1700 -8589.723 0.024 0.032
Chain 1: 1800 -8197.433 0.026 0.032
Chain 1: 1900 -8298.670 0.025 0.032
Chain 1: 2000 -8268.709 0.022 0.012
Chain 1: 2100 -8407.233 0.020 0.012
Chain 1: 2200 -8188.881 0.018 0.012
Chain 1: 2300 -8330.964 0.016 0.012
Chain 1: 2400 -8341.189 0.016 0.012
Chain 1: 2500 -8307.169 0.015 0.012
Chain 1: 2600 -8303.692 0.014 0.012
Chain 1: 2700 -8214.046 0.014 0.012
Chain 1: 2800 -8194.009 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003196 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.96 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8400002.482 1.000 1.000
Chain 1: 200 -1584605.196 2.651 4.301
Chain 1: 300 -891052.393 2.026 1.000
Chain 1: 400 -457736.126 1.757 1.000
Chain 1: 500 -358054.940 1.461 0.947
Chain 1: 600 -232864.258 1.307 0.947
Chain 1: 700 -119009.303 1.257 0.947
Chain 1: 800 -86225.585 1.147 0.947
Chain 1: 900 -66548.840 1.053 0.778
Chain 1: 1000 -51330.530 0.977 0.778
Chain 1: 1100 -38802.813 0.909 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37972.661 0.481 0.380
Chain 1: 1300 -25932.721 0.450 0.380
Chain 1: 1400 -25649.540 0.357 0.323
Chain 1: 1500 -22239.112 0.344 0.323
Chain 1: 1600 -21455.524 0.294 0.296
Chain 1: 1700 -20330.188 0.204 0.296
Chain 1: 1800 -20274.231 0.166 0.153
Chain 1: 1900 -20599.760 0.138 0.055
Chain 1: 2000 -19112.721 0.116 0.055
Chain 1: 2100 -19350.834 0.085 0.037
Chain 1: 2200 -19576.945 0.084 0.037
Chain 1: 2300 -19194.633 0.040 0.020
Chain 1: 2400 -18966.959 0.040 0.020
Chain 1: 2500 -18769.101 0.025 0.016
Chain 1: 2600 -18399.852 0.024 0.016
Chain 1: 2700 -18356.962 0.019 0.012
Chain 1: 2800 -18074.186 0.020 0.016
Chain 1: 2900 -18355.120 0.020 0.015
Chain 1: 3000 -18341.292 0.012 0.012
Chain 1: 3100 -18426.215 0.011 0.012
Chain 1: 3200 -18117.302 0.012 0.015
Chain 1: 3300 -18321.697 0.011 0.012
Chain 1: 3400 -17797.422 0.013 0.015
Chain 1: 3500 -18408.146 0.015 0.016
Chain 1: 3600 -17716.304 0.017 0.016
Chain 1: 3700 -18102.042 0.019 0.017
Chain 1: 3800 -17064.087 0.023 0.021
Chain 1: 3900 -17060.303 0.022 0.021
Chain 1: 4000 -17177.579 0.022 0.021
Chain 1: 4100 -17091.506 0.022 0.021
Chain 1: 4200 -16908.224 0.022 0.021
Chain 1: 4300 -17046.246 0.021 0.021
Chain 1: 4400 -17003.481 0.019 0.011
Chain 1: 4500 -16906.114 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001379 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.79 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12463.278 1.000 1.000
Chain 1: 200 -9312.598 0.669 1.000
Chain 1: 300 -8031.434 0.499 0.338
Chain 1: 400 -8197.269 0.380 0.338
Chain 1: 500 -8075.637 0.307 0.160
Chain 1: 600 -7983.526 0.257 0.160
Chain 1: 700 -7881.742 0.223 0.020
Chain 1: 800 -7890.680 0.195 0.020
Chain 1: 900 -7782.936 0.175 0.015
Chain 1: 1000 -7949.099 0.159 0.020
Chain 1: 1100 -7975.888 0.060 0.015
Chain 1: 1200 -7896.753 0.027 0.014
Chain 1: 1300 -7852.581 0.011 0.013
Chain 1: 1400 -7880.900 0.010 0.012 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001483 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.83 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58026.200 1.000 1.000
Chain 1: 200 -17668.950 1.642 2.284
Chain 1: 300 -8679.188 1.440 1.036
Chain 1: 400 -8149.227 1.096 1.036
Chain 1: 500 -8356.549 0.882 1.000
Chain 1: 600 -9067.071 0.748 1.000
Chain 1: 700 -8248.558 0.655 0.099
Chain 1: 800 -8027.328 0.577 0.099
Chain 1: 900 -7639.505 0.518 0.078
Chain 1: 1000 -7715.373 0.468 0.078
Chain 1: 1100 -7682.668 0.368 0.065
Chain 1: 1200 -7579.772 0.141 0.051
Chain 1: 1300 -7593.349 0.038 0.028
Chain 1: 1400 -7720.844 0.033 0.025
Chain 1: 1500 -7611.193 0.032 0.017
Chain 1: 1600 -7673.732 0.025 0.014
Chain 1: 1700 -7515.394 0.017 0.014
Chain 1: 1800 -7537.905 0.014 0.014
Chain 1: 1900 -7573.586 0.010 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003415 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.15 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86245.271 1.000 1.000
Chain 1: 200 -13495.211 3.195 5.391
Chain 1: 300 -9862.949 2.253 1.000
Chain 1: 400 -10766.049 1.711 1.000
Chain 1: 500 -8840.242 1.412 0.368
Chain 1: 600 -8756.459 1.178 0.368
Chain 1: 700 -8436.665 1.015 0.218
Chain 1: 800 -9058.072 0.897 0.218
Chain 1: 900 -8676.565 0.802 0.084
Chain 1: 1000 -8488.765 0.724 0.084
Chain 1: 1100 -8684.330 0.627 0.069 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8221.744 0.093 0.056
Chain 1: 1300 -8541.379 0.060 0.044
Chain 1: 1400 -8554.921 0.052 0.038
Chain 1: 1500 -8428.341 0.031 0.037
Chain 1: 1600 -8536.514 0.032 0.037
Chain 1: 1700 -8620.574 0.029 0.023
Chain 1: 1800 -8206.838 0.027 0.023
Chain 1: 1900 -8303.050 0.024 0.022
Chain 1: 2000 -8276.453 0.022 0.015
Chain 1: 2100 -8399.266 0.021 0.015
Chain 1: 2200 -8219.243 0.018 0.015
Chain 1: 2300 -8297.924 0.015 0.013
Chain 1: 2400 -8367.651 0.016 0.013
Chain 1: 2500 -8313.145 0.015 0.012
Chain 1: 2600 -8312.698 0.014 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003256 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.56 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8411224.314 1.000 1.000
Chain 1: 200 -1584936.826 2.653 4.307
Chain 1: 300 -890603.339 2.029 1.000
Chain 1: 400 -457543.339 1.758 1.000
Chain 1: 500 -357799.998 1.462 0.946
Chain 1: 600 -232823.780 1.308 0.946
Chain 1: 700 -119115.569 1.258 0.946
Chain 1: 800 -86358.672 1.148 0.946
Chain 1: 900 -66717.266 1.053 0.780
Chain 1: 1000 -51532.203 0.977 0.780
Chain 1: 1100 -39027.458 0.909 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38204.910 0.481 0.379
Chain 1: 1300 -26178.926 0.449 0.379
Chain 1: 1400 -25899.513 0.355 0.320
Chain 1: 1500 -22491.494 0.342 0.320
Chain 1: 1600 -21709.537 0.292 0.295
Chain 1: 1700 -20585.269 0.202 0.294
Chain 1: 1800 -20529.886 0.165 0.152
Chain 1: 1900 -20855.998 0.137 0.055
Chain 1: 2000 -19368.296 0.115 0.055
Chain 1: 2100 -19606.625 0.084 0.036
Chain 1: 2200 -19832.915 0.083 0.036
Chain 1: 2300 -19450.241 0.039 0.020
Chain 1: 2400 -19222.358 0.039 0.020
Chain 1: 2500 -19024.328 0.025 0.016
Chain 1: 2600 -18654.639 0.024 0.016
Chain 1: 2700 -18611.608 0.018 0.012
Chain 1: 2800 -18328.506 0.020 0.015
Chain 1: 2900 -18609.668 0.019 0.015
Chain 1: 3000 -18595.888 0.012 0.012
Chain 1: 3100 -18680.891 0.011 0.012
Chain 1: 3200 -18371.591 0.012 0.015
Chain 1: 3300 -18576.278 0.011 0.012
Chain 1: 3400 -18051.284 0.013 0.015
Chain 1: 3500 -18663.035 0.015 0.015
Chain 1: 3600 -17969.810 0.017 0.015
Chain 1: 3700 -18356.547 0.019 0.017
Chain 1: 3800 -17316.434 0.023 0.021
Chain 1: 3900 -17312.555 0.021 0.021
Chain 1: 4000 -17429.875 0.022 0.021
Chain 1: 4100 -17343.669 0.022 0.021
Chain 1: 4200 -17159.922 0.022 0.021
Chain 1: 4300 -17298.326 0.021 0.021
Chain 1: 4400 -17255.178 0.019 0.011
Chain 1: 4500 -17157.689 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001263 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.63 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12505.422 1.000 1.000
Chain 1: 200 -9258.595 0.675 1.000
Chain 1: 300 -8071.303 0.499 0.351
Chain 1: 400 -8236.589 0.379 0.351
Chain 1: 500 -7864.011 0.313 0.147
Chain 1: 600 -7952.603 0.263 0.147
Chain 1: 700 -7880.775 0.226 0.047
Chain 1: 800 -7822.175 0.199 0.047
Chain 1: 900 -7845.481 0.177 0.020
Chain 1: 1000 -7878.039 0.160 0.020
Chain 1: 1100 -7901.298 0.060 0.011
Chain 1: 1200 -7888.662 0.025 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001598 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.98 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -59017.755 1.000 1.000
Chain 1: 200 -17962.570 1.643 2.286
Chain 1: 300 -8991.841 1.428 1.000
Chain 1: 400 -8394.451 1.089 1.000
Chain 1: 500 -8540.478 0.874 0.998
Chain 1: 600 -9104.928 0.739 0.998
Chain 1: 700 -8298.323 0.647 0.097
Chain 1: 800 -8202.408 0.568 0.097
Chain 1: 900 -8490.222 0.508 0.071
Chain 1: 1000 -8050.634 0.463 0.071
Chain 1: 1100 -7762.508 0.367 0.062
Chain 1: 1200 -7766.539 0.138 0.055
Chain 1: 1300 -7807.372 0.039 0.037
Chain 1: 1400 -8016.351 0.035 0.034
Chain 1: 1500 -7647.024 0.038 0.037
Chain 1: 1600 -7929.445 0.035 0.036
Chain 1: 1700 -7542.305 0.030 0.036
Chain 1: 1800 -7681.961 0.031 0.036
Chain 1: 1900 -7632.710 0.028 0.036
Chain 1: 2000 -7691.327 0.024 0.026
Chain 1: 2100 -7597.474 0.021 0.018
Chain 1: 2200 -7744.760 0.023 0.019
Chain 1: 2300 -7594.008 0.024 0.020
Chain 1: 2400 -7748.184 0.024 0.020
Chain 1: 2500 -7678.476 0.020 0.019
Chain 1: 2600 -7586.734 0.018 0.018
Chain 1: 2700 -7506.496 0.014 0.012
Chain 1: 2800 -7545.142 0.012 0.012
Chain 1: 2900 -7431.435 0.013 0.012
Chain 1: 3000 -7572.005 0.014 0.015
Chain 1: 3100 -7568.049 0.013 0.015
Chain 1: 3200 -7774.877 0.014 0.015
Chain 1: 3300 -7498.565 0.015 0.015
Chain 1: 3400 -7715.775 0.016 0.015
Chain 1: 3500 -7482.004 0.019 0.019
Chain 1: 3600 -7548.941 0.018 0.019
Chain 1: 3700 -7497.053 0.018 0.019
Chain 1: 3800 -7496.966 0.017 0.019
Chain 1: 3900 -7464.125 0.016 0.019
Chain 1: 4000 -7458.376 0.014 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00326 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.6 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86039.157 1.000 1.000
Chain 1: 200 -13739.867 3.131 5.262
Chain 1: 300 -9992.651 2.212 1.000
Chain 1: 400 -11629.834 1.694 1.000
Chain 1: 500 -8755.885 1.421 0.375
Chain 1: 600 -8429.965 1.191 0.375
Chain 1: 700 -8507.807 1.022 0.328
Chain 1: 800 -9296.577 0.905 0.328
Chain 1: 900 -8844.978 0.810 0.141
Chain 1: 1000 -8655.907 0.731 0.141
Chain 1: 1100 -8840.682 0.633 0.085 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8327.252 0.113 0.062
Chain 1: 1300 -8638.983 0.079 0.051
Chain 1: 1400 -8472.295 0.067 0.039
Chain 1: 1500 -8493.528 0.035 0.036
Chain 1: 1600 -8594.215 0.032 0.022
Chain 1: 1700 -8651.993 0.032 0.022
Chain 1: 1800 -8202.251 0.029 0.022
Chain 1: 1900 -8311.592 0.025 0.021
Chain 1: 2000 -8287.215 0.023 0.020
Chain 1: 2100 -8416.648 0.022 0.015
Chain 1: 2200 -8205.850 0.019 0.015
Chain 1: 2300 -8306.969 0.016 0.013
Chain 1: 2400 -8368.710 0.015 0.012
Chain 1: 2500 -8319.712 0.016 0.012
Chain 1: 2600 -8335.588 0.015 0.012
Chain 1: 2700 -8241.660 0.015 0.012
Chain 1: 2800 -8186.478 0.010 0.011
Chain 1: 2900 -8287.710 0.010 0.011
Chain 1: 3000 -8135.925 0.012 0.012
Chain 1: 3100 -8272.364 0.012 0.012
Chain 1: 3200 -8140.951 0.011 0.012
Chain 1: 3300 -8378.296 0.013 0.012
Chain 1: 3400 -8381.761 0.012 0.012
Chain 1: 3500 -8249.346 0.013 0.016
Chain 1: 3600 -8098.101 0.015 0.016
Chain 1: 3700 -8245.756 0.015 0.016
Chain 1: 3800 -8100.244 0.016 0.018
Chain 1: 3900 -8032.270 0.016 0.018
Chain 1: 4000 -8148.164 0.015 0.016
Chain 1: 4100 -8107.896 0.014 0.016
Chain 1: 4200 -8093.737 0.013 0.016
Chain 1: 4300 -8127.273 0.010 0.014
Chain 1: 4400 -8084.269 0.011 0.014
Chain 1: 4500 -8182.409 0.011 0.012
Chain 1: 4600 -8073.456 0.010 0.012
Chain 1: 4700 -8280.756 0.011 0.012
Chain 1: 4800 -8162.408 0.010 0.012
Chain 1: 4900 -8172.406 0.010 0.012 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003469 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.69 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8408373.298 1.000 1.000
Chain 1: 200 -1585558.089 2.652 4.303
Chain 1: 300 -891846.539 2.027 1.000
Chain 1: 400 -458392.401 1.757 1.000
Chain 1: 500 -358698.149 1.461 0.946
Chain 1: 600 -233380.123 1.307 0.946
Chain 1: 700 -119537.671 1.256 0.946
Chain 1: 800 -86762.988 1.146 0.946
Chain 1: 900 -67091.663 1.052 0.778
Chain 1: 1000 -51894.011 0.976 0.778
Chain 1: 1100 -39370.597 0.908 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38552.754 0.479 0.378
Chain 1: 1300 -26493.000 0.447 0.378
Chain 1: 1400 -26213.682 0.354 0.318
Chain 1: 1500 -22797.131 0.341 0.318
Chain 1: 1600 -22013.896 0.291 0.293
Chain 1: 1700 -20884.688 0.201 0.293
Chain 1: 1800 -20828.835 0.163 0.150
Chain 1: 1900 -21155.598 0.136 0.054
Chain 1: 2000 -19664.472 0.114 0.054
Chain 1: 2100 -19902.752 0.083 0.036
Chain 1: 2200 -20130.021 0.082 0.036
Chain 1: 2300 -19746.421 0.039 0.019
Chain 1: 2400 -19518.285 0.039 0.019
Chain 1: 2500 -19320.475 0.025 0.015
Chain 1: 2600 -18949.759 0.023 0.015
Chain 1: 2700 -18906.569 0.018 0.012
Chain 1: 2800 -18623.172 0.019 0.015
Chain 1: 2900 -18904.795 0.019 0.015
Chain 1: 3000 -18890.834 0.012 0.012
Chain 1: 3100 -18975.915 0.011 0.012
Chain 1: 3200 -18666.144 0.012 0.015
Chain 1: 3300 -18871.288 0.011 0.012
Chain 1: 3400 -18345.409 0.012 0.015
Chain 1: 3500 -18958.485 0.015 0.015
Chain 1: 3600 -18263.680 0.016 0.015
Chain 1: 3700 -18651.558 0.018 0.017
Chain 1: 3800 -17608.949 0.023 0.021
Chain 1: 3900 -17605.094 0.021 0.021
Chain 1: 4000 -17722.364 0.022 0.021
Chain 1: 4100 -17635.989 0.022 0.021
Chain 1: 4200 -17451.786 0.021 0.021
Chain 1: 4300 -17590.464 0.021 0.021
Chain 1: 4400 -17546.858 0.018 0.011
Chain 1: 4500 -17449.383 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001573 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.73 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12704.555 1.000 1.000
Chain 1: 200 -9611.983 0.661 1.000
Chain 1: 300 -8310.309 0.493 0.322
Chain 1: 400 -8492.972 0.375 0.322
Chain 1: 500 -8401.863 0.302 0.157
Chain 1: 600 -8252.785 0.255 0.157
Chain 1: 700 -8229.784 0.219 0.022
Chain 1: 800 -8179.230 0.192 0.022
Chain 1: 900 -8237.492 0.172 0.018
Chain 1: 1000 -8237.679 0.154 0.018
Chain 1: 1100 -8257.133 0.055 0.011
Chain 1: 1200 -8170.569 0.024 0.011
Chain 1: 1300 -8129.611 0.008 0.007 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001443 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.43 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58481.042 1.000 1.000
Chain 1: 200 -17974.845 1.627 2.253
Chain 1: 300 -8856.182 1.428 1.030
Chain 1: 400 -8245.985 1.089 1.030
Chain 1: 500 -8779.949 0.884 1.000
Chain 1: 600 -8906.250 0.739 1.000
Chain 1: 700 -8376.219 0.642 0.074
Chain 1: 800 -8314.526 0.563 0.074
Chain 1: 900 -8063.934 0.504 0.063
Chain 1: 1000 -8073.280 0.454 0.063
Chain 1: 1100 -7928.905 0.355 0.061
Chain 1: 1200 -7660.413 0.133 0.035
Chain 1: 1300 -7723.139 0.031 0.031
Chain 1: 1400 -7968.875 0.027 0.031
Chain 1: 1500 -7700.277 0.024 0.031
Chain 1: 1600 -7912.940 0.026 0.031
Chain 1: 1700 -7544.261 0.024 0.031
Chain 1: 1800 -7731.770 0.026 0.031
Chain 1: 1900 -7647.135 0.024 0.027
Chain 1: 2000 -7742.605 0.025 0.027
Chain 1: 2100 -7675.370 0.024 0.027
Chain 1: 2200 -7800.085 0.022 0.024
Chain 1: 2300 -7709.060 0.023 0.024
Chain 1: 2400 -7646.550 0.020 0.016
Chain 1: 2500 -7697.493 0.017 0.012
Chain 1: 2600 -7617.361 0.016 0.012
Chain 1: 2700 -7564.924 0.012 0.011
Chain 1: 2800 -7531.752 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003144 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.44 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86794.535 1.000 1.000
Chain 1: 200 -13791.564 3.147 5.293
Chain 1: 300 -10149.207 2.217 1.000
Chain 1: 400 -11069.035 1.684 1.000
Chain 1: 500 -9115.816 1.390 0.359
Chain 1: 600 -8616.916 1.168 0.359
Chain 1: 700 -8646.660 1.002 0.214
Chain 1: 800 -9403.554 0.886 0.214
Chain 1: 900 -8886.774 0.794 0.083
Chain 1: 1000 -8776.866 0.716 0.083
Chain 1: 1100 -9010.073 0.619 0.080 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8507.797 0.095 0.059
Chain 1: 1300 -8838.817 0.063 0.058
Chain 1: 1400 -8844.092 0.055 0.058
Chain 1: 1500 -8715.224 0.035 0.037
Chain 1: 1600 -8823.428 0.030 0.026
Chain 1: 1700 -8904.480 0.031 0.026
Chain 1: 1800 -8488.115 0.028 0.026
Chain 1: 1900 -8585.225 0.023 0.015
Chain 1: 2000 -8558.928 0.022 0.015
Chain 1: 2100 -8682.308 0.021 0.014
Chain 1: 2200 -8500.156 0.017 0.014
Chain 1: 2300 -8579.845 0.015 0.012
Chain 1: 2400 -8649.541 0.015 0.012
Chain 1: 2500 -8595.342 0.014 0.011
Chain 1: 2600 -8595.280 0.013 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004949 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 49.49 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8426745.839 1.000 1.000
Chain 1: 200 -1589714.112 2.650 4.301
Chain 1: 300 -890990.626 2.028 1.000
Chain 1: 400 -457701.989 1.758 1.000
Chain 1: 500 -357525.589 1.462 0.947
Chain 1: 600 -232527.574 1.308 0.947
Chain 1: 700 -119095.719 1.257 0.947
Chain 1: 800 -86412.328 1.148 0.947
Chain 1: 900 -66835.937 1.053 0.784
Chain 1: 1000 -51699.574 0.977 0.784
Chain 1: 1100 -39240.177 0.908 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38424.608 0.480 0.378
Chain 1: 1300 -26445.648 0.447 0.378
Chain 1: 1400 -26171.173 0.354 0.318
Chain 1: 1500 -22775.388 0.341 0.318
Chain 1: 1600 -21997.229 0.290 0.293
Chain 1: 1700 -20878.522 0.200 0.293
Chain 1: 1800 -20824.495 0.163 0.149
Chain 1: 1900 -21150.670 0.135 0.054
Chain 1: 2000 -19665.689 0.113 0.054
Chain 1: 2100 -19903.973 0.083 0.035
Chain 1: 2200 -20129.820 0.082 0.035
Chain 1: 2300 -19747.511 0.038 0.019
Chain 1: 2400 -19519.639 0.039 0.019
Chain 1: 2500 -19321.424 0.025 0.015
Chain 1: 2600 -18951.931 0.023 0.015
Chain 1: 2700 -18908.917 0.018 0.012
Chain 1: 2800 -18625.697 0.019 0.015
Chain 1: 2900 -18906.840 0.019 0.015
Chain 1: 3000 -18893.068 0.012 0.012
Chain 1: 3100 -18978.095 0.011 0.012
Chain 1: 3200 -18668.819 0.011 0.015
Chain 1: 3300 -18873.488 0.011 0.012
Chain 1: 3400 -18348.449 0.012 0.015
Chain 1: 3500 -18960.206 0.015 0.015
Chain 1: 3600 -18266.921 0.016 0.015
Chain 1: 3700 -18653.674 0.018 0.017
Chain 1: 3800 -17613.469 0.023 0.021
Chain 1: 3900 -17609.545 0.021 0.021
Chain 1: 4000 -17726.899 0.022 0.021
Chain 1: 4100 -17640.687 0.022 0.021
Chain 1: 4200 -17456.890 0.021 0.021
Chain 1: 4300 -17595.342 0.021 0.021
Chain 1: 4400 -17552.186 0.018 0.011
Chain 1: 4500 -17454.646 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001404 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.04 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12450.274 1.000 1.000
Chain 1: 200 -9372.294 0.664 1.000
Chain 1: 300 -8099.091 0.495 0.328
Chain 1: 400 -8303.158 0.378 0.328
Chain 1: 500 -8246.972 0.303 0.157
Chain 1: 600 -8066.146 0.257 0.157
Chain 1: 700 -7976.082 0.222 0.025
Chain 1: 800 -7984.902 0.194 0.025
Chain 1: 900 -7876.377 0.174 0.022
Chain 1: 1000 -8030.026 0.158 0.022
Chain 1: 1100 -8127.230 0.060 0.019
Chain 1: 1200 -7997.664 0.028 0.016
Chain 1: 1300 -7943.103 0.013 0.014
Chain 1: 1400 -7972.435 0.011 0.012
Chain 1: 1500 -8069.829 0.012 0.012
Chain 1: 1600 -8020.027 0.010 0.012
Chain 1: 1700 -7947.063 0.010 0.012
Chain 1: 1800 -7926.218 0.010 0.012
Chain 1: 1900 -7926.204 0.009 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001395 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.95 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -63626.508 1.000 1.000
Chain 1: 200 -18355.524 1.733 2.466
Chain 1: 300 -8863.776 1.512 1.071
Chain 1: 400 -8164.606 1.156 1.071
Chain 1: 500 -8456.404 0.931 1.000
Chain 1: 600 -9055.984 0.787 1.000
Chain 1: 700 -7997.366 0.694 0.132
Chain 1: 800 -8372.243 0.613 0.132
Chain 1: 900 -8208.034 0.547 0.086
Chain 1: 1000 -7789.731 0.497 0.086
Chain 1: 1100 -7765.971 0.398 0.066
Chain 1: 1200 -7676.955 0.152 0.054
Chain 1: 1300 -7829.478 0.047 0.045
Chain 1: 1400 -7915.656 0.040 0.035
Chain 1: 1500 -7686.555 0.039 0.030
Chain 1: 1600 -7866.508 0.035 0.023
Chain 1: 1700 -7605.679 0.025 0.023
Chain 1: 1800 -7663.318 0.021 0.020
Chain 1: 1900 -7675.490 0.019 0.019
Chain 1: 2000 -7683.374 0.014 0.012
Chain 1: 2100 -7670.654 0.014 0.012
Chain 1: 2200 -7772.900 0.014 0.013
Chain 1: 2300 -7670.337 0.014 0.013
Chain 1: 2400 -7725.100 0.013 0.013
Chain 1: 2500 -7634.440 0.011 0.012
Chain 1: 2600 -7603.919 0.010 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003197 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.97 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86535.010 1.000 1.000
Chain 1: 200 -13581.581 3.186 5.371
Chain 1: 300 -9951.087 2.245 1.000
Chain 1: 400 -10832.534 1.704 1.000
Chain 1: 500 -8923.332 1.406 0.365
Chain 1: 600 -8620.356 1.178 0.365
Chain 1: 700 -8663.264 1.010 0.214
Chain 1: 800 -9128.117 0.890 0.214
Chain 1: 900 -8824.436 0.795 0.081
Chain 1: 1000 -8596.465 0.718 0.081
Chain 1: 1100 -8803.447 0.621 0.051 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8445.482 0.088 0.042
Chain 1: 1300 -8640.689 0.054 0.035
Chain 1: 1400 -8653.685 0.046 0.034
Chain 1: 1500 -8517.749 0.026 0.027
Chain 1: 1600 -8629.348 0.024 0.024
Chain 1: 1700 -8713.307 0.024 0.024
Chain 1: 1800 -8301.465 0.024 0.024
Chain 1: 1900 -8397.398 0.022 0.023
Chain 1: 2000 -8370.544 0.019 0.016
Chain 1: 2100 -8492.855 0.018 0.014
Chain 1: 2200 -8312.540 0.016 0.014
Chain 1: 2300 -8392.541 0.015 0.013
Chain 1: 2400 -8462.125 0.016 0.013
Chain 1: 2500 -8407.428 0.015 0.011
Chain 1: 2600 -8406.678 0.013 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003464 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.64 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8431499.442 1.000 1.000
Chain 1: 200 -1590329.309 2.651 4.302
Chain 1: 300 -892568.231 2.028 1.000
Chain 1: 400 -458035.887 1.758 1.000
Chain 1: 500 -357772.593 1.462 0.949
Chain 1: 600 -232605.610 1.308 0.949
Chain 1: 700 -119046.751 1.258 0.949
Chain 1: 800 -86285.516 1.148 0.949
Chain 1: 900 -66685.120 1.053 0.782
Chain 1: 1000 -51531.516 0.977 0.782
Chain 1: 1100 -39052.026 0.909 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38236.313 0.481 0.380
Chain 1: 1300 -26241.532 0.449 0.380
Chain 1: 1400 -25964.783 0.355 0.320
Chain 1: 1500 -22564.216 0.342 0.320
Chain 1: 1600 -21784.422 0.292 0.294
Chain 1: 1700 -20664.247 0.202 0.294
Chain 1: 1800 -20609.875 0.164 0.151
Chain 1: 1900 -20935.955 0.136 0.054
Chain 1: 2000 -19450.497 0.114 0.054
Chain 1: 2100 -19688.649 0.084 0.036
Chain 1: 2200 -19914.453 0.083 0.036
Chain 1: 2300 -19532.308 0.039 0.020
Chain 1: 2400 -19304.528 0.039 0.020
Chain 1: 2500 -19106.221 0.025 0.016
Chain 1: 2600 -18736.678 0.023 0.016
Chain 1: 2700 -18693.847 0.018 0.012
Chain 1: 2800 -18410.506 0.019 0.015
Chain 1: 2900 -18691.761 0.019 0.015
Chain 1: 3000 -18678.019 0.012 0.012
Chain 1: 3100 -18762.920 0.011 0.012
Chain 1: 3200 -18453.713 0.012 0.015
Chain 1: 3300 -18658.426 0.011 0.012
Chain 1: 3400 -18133.338 0.012 0.015
Chain 1: 3500 -18745.011 0.015 0.015
Chain 1: 3600 -18052.091 0.017 0.015
Chain 1: 3700 -18438.487 0.018 0.017
Chain 1: 3800 -17398.619 0.023 0.021
Chain 1: 3900 -17394.772 0.021 0.021
Chain 1: 4000 -17512.126 0.022 0.021
Chain 1: 4100 -17425.787 0.022 0.021
Chain 1: 4200 -17242.244 0.021 0.021
Chain 1: 4300 -17380.537 0.021 0.021
Chain 1: 4400 -17337.447 0.018 0.011
Chain 1: 4500 -17240.004 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001422 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.22 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12854.567 1.000 1.000
Chain 1: 200 -9754.506 0.659 1.000
Chain 1: 300 -8322.125 0.497 0.318
Chain 1: 400 -8545.387 0.379 0.318
Chain 1: 500 -8417.300 0.306 0.172
Chain 1: 600 -8268.341 0.258 0.172
Chain 1: 700 -8354.311 0.223 0.026
Chain 1: 800 -8223.209 0.197 0.026
Chain 1: 900 -8293.130 0.176 0.018
Chain 1: 1000 -8233.026 0.159 0.018
Chain 1: 1100 -8311.300 0.060 0.016
Chain 1: 1200 -8180.467 0.030 0.016
Chain 1: 1300 -8124.962 0.013 0.015
Chain 1: 1400 -8159.610 0.011 0.010
Chain 1: 1500 -8249.082 0.011 0.010
Chain 1: 1600 -8181.810 0.010 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001435 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.35 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57517.657 1.000 1.000
Chain 1: 200 -17922.003 1.605 2.209
Chain 1: 300 -9010.198 1.399 1.000
Chain 1: 400 -8381.807 1.068 1.000
Chain 1: 500 -9340.377 0.875 0.989
Chain 1: 600 -8634.177 0.743 0.989
Chain 1: 700 -8068.426 0.647 0.103
Chain 1: 800 -8447.693 0.572 0.103
Chain 1: 900 -8001.571 0.514 0.082
Chain 1: 1000 -8180.777 0.465 0.082
Chain 1: 1100 -7873.381 0.369 0.075
Chain 1: 1200 -7739.271 0.150 0.070
Chain 1: 1300 -7806.712 0.052 0.056
Chain 1: 1400 -7900.817 0.045 0.045
Chain 1: 1500 -7689.251 0.038 0.039
Chain 1: 1600 -7840.756 0.032 0.028
Chain 1: 1700 -7656.493 0.027 0.024
Chain 1: 1800 -7681.007 0.023 0.022
Chain 1: 1900 -7678.774 0.017 0.019
Chain 1: 2000 -7834.509 0.017 0.019
Chain 1: 2100 -7689.284 0.015 0.019
Chain 1: 2200 -7824.146 0.015 0.019
Chain 1: 2300 -7651.772 0.016 0.019
Chain 1: 2400 -7741.530 0.016 0.019
Chain 1: 2500 -7494.672 0.017 0.019
Chain 1: 2600 -7614.773 0.017 0.019
Chain 1: 2700 -7608.300 0.014 0.017
Chain 1: 2800 -7591.183 0.014 0.017
Chain 1: 2900 -7459.660 0.016 0.018
Chain 1: 3000 -7619.306 0.016 0.018
Chain 1: 3100 -7608.328 0.014 0.017
Chain 1: 3200 -7815.877 0.015 0.018
Chain 1: 3300 -7527.733 0.017 0.018
Chain 1: 3400 -7767.441 0.019 0.021
Chain 1: 3500 -7514.055 0.019 0.021
Chain 1: 3600 -7581.517 0.018 0.021
Chain 1: 3700 -7530.739 0.019 0.021
Chain 1: 3800 -7530.881 0.019 0.021
Chain 1: 3900 -7492.426 0.017 0.021
Chain 1: 4000 -7484.066 0.015 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002995 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.95 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87478.036 1.000 1.000
Chain 1: 200 -13956.744 3.134 5.268
Chain 1: 300 -10241.431 2.210 1.000
Chain 1: 400 -11602.373 1.687 1.000
Chain 1: 500 -9205.899 1.402 0.363
Chain 1: 600 -8633.070 1.179 0.363
Chain 1: 700 -9048.448 1.017 0.260
Chain 1: 800 -9351.288 0.894 0.260
Chain 1: 900 -9062.948 0.798 0.117
Chain 1: 1000 -8929.637 0.720 0.117
Chain 1: 1100 -8991.447 0.621 0.066 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8514.782 0.099 0.056
Chain 1: 1300 -8827.295 0.067 0.046
Chain 1: 1400 -8833.783 0.055 0.035
Chain 1: 1500 -8768.464 0.030 0.032
Chain 1: 1600 -8876.551 0.024 0.032
Chain 1: 1700 -8937.612 0.020 0.015
Chain 1: 1800 -8502.965 0.022 0.015
Chain 1: 1900 -8606.763 0.020 0.012
Chain 1: 2000 -8582.226 0.019 0.012
Chain 1: 2100 -8719.360 0.020 0.012
Chain 1: 2200 -8513.469 0.017 0.012
Chain 1: 2300 -8671.476 0.015 0.012
Chain 1: 2400 -8510.428 0.017 0.016
Chain 1: 2500 -8580.813 0.017 0.016
Chain 1: 2600 -8493.070 0.017 0.016
Chain 1: 2700 -8526.730 0.017 0.016
Chain 1: 2800 -8487.092 0.012 0.012
Chain 1: 2900 -8580.047 0.012 0.011
Chain 1: 3000 -8411.275 0.014 0.016
Chain 1: 3100 -8569.434 0.014 0.018
Chain 1: 3200 -8441.732 0.013 0.015
Chain 1: 3300 -8449.514 0.011 0.011
Chain 1: 3400 -8607.107 0.011 0.011
Chain 1: 3500 -8611.212 0.010 0.011
Chain 1: 3600 -8398.406 0.012 0.015
Chain 1: 3700 -8543.581 0.013 0.017
Chain 1: 3800 -8405.034 0.014 0.017
Chain 1: 3900 -8339.773 0.014 0.017
Chain 1: 4000 -8414.671 0.013 0.016
Chain 1: 4100 -8405.645 0.011 0.015
Chain 1: 4200 -8395.191 0.010 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003299 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.99 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8425596.264 1.000 1.000
Chain 1: 200 -1588602.763 2.652 4.304
Chain 1: 300 -891630.848 2.028 1.000
Chain 1: 400 -458011.011 1.758 1.000
Chain 1: 500 -357920.246 1.462 0.947
Chain 1: 600 -232935.865 1.308 0.947
Chain 1: 700 -119413.762 1.257 0.947
Chain 1: 800 -86690.744 1.147 0.947
Chain 1: 900 -67094.362 1.052 0.782
Chain 1: 1000 -51942.517 0.976 0.782
Chain 1: 1100 -39463.284 0.908 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38649.878 0.479 0.377
Chain 1: 1300 -26643.686 0.446 0.377
Chain 1: 1400 -26368.647 0.353 0.316
Chain 1: 1500 -22964.859 0.340 0.316
Chain 1: 1600 -22184.827 0.289 0.292
Chain 1: 1700 -21062.473 0.200 0.292
Chain 1: 1800 -21007.860 0.162 0.148
Chain 1: 1900 -21334.404 0.134 0.053
Chain 1: 2000 -19846.875 0.113 0.053
Chain 1: 2100 -20085.304 0.082 0.035
Chain 1: 2200 -20311.625 0.081 0.035
Chain 1: 2300 -19928.823 0.038 0.019
Chain 1: 2400 -19700.809 0.038 0.019
Chain 1: 2500 -19502.621 0.025 0.015
Chain 1: 2600 -19132.633 0.023 0.015
Chain 1: 2700 -19089.555 0.018 0.012
Chain 1: 2800 -18806.147 0.019 0.015
Chain 1: 2900 -19087.519 0.019 0.015
Chain 1: 3000 -19073.765 0.012 0.012
Chain 1: 3100 -19158.793 0.011 0.012
Chain 1: 3200 -18849.255 0.011 0.015
Chain 1: 3300 -19054.162 0.011 0.012
Chain 1: 3400 -18528.623 0.012 0.015
Chain 1: 3500 -19141.093 0.014 0.015
Chain 1: 3600 -18446.979 0.016 0.015
Chain 1: 3700 -18834.328 0.018 0.016
Chain 1: 3800 -17792.749 0.022 0.021
Chain 1: 3900 -17788.814 0.021 0.021
Chain 1: 4000 -17906.172 0.022 0.021
Chain 1: 4100 -17819.834 0.022 0.021
Chain 1: 4200 -17635.806 0.021 0.021
Chain 1: 4300 -17774.434 0.021 0.021
Chain 1: 4400 -17731.031 0.018 0.010
Chain 1: 4500 -17633.474 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001393 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.93 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49574.137 1.000 1.000
Chain 1: 200 -20270.649 1.223 1.446
Chain 1: 300 -16342.016 0.895 1.000
Chain 1: 400 -18632.662 0.702 1.000
Chain 1: 500 -13566.245 0.636 0.373
Chain 1: 600 -15279.525 0.549 0.373
Chain 1: 700 -14886.565 0.474 0.240
Chain 1: 800 -13541.448 0.428 0.240
Chain 1: 900 -21314.638 0.421 0.240
Chain 1: 1000 -10934.859 0.473 0.365
Chain 1: 1100 -11359.772 0.377 0.240
Chain 1: 1200 -10831.194 0.237 0.123
Chain 1: 1300 -12294.193 0.225 0.119
Chain 1: 1400 -18133.683 0.245 0.119
Chain 1: 1500 -10086.613 0.288 0.119
Chain 1: 1600 -9698.239 0.280 0.119
Chain 1: 1700 -10009.903 0.281 0.119
Chain 1: 1800 -10802.718 0.278 0.119
Chain 1: 1900 -18169.797 0.282 0.119
Chain 1: 2000 -10372.929 0.263 0.119
Chain 1: 2100 -10051.935 0.262 0.119
Chain 1: 2200 -11181.225 0.267 0.119
Chain 1: 2300 -11859.755 0.261 0.101
Chain 1: 2400 -9624.349 0.252 0.101
Chain 1: 2500 -9280.273 0.176 0.073
Chain 1: 2600 -11180.896 0.189 0.101
Chain 1: 2700 -12072.337 0.193 0.101
Chain 1: 2800 -9229.306 0.217 0.170
Chain 1: 2900 -9455.938 0.179 0.101
Chain 1: 3000 -11711.891 0.123 0.101
Chain 1: 3100 -10537.914 0.131 0.111
Chain 1: 3200 -9695.749 0.129 0.111
Chain 1: 3300 -10841.766 0.134 0.111
Chain 1: 3400 -9483.206 0.125 0.111
Chain 1: 3500 -9228.512 0.124 0.111
Chain 1: 3600 -17657.703 0.155 0.111
Chain 1: 3700 -10542.743 0.215 0.143
Chain 1: 3800 -8872.642 0.203 0.143
Chain 1: 3900 -13080.779 0.233 0.188
Chain 1: 4000 -9581.848 0.250 0.188
Chain 1: 4100 -9453.133 0.240 0.188
Chain 1: 4200 -8843.075 0.239 0.188
Chain 1: 4300 -10144.092 0.241 0.188
Chain 1: 4400 -14833.688 0.258 0.316
Chain 1: 4500 -9141.660 0.318 0.322
Chain 1: 4600 -13848.588 0.304 0.322
Chain 1: 4700 -9594.389 0.281 0.322
Chain 1: 4800 -8640.790 0.273 0.322
Chain 1: 4900 -9471.766 0.250 0.316
Chain 1: 5000 -9818.791 0.217 0.128
Chain 1: 5100 -8851.233 0.226 0.128
Chain 1: 5200 -8582.561 0.222 0.128
Chain 1: 5300 -8600.410 0.210 0.110
Chain 1: 5400 -8880.558 0.181 0.109
Chain 1: 5500 -12682.567 0.149 0.109
Chain 1: 5600 -9237.221 0.152 0.109
Chain 1: 5700 -14203.595 0.143 0.109
Chain 1: 5800 -9168.139 0.187 0.109
Chain 1: 5900 -8751.865 0.183 0.109
Chain 1: 6000 -9898.858 0.191 0.116
Chain 1: 6100 -8816.201 0.192 0.123
Chain 1: 6200 -8853.061 0.190 0.123
Chain 1: 6300 -8625.562 0.192 0.123
Chain 1: 6400 -9008.053 0.193 0.123
Chain 1: 6500 -10479.335 0.177 0.123
Chain 1: 6600 -9105.514 0.155 0.123
Chain 1: 6700 -8720.870 0.124 0.116
Chain 1: 6800 -12361.295 0.099 0.116
Chain 1: 6900 -12698.758 0.097 0.116
Chain 1: 7000 -11800.228 0.093 0.076
Chain 1: 7100 -9593.001 0.104 0.076
Chain 1: 7200 -9277.921 0.107 0.076
Chain 1: 7300 -11587.073 0.124 0.140
Chain 1: 7400 -12064.295 0.124 0.140
Chain 1: 7500 -11676.441 0.113 0.076
Chain 1: 7600 -8934.364 0.128 0.076
Chain 1: 7700 -14825.138 0.164 0.199
Chain 1: 7800 -10919.564 0.170 0.199
Chain 1: 7900 -9523.937 0.182 0.199
Chain 1: 8000 -9762.903 0.177 0.199
Chain 1: 8100 -8554.156 0.168 0.147
Chain 1: 8200 -8668.687 0.166 0.147
Chain 1: 8300 -8388.538 0.149 0.141
Chain 1: 8400 -8713.388 0.149 0.141
Chain 1: 8500 -8206.375 0.152 0.141
Chain 1: 8600 -8766.648 0.128 0.064
Chain 1: 8700 -8291.451 0.094 0.062
Chain 1: 8800 -9095.124 0.067 0.062
Chain 1: 8900 -10753.420 0.068 0.062
Chain 1: 9000 -10515.088 0.067 0.062
Chain 1: 9100 -8431.845 0.078 0.062
Chain 1: 9200 -8863.506 0.081 0.062
Chain 1: 9300 -11604.029 0.102 0.064
Chain 1: 9400 -8815.205 0.130 0.088
Chain 1: 9500 -10098.523 0.136 0.127
Chain 1: 9600 -8563.586 0.148 0.154
Chain 1: 9700 -9149.486 0.148 0.154
Chain 1: 9800 -8256.437 0.150 0.154
Chain 1: 9900 -9755.654 0.150 0.154
Chain 1: 10000 -8503.544 0.163 0.154
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002167 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 21.67 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57235.231 1.000 1.000
Chain 1: 200 -17805.607 1.607 2.214
Chain 1: 300 -8768.702 1.415 1.031
Chain 1: 400 -7975.950 1.086 1.031
Chain 1: 500 -8691.246 0.885 1.000
Chain 1: 600 -7896.112 0.755 1.000
Chain 1: 700 -8612.673 0.659 0.101
Chain 1: 800 -8065.328 0.585 0.101
Chain 1: 900 -7908.639 0.522 0.099
Chain 1: 1000 -7870.263 0.470 0.099
Chain 1: 1100 -7653.505 0.373 0.083
Chain 1: 1200 -7559.995 0.153 0.082
Chain 1: 1300 -7739.045 0.052 0.068
Chain 1: 1400 -7749.213 0.042 0.028
Chain 1: 1500 -7514.957 0.037 0.028
Chain 1: 1600 -7733.079 0.030 0.028
Chain 1: 1700 -7534.754 0.024 0.026
Chain 1: 1800 -7624.843 0.019 0.023
Chain 1: 1900 -7574.534 0.017 0.023
Chain 1: 2000 -7606.025 0.017 0.023
Chain 1: 2100 -7512.200 0.016 0.012
Chain 1: 2200 -7678.360 0.017 0.022
Chain 1: 2300 -7504.149 0.017 0.022
Chain 1: 2400 -7623.607 0.018 0.022
Chain 1: 2500 -7576.335 0.016 0.016
Chain 1: 2600 -7475.546 0.014 0.013
Chain 1: 2700 -7464.735 0.012 0.012
Chain 1: 2800 -7461.297 0.011 0.012
Chain 1: 2900 -7318.390 0.012 0.013
Chain 1: 3000 -7468.566 0.013 0.016
Chain 1: 3100 -7468.341 0.012 0.016
Chain 1: 3200 -7688.327 0.013 0.016
Chain 1: 3300 -7401.384 0.014 0.016
Chain 1: 3400 -7641.669 0.016 0.020
Chain 1: 3500 -7379.640 0.019 0.020
Chain 1: 3600 -7446.364 0.018 0.020
Chain 1: 3700 -7396.124 0.019 0.020
Chain 1: 3800 -7397.720 0.019 0.020
Chain 1: 3900 -7354.316 0.018 0.020
Chain 1: 4000 -7347.074 0.016 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003368 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.68 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85332.876 1.000 1.000
Chain 1: 200 -13855.045 3.079 5.159
Chain 1: 300 -10140.226 2.175 1.000
Chain 1: 400 -11271.280 1.656 1.000
Chain 1: 500 -9072.376 1.374 0.366
Chain 1: 600 -9494.819 1.152 0.366
Chain 1: 700 -9068.248 0.994 0.242
Chain 1: 800 -8503.870 0.878 0.242
Chain 1: 900 -8461.063 0.781 0.100
Chain 1: 1000 -8683.411 0.706 0.100
Chain 1: 1100 -8797.247 0.607 0.066 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8482.015 0.095 0.047
Chain 1: 1300 -8786.815 0.062 0.044
Chain 1: 1400 -8691.528 0.053 0.037
Chain 1: 1500 -8623.500 0.029 0.035
Chain 1: 1600 -8732.289 0.026 0.026
Chain 1: 1700 -8788.533 0.022 0.013
Chain 1: 1800 -8344.238 0.021 0.013
Chain 1: 1900 -8449.690 0.021 0.013
Chain 1: 2000 -8434.028 0.019 0.012
Chain 1: 2100 -8556.522 0.019 0.012
Chain 1: 2200 -8345.202 0.018 0.012
Chain 1: 2300 -8444.918 0.016 0.012
Chain 1: 2400 -8508.768 0.015 0.012
Chain 1: 2500 -8458.645 0.015 0.012
Chain 1: 2600 -8472.002 0.014 0.012
Chain 1: 2700 -8378.959 0.015 0.012
Chain 1: 2800 -8325.129 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003244 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.44 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8369137.740 1.000 1.000
Chain 1: 200 -1578823.485 2.650 4.301
Chain 1: 300 -890001.552 2.025 1.000
Chain 1: 400 -457694.089 1.755 1.000
Chain 1: 500 -358625.877 1.459 0.945
Chain 1: 600 -233796.366 1.305 0.945
Chain 1: 700 -119864.189 1.254 0.945
Chain 1: 800 -87050.570 1.145 0.945
Chain 1: 900 -67346.426 1.050 0.774
Chain 1: 1000 -52115.650 0.974 0.774
Chain 1: 1100 -39554.325 0.906 0.534 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38731.943 0.478 0.377
Chain 1: 1300 -26634.386 0.446 0.377
Chain 1: 1400 -26351.903 0.353 0.318
Chain 1: 1500 -22924.584 0.340 0.318
Chain 1: 1600 -22138.102 0.290 0.293
Chain 1: 1700 -21004.276 0.200 0.292
Chain 1: 1800 -20947.350 0.163 0.150
Chain 1: 1900 -21274.047 0.135 0.054
Chain 1: 2000 -19780.313 0.114 0.054
Chain 1: 2100 -20018.981 0.083 0.036
Chain 1: 2200 -20246.474 0.082 0.036
Chain 1: 2300 -19862.637 0.039 0.019
Chain 1: 2400 -19634.414 0.039 0.019
Chain 1: 2500 -19436.743 0.025 0.015
Chain 1: 2600 -19066.067 0.023 0.015
Chain 1: 2700 -19022.786 0.018 0.012
Chain 1: 2800 -18739.462 0.019 0.015
Chain 1: 2900 -19021.089 0.019 0.015
Chain 1: 3000 -19007.180 0.012 0.012
Chain 1: 3100 -19092.273 0.011 0.012
Chain 1: 3200 -18782.508 0.011 0.015
Chain 1: 3300 -18987.601 0.011 0.012
Chain 1: 3400 -18461.818 0.012 0.015
Chain 1: 3500 -19074.869 0.014 0.015
Chain 1: 3600 -18380.045 0.016 0.015
Chain 1: 3700 -18767.977 0.018 0.016
Chain 1: 3800 -17725.464 0.023 0.021
Chain 1: 3900 -17721.602 0.021 0.021
Chain 1: 4000 -17838.859 0.022 0.021
Chain 1: 4100 -17752.506 0.022 0.021
Chain 1: 4200 -17568.293 0.021 0.021
Chain 1: 4300 -17706.997 0.021 0.021
Chain 1: 4400 -17663.399 0.018 0.010
Chain 1: 4500 -17565.905 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001339 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.39 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12611.206 1.000 1.000
Chain 1: 200 -9447.306 0.667 1.000
Chain 1: 300 -8237.118 0.494 0.335
Chain 1: 400 -8421.987 0.376 0.335
Chain 1: 500 -8438.037 0.301 0.147
Chain 1: 600 -8147.074 0.257 0.147
Chain 1: 700 -8076.112 0.221 0.036
Chain 1: 800 -8086.379 0.194 0.036
Chain 1: 900 -8016.696 0.173 0.022
Chain 1: 1000 -8196.150 0.158 0.022
Chain 1: 1100 -8214.177 0.058 0.022
Chain 1: 1200 -8096.290 0.026 0.015
Chain 1: 1300 -8055.354 0.012 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001917 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 19.17 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57872.100 1.000 1.000
Chain 1: 200 -17641.480 1.640 2.280
Chain 1: 300 -8829.302 1.426 1.000
Chain 1: 400 -8211.176 1.088 1.000
Chain 1: 500 -8124.815 0.873 0.998
Chain 1: 600 -8623.598 0.737 0.998
Chain 1: 700 -8334.829 0.637 0.075
Chain 1: 800 -8127.619 0.560 0.075
Chain 1: 900 -8042.027 0.499 0.058
Chain 1: 1000 -7870.806 0.451 0.058
Chain 1: 1100 -7665.852 0.354 0.035
Chain 1: 1200 -7675.692 0.126 0.027
Chain 1: 1300 -7606.123 0.027 0.025
Chain 1: 1400 -7852.763 0.023 0.025
Chain 1: 1500 -7665.955 0.024 0.025
Chain 1: 1600 -7846.185 0.021 0.024
Chain 1: 1700 -7586.973 0.021 0.024
Chain 1: 1800 -7677.431 0.019 0.023
Chain 1: 1900 -7610.529 0.019 0.023
Chain 1: 2000 -7661.952 0.018 0.023
Chain 1: 2100 -7588.494 0.016 0.012
Chain 1: 2200 -7723.528 0.018 0.017
Chain 1: 2300 -7571.367 0.019 0.020
Chain 1: 2400 -7602.323 0.016 0.017
Chain 1: 2500 -7611.366 0.014 0.012
Chain 1: 2600 -7532.034 0.012 0.011
Chain 1: 2700 -7549.186 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003233 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.33 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86022.062 1.000 1.000
Chain 1: 200 -13738.606 3.131 5.261
Chain 1: 300 -10095.713 2.207 1.000
Chain 1: 400 -11081.762 1.678 1.000
Chain 1: 500 -8841.662 1.393 0.361
Chain 1: 600 -8994.518 1.164 0.361
Chain 1: 700 -8976.381 0.998 0.253
Chain 1: 800 -8489.705 0.880 0.253
Chain 1: 900 -8542.699 0.783 0.089
Chain 1: 1000 -8719.948 0.707 0.089
Chain 1: 1100 -8915.848 0.609 0.057 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8523.344 0.087 0.046
Chain 1: 1300 -8853.356 0.055 0.037
Chain 1: 1400 -8738.882 0.047 0.022
Chain 1: 1500 -8641.780 0.023 0.020
Chain 1: 1600 -8744.738 0.023 0.020
Chain 1: 1700 -8831.650 0.024 0.020
Chain 1: 1800 -8408.738 0.023 0.020
Chain 1: 1900 -8508.990 0.023 0.020
Chain 1: 2000 -8483.142 0.022 0.013
Chain 1: 2100 -8607.938 0.021 0.013
Chain 1: 2200 -8414.471 0.019 0.013
Chain 1: 2300 -8503.514 0.016 0.012
Chain 1: 2400 -8572.648 0.015 0.012
Chain 1: 2500 -8518.820 0.015 0.012
Chain 1: 2600 -8519.736 0.014 0.010
Chain 1: 2700 -8436.651 0.014 0.010
Chain 1: 2800 -8397.209 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004608 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 46.08 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8386589.940 1.000 1.000
Chain 1: 200 -1583824.564 2.648 4.295
Chain 1: 300 -890828.489 2.024 1.000
Chain 1: 400 -457612.793 1.755 1.000
Chain 1: 500 -358261.910 1.459 0.947
Chain 1: 600 -233228.579 1.306 0.947
Chain 1: 700 -119466.238 1.255 0.947
Chain 1: 800 -86697.246 1.145 0.947
Chain 1: 900 -67039.717 1.051 0.778
Chain 1: 1000 -51842.272 0.975 0.778
Chain 1: 1100 -39318.578 0.907 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38496.688 0.479 0.378
Chain 1: 1300 -26446.894 0.447 0.378
Chain 1: 1400 -26166.151 0.354 0.319
Chain 1: 1500 -22751.863 0.341 0.319
Chain 1: 1600 -21968.231 0.291 0.293
Chain 1: 1700 -20840.934 0.201 0.293
Chain 1: 1800 -20785.095 0.164 0.150
Chain 1: 1900 -21111.346 0.136 0.054
Chain 1: 2000 -19621.820 0.114 0.054
Chain 1: 2100 -19860.224 0.083 0.036
Chain 1: 2200 -20086.868 0.082 0.036
Chain 1: 2300 -19703.871 0.039 0.019
Chain 1: 2400 -19475.916 0.039 0.019
Chain 1: 2500 -19278.006 0.025 0.015
Chain 1: 2600 -18908.073 0.023 0.015
Chain 1: 2700 -18864.992 0.018 0.012
Chain 1: 2800 -18581.836 0.019 0.015
Chain 1: 2900 -18863.137 0.019 0.015
Chain 1: 3000 -18849.289 0.012 0.012
Chain 1: 3100 -18934.313 0.011 0.012
Chain 1: 3200 -18624.927 0.012 0.015
Chain 1: 3300 -18829.704 0.011 0.012
Chain 1: 3400 -18304.536 0.012 0.015
Chain 1: 3500 -18916.604 0.015 0.015
Chain 1: 3600 -18223.021 0.016 0.015
Chain 1: 3700 -18610.022 0.018 0.017
Chain 1: 3800 -17569.384 0.023 0.021
Chain 1: 3900 -17565.530 0.021 0.021
Chain 1: 4000 -17682.813 0.022 0.021
Chain 1: 4100 -17596.577 0.022 0.021
Chain 1: 4200 -17412.748 0.021 0.021
Chain 1: 4300 -17551.188 0.021 0.021
Chain 1: 4400 -17507.935 0.018 0.011
Chain 1: 4500 -17410.480 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001574 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.74 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49002.576 1.000 1.000
Chain 1: 200 -21370.796 1.146 1.293
Chain 1: 300 -31313.837 0.870 1.000
Chain 1: 400 -13575.964 0.979 1.293
Chain 1: 500 -17043.910 0.824 1.000
Chain 1: 600 -13508.656 0.730 1.000
Chain 1: 700 -13906.142 0.630 0.318
Chain 1: 800 -11872.914 0.573 0.318
Chain 1: 900 -14714.211 0.531 0.262
Chain 1: 1000 -14127.712 0.482 0.262
Chain 1: 1100 -12142.236 0.398 0.203
Chain 1: 1200 -11997.630 0.270 0.193
Chain 1: 1300 -10249.386 0.255 0.171
Chain 1: 1400 -15892.581 0.160 0.171
Chain 1: 1500 -12049.312 0.172 0.171
Chain 1: 1600 -12813.668 0.151 0.171
Chain 1: 1700 -10595.794 0.170 0.171
Chain 1: 1800 -13098.493 0.171 0.191
Chain 1: 1900 -10276.290 0.180 0.191
Chain 1: 2000 -10205.486 0.176 0.191
Chain 1: 2100 -10042.252 0.161 0.191
Chain 1: 2200 -11176.642 0.170 0.191
Chain 1: 2300 -10438.891 0.160 0.191
Chain 1: 2400 -9587.794 0.134 0.101
Chain 1: 2500 -9426.744 0.104 0.089
Chain 1: 2600 -11032.806 0.112 0.101
Chain 1: 2700 -9537.478 0.107 0.101
Chain 1: 2800 -9295.762 0.090 0.089
Chain 1: 2900 -9960.672 0.070 0.071
Chain 1: 3000 -8931.829 0.080 0.089
Chain 1: 3100 -9173.002 0.081 0.089
Chain 1: 3200 -8932.974 0.074 0.071
Chain 1: 3300 -9331.943 0.071 0.067
Chain 1: 3400 -16071.591 0.104 0.067
Chain 1: 3500 -9947.944 0.164 0.115
Chain 1: 3600 -9017.095 0.160 0.103
Chain 1: 3700 -9261.537 0.147 0.067
Chain 1: 3800 -9365.273 0.145 0.067
Chain 1: 3900 -12573.558 0.164 0.103
Chain 1: 4000 -9549.708 0.184 0.103
Chain 1: 4100 -9007.750 0.188 0.103
Chain 1: 4200 -9216.208 0.187 0.103
Chain 1: 4300 -9936.889 0.190 0.103
Chain 1: 4400 -9106.823 0.157 0.091
Chain 1: 4500 -9087.392 0.096 0.073
Chain 1: 4600 -9985.223 0.095 0.073
Chain 1: 4700 -11056.220 0.102 0.090
Chain 1: 4800 -10369.447 0.107 0.090
Chain 1: 4900 -8841.053 0.099 0.090
Chain 1: 5000 -10370.611 0.082 0.090
Chain 1: 5100 -8778.159 0.094 0.091
Chain 1: 5200 -9036.925 0.095 0.091
Chain 1: 5300 -11599.313 0.110 0.097
Chain 1: 5400 -9249.739 0.126 0.147
Chain 1: 5500 -8458.240 0.135 0.147
Chain 1: 5600 -8566.611 0.127 0.147
Chain 1: 5700 -9887.983 0.131 0.147
Chain 1: 5800 -9256.103 0.131 0.147
Chain 1: 5900 -9211.534 0.115 0.134
Chain 1: 6000 -12060.636 0.123 0.134
Chain 1: 6100 -12050.336 0.105 0.094
Chain 1: 6200 -10570.464 0.116 0.134
Chain 1: 6300 -8833.635 0.114 0.134
Chain 1: 6400 -10347.586 0.103 0.134
Chain 1: 6500 -12613.862 0.112 0.140
Chain 1: 6600 -8943.797 0.152 0.146
Chain 1: 6700 -12253.046 0.165 0.180
Chain 1: 6800 -9282.922 0.190 0.197
Chain 1: 6900 -10675.444 0.203 0.197
Chain 1: 7000 -8608.820 0.203 0.197
Chain 1: 7100 -8474.302 0.205 0.197
Chain 1: 7200 -12132.122 0.221 0.240
Chain 1: 7300 -9531.900 0.229 0.270
Chain 1: 7400 -10637.973 0.224 0.270
Chain 1: 7500 -8448.096 0.232 0.270
Chain 1: 7600 -8936.448 0.197 0.259
Chain 1: 7700 -8594.715 0.174 0.240
Chain 1: 7800 -11345.608 0.166 0.240
Chain 1: 7900 -8810.814 0.182 0.242
Chain 1: 8000 -10473.085 0.174 0.242
Chain 1: 8100 -11308.840 0.179 0.242
Chain 1: 8200 -10021.313 0.162 0.159
Chain 1: 8300 -8917.257 0.147 0.128
Chain 1: 8400 -12557.031 0.166 0.159
Chain 1: 8500 -8453.436 0.188 0.159
Chain 1: 8600 -8342.570 0.184 0.159
Chain 1: 8700 -11416.167 0.207 0.242
Chain 1: 8800 -8382.478 0.219 0.269
Chain 1: 8900 -8806.770 0.195 0.159
Chain 1: 9000 -9155.443 0.183 0.128
Chain 1: 9100 -9176.675 0.176 0.128
Chain 1: 9200 -10308.892 0.174 0.124
Chain 1: 9300 -8478.833 0.183 0.216
Chain 1: 9400 -9491.601 0.165 0.110
Chain 1: 9500 -8742.064 0.125 0.107
Chain 1: 9600 -10734.950 0.142 0.110
Chain 1: 9700 -8378.329 0.144 0.110
Chain 1: 9800 -8536.196 0.109 0.107
Chain 1: 9900 -8234.501 0.108 0.107
Chain 1: 10000 -8292.504 0.105 0.107
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001931 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 19.31 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57060.414 1.000 1.000
Chain 1: 200 -17619.216 1.619 2.239
Chain 1: 300 -8810.055 1.413 1.000
Chain 1: 400 -8344.800 1.074 1.000
Chain 1: 500 -8467.814 0.862 1.000
Chain 1: 600 -9099.922 0.730 1.000
Chain 1: 700 -7884.705 0.647 0.154
Chain 1: 800 -8100.037 0.570 0.154
Chain 1: 900 -8029.903 0.508 0.069
Chain 1: 1000 -7907.427 0.458 0.069
Chain 1: 1100 -7640.103 0.362 0.056
Chain 1: 1200 -7704.642 0.139 0.035
Chain 1: 1300 -7600.543 0.040 0.027
Chain 1: 1400 -7598.757 0.035 0.015
Chain 1: 1500 -7581.266 0.033 0.015
Chain 1: 1600 -7769.536 0.029 0.015
Chain 1: 1700 -7535.924 0.017 0.015
Chain 1: 1800 -7691.043 0.016 0.015
Chain 1: 1900 -7562.891 0.017 0.017
Chain 1: 2000 -7613.298 0.016 0.017
Chain 1: 2100 -7479.097 0.014 0.017
Chain 1: 2200 -7801.999 0.017 0.018
Chain 1: 2300 -7556.257 0.019 0.020
Chain 1: 2400 -7631.699 0.020 0.020
Chain 1: 2500 -7534.802 0.021 0.020
Chain 1: 2600 -7499.467 0.019 0.018
Chain 1: 2700 -7500.875 0.016 0.017
Chain 1: 2800 -7479.126 0.015 0.013
Chain 1: 2900 -7374.175 0.014 0.013
Chain 1: 3000 -7515.799 0.016 0.014
Chain 1: 3100 -7508.218 0.014 0.013
Chain 1: 3200 -7705.006 0.012 0.013
Chain 1: 3300 -7436.929 0.013 0.013
Chain 1: 3400 -7649.165 0.014 0.014
Chain 1: 3500 -7416.937 0.016 0.019
Chain 1: 3600 -7482.240 0.017 0.019
Chain 1: 3700 -7431.106 0.017 0.019
Chain 1: 3800 -7432.686 0.017 0.019
Chain 1: 3900 -7398.716 0.016 0.019
Chain 1: 4000 -7393.847 0.014 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003246 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.46 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87102.427 1.000 1.000
Chain 1: 200 -13697.812 3.179 5.359
Chain 1: 300 -10046.515 2.241 1.000
Chain 1: 400 -10958.859 1.701 1.000
Chain 1: 500 -8935.008 1.406 0.363
Chain 1: 600 -8577.325 1.179 0.363
Chain 1: 700 -8487.778 1.012 0.227
Chain 1: 800 -8751.634 0.889 0.227
Chain 1: 900 -8795.012 0.791 0.083
Chain 1: 1000 -8703.998 0.713 0.083
Chain 1: 1100 -8839.549 0.615 0.042 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8372.853 0.084 0.042
Chain 1: 1300 -8731.097 0.052 0.041
Chain 1: 1400 -8702.472 0.044 0.030
Chain 1: 1500 -8606.211 0.022 0.015
Chain 1: 1600 -8709.173 0.019 0.012
Chain 1: 1700 -8787.730 0.019 0.012
Chain 1: 1800 -8368.749 0.021 0.012
Chain 1: 1900 -8467.028 0.022 0.012
Chain 1: 2000 -8441.069 0.021 0.012
Chain 1: 2100 -8565.303 0.021 0.012
Chain 1: 2200 -8377.013 0.018 0.012
Chain 1: 2300 -8461.677 0.015 0.012
Chain 1: 2400 -8531.122 0.015 0.012
Chain 1: 2500 -8477.134 0.015 0.012
Chain 1: 2600 -8477.592 0.014 0.010
Chain 1: 2700 -8394.691 0.014 0.010
Chain 1: 2800 -8356.025 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003403 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.03 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8411992.356 1.000 1.000
Chain 1: 200 -1586216.987 2.652 4.303
Chain 1: 300 -891627.354 2.027 1.000
Chain 1: 400 -458176.931 1.757 1.000
Chain 1: 500 -358315.866 1.461 0.946
Chain 1: 600 -233116.550 1.307 0.946
Chain 1: 700 -119355.483 1.257 0.946
Chain 1: 800 -86578.016 1.147 0.946
Chain 1: 900 -66931.647 1.052 0.779
Chain 1: 1000 -51741.121 0.976 0.779
Chain 1: 1100 -39233.014 0.908 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38409.756 0.480 0.379
Chain 1: 1300 -26379.700 0.448 0.379
Chain 1: 1400 -26099.891 0.354 0.319
Chain 1: 1500 -22690.449 0.341 0.319
Chain 1: 1600 -21908.171 0.291 0.294
Chain 1: 1700 -20783.412 0.201 0.294
Chain 1: 1800 -20727.885 0.164 0.150
Chain 1: 1900 -21054.106 0.136 0.054
Chain 1: 2000 -19565.878 0.114 0.054
Chain 1: 2100 -19804.258 0.083 0.036
Chain 1: 2200 -20030.659 0.082 0.036
Chain 1: 2300 -19647.848 0.039 0.019
Chain 1: 2400 -19419.922 0.039 0.019
Chain 1: 2500 -19221.868 0.025 0.015
Chain 1: 2600 -18852.102 0.023 0.015
Chain 1: 2700 -18809.065 0.018 0.012
Chain 1: 2800 -18525.909 0.019 0.015
Chain 1: 2900 -18807.107 0.019 0.015
Chain 1: 3000 -18793.310 0.012 0.012
Chain 1: 3100 -18878.327 0.011 0.012
Chain 1: 3200 -18569.002 0.012 0.015
Chain 1: 3300 -18773.708 0.011 0.012
Chain 1: 3400 -18248.609 0.012 0.015
Chain 1: 3500 -18860.564 0.015 0.015
Chain 1: 3600 -18167.082 0.016 0.015
Chain 1: 3700 -18553.994 0.018 0.017
Chain 1: 3800 -17513.515 0.023 0.021
Chain 1: 3900 -17509.624 0.021 0.021
Chain 1: 4000 -17626.930 0.022 0.021
Chain 1: 4100 -17540.726 0.022 0.021
Chain 1: 4200 -17356.879 0.021 0.021
Chain 1: 4300 -17495.346 0.021 0.021
Chain 1: 4400 -17452.143 0.018 0.011
Chain 1: 4500 -17354.627 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00144 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.4 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12183.499 1.000 1.000
Chain 1: 200 -9044.733 0.674 1.000
Chain 1: 300 -7855.142 0.499 0.347
Chain 1: 400 -8073.934 0.381 0.347
Chain 1: 500 -7920.276 0.309 0.151
Chain 1: 600 -7799.987 0.260 0.151
Chain 1: 700 -7720.624 0.224 0.027
Chain 1: 800 -7728.894 0.196 0.027
Chain 1: 900 -7632.280 0.176 0.019
Chain 1: 1000 -7748.384 0.160 0.019
Chain 1: 1100 -7838.466 0.061 0.015
Chain 1: 1200 -7756.478 0.027 0.015
Chain 1: 1300 -7697.817 0.013 0.013
Chain 1: 1400 -7728.500 0.011 0.011
Chain 1: 1500 -7808.704 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001421 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.21 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -63078.501 1.000 1.000
Chain 1: 200 -17918.836 1.760 2.520
Chain 1: 300 -8658.291 1.530 1.070
Chain 1: 400 -8348.999 1.157 1.070
Chain 1: 500 -8074.441 0.932 1.000
Chain 1: 600 -8496.202 0.785 1.000
Chain 1: 700 -8573.634 0.674 0.050
Chain 1: 800 -8014.354 0.599 0.070
Chain 1: 900 -7942.703 0.533 0.050
Chain 1: 1000 -7791.841 0.482 0.050
Chain 1: 1100 -7678.668 0.383 0.037
Chain 1: 1200 -7541.857 0.133 0.034
Chain 1: 1300 -7683.709 0.028 0.019
Chain 1: 1400 -7822.503 0.026 0.018
Chain 1: 1500 -7575.800 0.026 0.018
Chain 1: 1600 -7508.902 0.022 0.018
Chain 1: 1700 -7479.512 0.021 0.018
Chain 1: 1800 -7561.465 0.015 0.018
Chain 1: 1900 -7553.686 0.015 0.018
Chain 1: 2000 -7563.016 0.013 0.015
Chain 1: 2100 -7551.446 0.011 0.011
Chain 1: 2200 -7648.616 0.011 0.011
Chain 1: 2300 -7729.174 0.010 0.010
Chain 1: 2400 -7604.187 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003655 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 36.55 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86291.366 1.000 1.000
Chain 1: 200 -13254.784 3.255 5.510
Chain 1: 300 -9668.591 2.294 1.000
Chain 1: 400 -10500.562 1.740 1.000
Chain 1: 500 -8610.488 1.436 0.371
Chain 1: 600 -8308.756 1.203 0.371
Chain 1: 700 -8518.757 1.034 0.220
Chain 1: 800 -9141.188 0.914 0.220
Chain 1: 900 -8531.394 0.820 0.079
Chain 1: 1000 -8253.725 0.741 0.079
Chain 1: 1100 -8532.464 0.645 0.071 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8196.501 0.098 0.068
Chain 1: 1300 -8375.241 0.063 0.041
Chain 1: 1400 -8382.051 0.055 0.036
Chain 1: 1500 -8245.970 0.035 0.034
Chain 1: 1600 -8356.936 0.032 0.033
Chain 1: 1700 -8443.296 0.031 0.033
Chain 1: 1800 -8040.618 0.029 0.033
Chain 1: 1900 -8138.503 0.023 0.021
Chain 1: 2000 -8110.102 0.020 0.017
Chain 1: 2100 -8229.978 0.018 0.015
Chain 1: 2200 -8039.123 0.017 0.015
Chain 1: 2300 -8174.165 0.016 0.015
Chain 1: 2400 -8049.032 0.018 0.016
Chain 1: 2500 -8113.658 0.017 0.015
Chain 1: 2600 -8137.470 0.016 0.015
Chain 1: 2700 -8055.725 0.016 0.015
Chain 1: 2800 -8028.234 0.011 0.012
Chain 1: 2900 -8083.663 0.011 0.010
Chain 1: 3000 -7966.917 0.012 0.015
Chain 1: 3100 -8105.944 0.012 0.015
Chain 1: 3200 -7985.210 0.011 0.015
Chain 1: 3300 -8007.503 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00321 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.1 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8440860.020 1.000 1.000
Chain 1: 200 -1590325.601 2.654 4.308
Chain 1: 300 -890086.547 2.031 1.000
Chain 1: 400 -456563.204 1.761 1.000
Chain 1: 500 -356167.741 1.465 0.950
Chain 1: 600 -231355.947 1.311 0.950
Chain 1: 700 -118232.219 1.260 0.950
Chain 1: 800 -85640.076 1.150 0.950
Chain 1: 900 -66125.472 1.055 0.787
Chain 1: 1000 -51036.490 0.979 0.787
Chain 1: 1100 -38621.271 0.911 0.539 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37810.681 0.483 0.381
Chain 1: 1300 -25872.704 0.450 0.381
Chain 1: 1400 -25602.808 0.356 0.321
Chain 1: 1500 -22217.588 0.343 0.321
Chain 1: 1600 -21442.621 0.293 0.296
Chain 1: 1700 -20328.615 0.203 0.295
Chain 1: 1800 -20275.784 0.165 0.152
Chain 1: 1900 -20601.744 0.137 0.055
Chain 1: 2000 -19119.840 0.115 0.055
Chain 1: 2100 -19357.868 0.084 0.036
Chain 1: 2200 -19583.153 0.083 0.036
Chain 1: 2300 -19201.433 0.039 0.020
Chain 1: 2400 -18973.684 0.039 0.020
Chain 1: 2500 -18775.383 0.025 0.016
Chain 1: 2600 -18406.178 0.024 0.016
Chain 1: 2700 -18363.353 0.018 0.012
Chain 1: 2800 -18080.150 0.020 0.016
Chain 1: 2900 -18361.159 0.020 0.015
Chain 1: 3000 -18347.471 0.012 0.012
Chain 1: 3100 -18432.395 0.011 0.012
Chain 1: 3200 -18123.316 0.012 0.015
Chain 1: 3300 -18327.878 0.011 0.012
Chain 1: 3400 -17803.099 0.013 0.015
Chain 1: 3500 -18414.354 0.015 0.016
Chain 1: 3600 -17721.768 0.017 0.016
Chain 1: 3700 -18107.934 0.019 0.017
Chain 1: 3800 -17068.735 0.023 0.021
Chain 1: 3900 -17064.834 0.022 0.021
Chain 1: 4000 -17182.210 0.022 0.021
Chain 1: 4100 -17095.981 0.022 0.021
Chain 1: 4200 -16912.479 0.022 0.021
Chain 1: 4300 -17050.741 0.022 0.021
Chain 1: 4400 -17007.743 0.019 0.011
Chain 1: 4500 -16910.253 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001326 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.26 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12291.020 1.000 1.000
Chain 1: 200 -9213.695 0.667 1.000
Chain 1: 300 -8099.806 0.491 0.334
Chain 1: 400 -8243.829 0.372 0.334
Chain 1: 500 -8169.858 0.300 0.138
Chain 1: 600 -8065.126 0.252 0.138
Chain 1: 700 -7930.905 0.218 0.017
Chain 1: 800 -7942.601 0.191 0.017
Chain 1: 900 -7897.360 0.171 0.017
Chain 1: 1000 -7980.834 0.155 0.017
Chain 1: 1100 -8048.953 0.055 0.013
Chain 1: 1200 -7932.203 0.023 0.013
Chain 1: 1300 -7907.864 0.010 0.010
Chain 1: 1400 -7915.313 0.008 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001486 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.86 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61461.284 1.000 1.000
Chain 1: 200 -17737.140 1.733 2.465
Chain 1: 300 -8803.538 1.493 1.015
Chain 1: 400 -9128.716 1.129 1.015
Chain 1: 500 -7973.376 0.932 1.000
Chain 1: 600 -8308.610 0.783 1.000
Chain 1: 700 -8100.713 0.675 0.145
Chain 1: 800 -8149.825 0.592 0.145
Chain 1: 900 -7952.538 0.529 0.040
Chain 1: 1000 -7921.832 0.476 0.040
Chain 1: 1100 -7698.471 0.379 0.036
Chain 1: 1200 -7789.556 0.134 0.029
Chain 1: 1300 -7759.364 0.033 0.026
Chain 1: 1400 -7675.440 0.030 0.025
Chain 1: 1500 -7589.027 0.017 0.012
Chain 1: 1600 -7845.626 0.016 0.012
Chain 1: 1700 -7526.360 0.018 0.012
Chain 1: 1800 -7620.156 0.018 0.012
Chain 1: 1900 -7477.110 0.018 0.012
Chain 1: 2000 -7604.176 0.019 0.017
Chain 1: 2100 -7659.810 0.017 0.012
Chain 1: 2200 -7690.702 0.016 0.012
Chain 1: 2300 -7595.183 0.017 0.013
Chain 1: 2400 -7645.060 0.017 0.013
Chain 1: 2500 -7586.611 0.016 0.013
Chain 1: 2600 -7527.824 0.014 0.012
Chain 1: 2700 -7566.969 0.010 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003724 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 37.24 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85875.843 1.000 1.000
Chain 1: 200 -13424.825 3.198 5.397
Chain 1: 300 -9841.362 2.254 1.000
Chain 1: 400 -10723.488 1.711 1.000
Chain 1: 500 -8776.731 1.413 0.364
Chain 1: 600 -8367.224 1.186 0.364
Chain 1: 700 -8482.580 1.018 0.222
Chain 1: 800 -9128.869 0.900 0.222
Chain 1: 900 -8676.307 0.806 0.082
Chain 1: 1000 -8449.551 0.728 0.082
Chain 1: 1100 -8746.417 0.631 0.071 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8267.830 0.097 0.058
Chain 1: 1300 -8532.752 0.064 0.052
Chain 1: 1400 -8544.629 0.056 0.049
Chain 1: 1500 -8461.955 0.035 0.034
Chain 1: 1600 -8555.165 0.031 0.031
Chain 1: 1700 -8633.534 0.030 0.031
Chain 1: 1800 -8239.364 0.028 0.031
Chain 1: 1900 -8340.149 0.024 0.027
Chain 1: 2000 -8311.187 0.022 0.012
Chain 1: 2100 -8432.365 0.020 0.012
Chain 1: 2200 -8210.307 0.017 0.012
Chain 1: 2300 -8369.281 0.015 0.012
Chain 1: 2400 -8380.882 0.015 0.012
Chain 1: 2500 -8353.229 0.015 0.012
Chain 1: 2600 -8355.763 0.014 0.012
Chain 1: 2700 -8261.895 0.014 0.012
Chain 1: 2800 -8232.201 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004724 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 47.24 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8377765.094 1.000 1.000
Chain 1: 200 -1581772.391 2.648 4.296
Chain 1: 300 -891023.673 2.024 1.000
Chain 1: 400 -457382.107 1.755 1.000
Chain 1: 500 -357999.358 1.459 0.948
Chain 1: 600 -233043.109 1.306 0.948
Chain 1: 700 -119249.952 1.255 0.948
Chain 1: 800 -86409.780 1.146 0.948
Chain 1: 900 -66746.013 1.051 0.775
Chain 1: 1000 -51526.169 0.976 0.775
Chain 1: 1100 -38985.142 0.908 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38161.167 0.480 0.380
Chain 1: 1300 -26106.486 0.449 0.380
Chain 1: 1400 -25823.970 0.355 0.322
Chain 1: 1500 -22407.153 0.343 0.322
Chain 1: 1600 -21621.987 0.293 0.295
Chain 1: 1700 -20494.737 0.203 0.295
Chain 1: 1800 -20438.641 0.165 0.152
Chain 1: 1900 -20764.423 0.137 0.055
Chain 1: 2000 -19275.526 0.116 0.055
Chain 1: 2100 -19514.066 0.085 0.036
Chain 1: 2200 -19740.204 0.084 0.036
Chain 1: 2300 -19357.778 0.039 0.020
Chain 1: 2400 -19129.951 0.039 0.020
Chain 1: 2500 -18931.920 0.025 0.016
Chain 1: 2600 -18562.522 0.024 0.016
Chain 1: 2700 -18519.658 0.018 0.012
Chain 1: 2800 -18236.554 0.020 0.016
Chain 1: 2900 -18517.734 0.020 0.015
Chain 1: 3000 -18503.989 0.012 0.012
Chain 1: 3100 -18588.873 0.011 0.012
Chain 1: 3200 -18279.820 0.012 0.015
Chain 1: 3300 -18484.376 0.011 0.012
Chain 1: 3400 -17959.678 0.013 0.015
Chain 1: 3500 -18570.966 0.015 0.016
Chain 1: 3600 -17878.502 0.017 0.016
Chain 1: 3700 -18264.617 0.019 0.017
Chain 1: 3800 -17225.623 0.023 0.021
Chain 1: 3900 -17221.809 0.022 0.021
Chain 1: 4000 -17339.108 0.022 0.021
Chain 1: 4100 -17252.855 0.022 0.021
Chain 1: 4200 -17069.464 0.022 0.021
Chain 1: 4300 -17207.633 0.021 0.021
Chain 1: 4400 -17164.696 0.019 0.011
Chain 1: 4500 -17067.275 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001818 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 18.18 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13355.653 1.000 1.000
Chain 1: 200 -9828.968 0.679 1.000
Chain 1: 300 -8485.440 0.506 0.359
Chain 1: 400 -8365.009 0.383 0.359
Chain 1: 500 -8194.975 0.310 0.158
Chain 1: 600 -8036.956 0.262 0.158
Chain 1: 700 -8175.692 0.227 0.021
Chain 1: 800 -7877.719 0.203 0.038
Chain 1: 900 -7899.872 0.181 0.021
Chain 1: 1000 -7981.535 0.164 0.021
Chain 1: 1100 -8078.135 0.065 0.020
Chain 1: 1200 -7942.488 0.031 0.017
Chain 1: 1300 -7930.180 0.015 0.017
Chain 1: 1400 -7899.213 0.014 0.017
Chain 1: 1500 -8005.866 0.014 0.013
Chain 1: 1600 -7935.427 0.012 0.012
Chain 1: 1700 -7889.635 0.011 0.010
Chain 1: 1800 -7862.392 0.008 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001496 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.96 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58389.051 1.000 1.000
Chain 1: 200 -18023.957 1.620 2.240
Chain 1: 300 -8866.361 1.424 1.033
Chain 1: 400 -8081.448 1.092 1.033
Chain 1: 500 -8684.419 0.888 1.000
Chain 1: 600 -8924.966 0.744 1.000
Chain 1: 700 -8630.400 0.643 0.097
Chain 1: 800 -7866.475 0.575 0.097
Chain 1: 900 -7927.886 0.512 0.097
Chain 1: 1000 -8183.249 0.464 0.097
Chain 1: 1100 -8168.187 0.364 0.069
Chain 1: 1200 -7878.254 0.144 0.037
Chain 1: 1300 -7840.502 0.041 0.034
Chain 1: 1400 -7921.242 0.032 0.031
Chain 1: 1500 -7611.546 0.029 0.031
Chain 1: 1600 -7763.873 0.028 0.031
Chain 1: 1700 -7584.006 0.027 0.024
Chain 1: 1800 -7658.495 0.019 0.020
Chain 1: 1900 -7714.982 0.019 0.020
Chain 1: 2000 -7709.275 0.016 0.010
Chain 1: 2100 -7635.311 0.016 0.010
Chain 1: 2200 -7766.401 0.014 0.010
Chain 1: 2300 -7605.695 0.016 0.017
Chain 1: 2400 -7703.438 0.016 0.017
Chain 1: 2500 -7474.718 0.015 0.017
Chain 1: 2600 -7567.637 0.014 0.013
Chain 1: 2700 -7562.330 0.012 0.012
Chain 1: 2800 -7520.839 0.012 0.012
Chain 1: 2900 -7416.290 0.012 0.013
Chain 1: 3000 -7560.052 0.014 0.014
Chain 1: 3100 -7560.813 0.013 0.014
Chain 1: 3200 -7774.843 0.014 0.014
Chain 1: 3300 -7486.337 0.016 0.014
Chain 1: 3400 -7720.294 0.018 0.019
Chain 1: 3500 -7472.199 0.018 0.019
Chain 1: 3600 -7540.097 0.018 0.019
Chain 1: 3700 -7488.766 0.018 0.019
Chain 1: 3800 -7487.086 0.018 0.019
Chain 1: 3900 -7451.870 0.017 0.019
Chain 1: 4000 -7444.584 0.015 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003302 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.02 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86937.509 1.000 1.000
Chain 1: 200 -13846.007 3.139 5.279
Chain 1: 300 -10089.471 2.217 1.000
Chain 1: 400 -11640.689 1.696 1.000
Chain 1: 500 -8957.690 1.417 0.372
Chain 1: 600 -8609.514 1.187 0.372
Chain 1: 700 -8702.457 1.019 0.300
Chain 1: 800 -8916.919 0.895 0.300
Chain 1: 900 -8872.015 0.796 0.133
Chain 1: 1000 -8910.257 0.717 0.133
Chain 1: 1100 -8922.054 0.617 0.040 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8402.304 0.095 0.040
Chain 1: 1300 -8712.342 0.062 0.036
Chain 1: 1400 -8647.352 0.049 0.024
Chain 1: 1500 -8571.272 0.020 0.011
Chain 1: 1600 -8667.965 0.017 0.011
Chain 1: 1700 -8725.769 0.017 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003159 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.59 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8441198.916 1.000 1.000
Chain 1: 200 -1589662.263 2.655 4.310
Chain 1: 300 -890457.774 2.032 1.000
Chain 1: 400 -457503.380 1.760 1.000
Chain 1: 500 -357363.891 1.464 0.946
Chain 1: 600 -232462.618 1.310 0.946
Chain 1: 700 -119132.941 1.259 0.946
Chain 1: 800 -86462.794 1.149 0.946
Chain 1: 900 -66903.327 1.053 0.785
Chain 1: 1000 -51781.127 0.977 0.785
Chain 1: 1100 -39332.929 0.909 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38525.193 0.480 0.378
Chain 1: 1300 -26542.836 0.447 0.378
Chain 1: 1400 -26272.072 0.353 0.316
Chain 1: 1500 -22874.111 0.340 0.316
Chain 1: 1600 -22096.575 0.290 0.292
Chain 1: 1700 -20976.555 0.200 0.292
Chain 1: 1800 -20922.840 0.162 0.149
Chain 1: 1900 -21249.718 0.135 0.053
Chain 1: 2000 -19762.969 0.113 0.053
Chain 1: 2100 -20001.343 0.082 0.035
Chain 1: 2200 -20227.651 0.082 0.035
Chain 1: 2300 -19844.780 0.038 0.019
Chain 1: 2400 -19616.656 0.038 0.019
Chain 1: 2500 -19418.421 0.025 0.015
Chain 1: 2600 -19048.139 0.023 0.015
Chain 1: 2700 -19005.060 0.018 0.012
Chain 1: 2800 -18721.425 0.019 0.015
Chain 1: 2900 -19002.946 0.019 0.015
Chain 1: 3000 -18989.177 0.012 0.012
Chain 1: 3100 -19074.222 0.011 0.012
Chain 1: 3200 -18764.544 0.011 0.015
Chain 1: 3300 -18969.595 0.011 0.012
Chain 1: 3400 -18443.670 0.012 0.015
Chain 1: 3500 -19056.686 0.014 0.015
Chain 1: 3600 -18361.866 0.016 0.015
Chain 1: 3700 -18749.665 0.018 0.017
Chain 1: 3800 -17706.999 0.023 0.021
Chain 1: 3900 -17703.031 0.021 0.021
Chain 1: 4000 -17820.398 0.022 0.021
Chain 1: 4100 -17733.962 0.022 0.021
Chain 1: 4200 -17549.733 0.021 0.021
Chain 1: 4300 -17688.514 0.021 0.021
Chain 1: 4400 -17644.904 0.018 0.010
Chain 1: 4500 -17547.311 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001655 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 16.55 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12721.529 1.000 1.000
Chain 1: 200 -9480.904 0.671 1.000
Chain 1: 300 -8145.238 0.502 0.342
Chain 1: 400 -8385.947 0.384 0.342
Chain 1: 500 -8411.817 0.308 0.164
Chain 1: 600 -8128.454 0.262 0.164
Chain 1: 700 -8017.337 0.227 0.035
Chain 1: 800 -8004.256 0.198 0.035
Chain 1: 900 -8007.839 0.176 0.029
Chain 1: 1000 -8164.705 0.161 0.029
Chain 1: 1100 -8311.070 0.063 0.019
Chain 1: 1200 -8047.627 0.032 0.019
Chain 1: 1300 -7988.309 0.016 0.018
Chain 1: 1400 -8013.917 0.013 0.014
Chain 1: 1500 -8111.093 0.014 0.014
Chain 1: 1600 -8029.307 0.012 0.012
Chain 1: 1700 -7990.936 0.011 0.010
Chain 1: 1800 -7962.242 0.011 0.010
Chain 1: 1900 -7988.166 0.011 0.010
Chain 1: 2000 -7925.505 0.010 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00148 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.8 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58988.355 1.000 1.000
Chain 1: 200 -18021.911 1.637 2.273
Chain 1: 300 -9032.501 1.423 1.000
Chain 1: 400 -8330.239 1.088 1.000
Chain 1: 500 -8312.931 0.871 0.995
Chain 1: 600 -8371.615 0.727 0.995
Chain 1: 700 -7760.772 0.634 0.084
Chain 1: 800 -8384.635 0.564 0.084
Chain 1: 900 -8087.422 0.506 0.079
Chain 1: 1000 -7774.272 0.459 0.079
Chain 1: 1100 -7730.669 0.360 0.074
Chain 1: 1200 -7892.437 0.134 0.040
Chain 1: 1300 -7690.942 0.038 0.037
Chain 1: 1400 -7654.609 0.030 0.026
Chain 1: 1500 -7551.340 0.031 0.026
Chain 1: 1600 -7743.673 0.033 0.026
Chain 1: 1700 -7634.093 0.026 0.025
Chain 1: 1800 -7624.746 0.019 0.020
Chain 1: 1900 -7591.199 0.016 0.014
Chain 1: 2000 -7674.303 0.013 0.014
Chain 1: 2100 -7592.861 0.013 0.014
Chain 1: 2200 -7755.596 0.013 0.014
Chain 1: 2300 -7562.438 0.013 0.014
Chain 1: 2400 -7561.146 0.013 0.014
Chain 1: 2500 -7679.656 0.013 0.014
Chain 1: 2600 -7555.355 0.012 0.014
Chain 1: 2700 -7514.291 0.011 0.011
Chain 1: 2800 -7537.655 0.011 0.011
Chain 1: 2900 -7411.428 0.013 0.015
Chain 1: 3000 -7561.792 0.013 0.016
Chain 1: 3100 -7551.753 0.013 0.016
Chain 1: 3200 -7761.891 0.013 0.016
Chain 1: 3300 -7475.694 0.014 0.016
Chain 1: 3400 -7713.432 0.017 0.017
Chain 1: 3500 -7464.448 0.019 0.020
Chain 1: 3600 -7530.857 0.019 0.020
Chain 1: 3700 -7480.468 0.019 0.020
Chain 1: 3800 -7479.551 0.018 0.020
Chain 1: 3900 -7442.665 0.017 0.020
Chain 1: 4000 -7434.140 0.015 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003643 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 36.43 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86748.889 1.000 1.000
Chain 1: 200 -13872.611 3.127 5.253
Chain 1: 300 -10136.444 2.207 1.000
Chain 1: 400 -11725.291 1.689 1.000
Chain 1: 500 -8825.929 1.417 0.369
Chain 1: 600 -9640.017 1.195 0.369
Chain 1: 700 -8798.398 1.038 0.329
Chain 1: 800 -8817.402 0.909 0.329
Chain 1: 900 -8860.408 0.808 0.136
Chain 1: 1000 -9068.351 0.730 0.136
Chain 1: 1100 -8788.092 0.633 0.096 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8483.693 0.111 0.084
Chain 1: 1300 -8854.985 0.078 0.042
Chain 1: 1400 -8557.714 0.068 0.036
Chain 1: 1500 -8633.454 0.036 0.035
Chain 1: 1600 -8742.177 0.029 0.032
Chain 1: 1700 -8803.178 0.020 0.023
Chain 1: 1800 -8361.490 0.025 0.032
Chain 1: 1900 -8465.117 0.026 0.032
Chain 1: 2000 -8453.859 0.024 0.032
Chain 1: 2100 -8569.742 0.022 0.014
Chain 1: 2200 -8363.706 0.021 0.014
Chain 1: 2300 -8459.106 0.018 0.012
Chain 1: 2400 -8526.065 0.015 0.012
Chain 1: 2500 -8474.534 0.015 0.012
Chain 1: 2600 -8488.474 0.014 0.011
Chain 1: 2700 -8395.952 0.014 0.011
Chain 1: 2800 -8343.434 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00332 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.2 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8422756.139 1.000 1.000
Chain 1: 200 -1587356.979 2.653 4.306
Chain 1: 300 -891399.521 2.029 1.000
Chain 1: 400 -457875.612 1.758 1.000
Chain 1: 500 -357769.466 1.463 0.947
Chain 1: 600 -232713.098 1.308 0.947
Chain 1: 700 -119258.839 1.257 0.947
Chain 1: 800 -86555.926 1.148 0.947
Chain 1: 900 -66974.495 1.052 0.781
Chain 1: 1000 -51839.727 0.976 0.781
Chain 1: 1100 -39374.035 0.908 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38564.887 0.480 0.378
Chain 1: 1300 -26568.129 0.447 0.378
Chain 1: 1400 -26295.399 0.353 0.317
Chain 1: 1500 -22893.981 0.340 0.317
Chain 1: 1600 -22115.074 0.290 0.292
Chain 1: 1700 -20993.605 0.200 0.292
Chain 1: 1800 -20939.568 0.162 0.149
Chain 1: 1900 -21266.189 0.135 0.053
Chain 1: 2000 -19779.088 0.113 0.053
Chain 1: 2100 -20017.398 0.083 0.035
Chain 1: 2200 -20243.742 0.082 0.035
Chain 1: 2300 -19860.958 0.038 0.019
Chain 1: 2400 -19632.914 0.038 0.019
Chain 1: 2500 -19434.708 0.025 0.015
Chain 1: 2600 -19064.495 0.023 0.015
Chain 1: 2700 -19021.484 0.018 0.012
Chain 1: 2800 -18737.923 0.019 0.015
Chain 1: 2900 -19019.403 0.019 0.015
Chain 1: 3000 -19005.632 0.012 0.012
Chain 1: 3100 -19090.635 0.011 0.012
Chain 1: 3200 -18781.018 0.011 0.015
Chain 1: 3300 -18986.044 0.011 0.012
Chain 1: 3400 -18460.273 0.012 0.015
Chain 1: 3500 -19073.017 0.014 0.015
Chain 1: 3600 -18378.635 0.016 0.015
Chain 1: 3700 -18766.113 0.018 0.016
Chain 1: 3800 -17724.050 0.023 0.021
Chain 1: 3900 -17720.131 0.021 0.021
Chain 1: 4000 -17837.483 0.022 0.021
Chain 1: 4100 -17751.052 0.022 0.021
Chain 1: 4200 -17567.008 0.021 0.021
Chain 1: 4300 -17705.644 0.021 0.021
Chain 1: 4400 -17662.140 0.018 0.010
Chain 1: 4500 -17564.615 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001561 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.61 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12343.614 1.000 1.000
Chain 1: 200 -9198.887 0.671 1.000
Chain 1: 300 -8054.114 0.495 0.342
Chain 1: 400 -8230.450 0.376 0.342
Chain 1: 500 -8091.483 0.305 0.142
Chain 1: 600 -8014.952 0.255 0.142
Chain 1: 700 -7928.839 0.220 0.021
Chain 1: 800 -7969.823 0.194 0.021
Chain 1: 900 -8090.010 0.174 0.017
Chain 1: 1000 -7984.715 0.158 0.017
Chain 1: 1100 -8183.007 0.060 0.017
Chain 1: 1200 -7946.357 0.029 0.017
Chain 1: 1300 -7901.787 0.015 0.015
Chain 1: 1400 -7918.052 0.013 0.013
Chain 1: 1500 -8016.346 0.013 0.012
Chain 1: 1600 -7963.622 0.012 0.012
Chain 1: 1700 -7898.032 0.012 0.012
Chain 1: 1800 -7885.793 0.012 0.012
Chain 1: 1900 -7864.910 0.011 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001615 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 16.15 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57156.656 1.000 1.000
Chain 1: 200 -17345.884 1.648 2.295
Chain 1: 300 -8695.089 1.430 1.000
Chain 1: 400 -8202.935 1.088 1.000
Chain 1: 500 -8645.563 0.880 0.995
Chain 1: 600 -8872.718 0.738 0.995
Chain 1: 700 -7912.482 0.650 0.121
Chain 1: 800 -8331.616 0.575 0.121
Chain 1: 900 -7915.995 0.517 0.060
Chain 1: 1000 -7778.659 0.467 0.060
Chain 1: 1100 -7693.102 0.368 0.053
Chain 1: 1200 -7596.826 0.140 0.051
Chain 1: 1300 -7753.183 0.042 0.050
Chain 1: 1400 -7826.133 0.037 0.026
Chain 1: 1500 -7611.644 0.035 0.026
Chain 1: 1600 -7646.568 0.033 0.020
Chain 1: 1700 -7536.154 0.022 0.018
Chain 1: 1800 -7575.280 0.018 0.015
Chain 1: 1900 -7588.378 0.013 0.013
Chain 1: 2000 -7656.289 0.012 0.011
Chain 1: 2100 -7616.161 0.011 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003343 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.43 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86105.252 1.000 1.000
Chain 1: 200 -13449.358 3.201 5.402
Chain 1: 300 -9864.786 2.255 1.000
Chain 1: 400 -10834.899 1.714 1.000
Chain 1: 500 -8791.832 1.417 0.363
Chain 1: 600 -8489.489 1.187 0.363
Chain 1: 700 -8465.388 1.018 0.232
Chain 1: 800 -8820.612 0.896 0.232
Chain 1: 900 -8722.214 0.797 0.090
Chain 1: 1000 -8402.776 0.722 0.090
Chain 1: 1100 -8718.442 0.625 0.040 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8436.459 0.088 0.038
Chain 1: 1300 -8579.539 0.054 0.036
Chain 1: 1400 -8573.247 0.045 0.036
Chain 1: 1500 -8449.785 0.023 0.033
Chain 1: 1600 -8559.561 0.021 0.017
Chain 1: 1700 -8645.023 0.021 0.017
Chain 1: 1800 -8244.872 0.022 0.017
Chain 1: 1900 -8344.160 0.022 0.017
Chain 1: 2000 -8315.414 0.019 0.015
Chain 1: 2100 -8435.299 0.017 0.014
Chain 1: 2200 -8226.248 0.016 0.014
Chain 1: 2300 -8375.982 0.016 0.014
Chain 1: 2400 -8256.719 0.017 0.014
Chain 1: 2500 -8319.707 0.017 0.014
Chain 1: 2600 -8341.283 0.016 0.014
Chain 1: 2700 -8260.267 0.016 0.014
Chain 1: 2800 -8234.143 0.011 0.012
Chain 1: 2900 -8289.515 0.011 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003383 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.83 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8390951.634 1.000 1.000
Chain 1: 200 -1585071.603 2.647 4.294
Chain 1: 300 -890408.145 2.025 1.000
Chain 1: 400 -457316.341 1.755 1.000
Chain 1: 500 -357700.274 1.460 0.947
Chain 1: 600 -232774.227 1.306 0.947
Chain 1: 700 -119076.499 1.256 0.947
Chain 1: 800 -86322.952 1.146 0.947
Chain 1: 900 -66681.760 1.052 0.780
Chain 1: 1000 -51489.304 0.976 0.780
Chain 1: 1100 -38976.924 0.908 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38153.218 0.481 0.379
Chain 1: 1300 -26122.133 0.449 0.379
Chain 1: 1400 -25841.808 0.355 0.321
Chain 1: 1500 -22432.571 0.343 0.321
Chain 1: 1600 -21649.909 0.293 0.295
Chain 1: 1700 -20525.285 0.203 0.295
Chain 1: 1800 -20469.787 0.165 0.152
Chain 1: 1900 -20795.695 0.137 0.055
Chain 1: 2000 -19308.178 0.115 0.055
Chain 1: 2100 -19546.507 0.084 0.036
Chain 1: 2200 -19772.658 0.083 0.036
Chain 1: 2300 -19390.169 0.039 0.020
Chain 1: 2400 -19162.346 0.039 0.020
Chain 1: 2500 -18964.374 0.025 0.016
Chain 1: 2600 -18594.888 0.024 0.016
Chain 1: 2700 -18551.943 0.018 0.012
Chain 1: 2800 -18268.910 0.020 0.015
Chain 1: 2900 -18550.039 0.020 0.015
Chain 1: 3000 -18536.225 0.012 0.012
Chain 1: 3100 -18621.176 0.011 0.012
Chain 1: 3200 -18312.070 0.012 0.015
Chain 1: 3300 -18516.637 0.011 0.012
Chain 1: 3400 -17991.911 0.013 0.015
Chain 1: 3500 -18603.280 0.015 0.015
Chain 1: 3600 -17910.601 0.017 0.015
Chain 1: 3700 -18296.916 0.019 0.017
Chain 1: 3800 -17257.656 0.023 0.021
Chain 1: 3900 -17253.818 0.022 0.021
Chain 1: 4000 -17371.113 0.022 0.021
Chain 1: 4100 -17284.931 0.022 0.021
Chain 1: 4200 -17101.395 0.022 0.021
Chain 1: 4300 -17239.634 0.021 0.021
Chain 1: 4400 -17196.636 0.019 0.011
Chain 1: 4500 -17099.194 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001312 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.12 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12376.746 1.000 1.000
Chain 1: 200 -9200.085 0.673 1.000
Chain 1: 300 -7948.740 0.501 0.345
Chain 1: 400 -8127.814 0.381 0.345
Chain 1: 500 -8023.162 0.308 0.157
Chain 1: 600 -7935.770 0.258 0.157
Chain 1: 700 -7839.737 0.223 0.022
Chain 1: 800 -7884.962 0.196 0.022
Chain 1: 900 -8009.712 0.176 0.016
Chain 1: 1000 -7947.174 0.159 0.016
Chain 1: 1100 -7939.376 0.059 0.013
Chain 1: 1200 -7861.122 0.026 0.012
Chain 1: 1300 -7809.747 0.011 0.011
Chain 1: 1400 -7828.956 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001607 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 16.07 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -46180.559 1.000 1.000
Chain 1: 200 -15454.876 1.494 1.988
Chain 1: 300 -8678.002 1.256 1.000
Chain 1: 400 -8536.579 0.946 1.000
Chain 1: 500 -8383.826 0.761 0.781
Chain 1: 600 -8644.416 0.639 0.781
Chain 1: 700 -7980.905 0.560 0.083
Chain 1: 800 -8256.427 0.494 0.083
Chain 1: 900 -7932.682 0.443 0.041
Chain 1: 1000 -7833.886 0.400 0.041
Chain 1: 1100 -7762.978 0.301 0.033
Chain 1: 1200 -7908.509 0.104 0.030
Chain 1: 1300 -7818.662 0.027 0.018
Chain 1: 1400 -7797.622 0.026 0.018
Chain 1: 1500 -7600.510 0.027 0.026
Chain 1: 1600 -7869.936 0.027 0.026
Chain 1: 1700 -7524.101 0.023 0.026
Chain 1: 1800 -7597.618 0.021 0.018
Chain 1: 1900 -7569.078 0.017 0.013
Chain 1: 2000 -7637.902 0.017 0.011
Chain 1: 2100 -7600.647 0.017 0.011
Chain 1: 2200 -7698.808 0.016 0.011
Chain 1: 2300 -7679.604 0.015 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003452 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.52 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86668.114 1.000 1.000
Chain 1: 200 -13449.795 3.222 5.444
Chain 1: 300 -9817.703 2.271 1.000
Chain 1: 400 -10733.023 1.725 1.000
Chain 1: 500 -8696.145 1.427 0.370
Chain 1: 600 -8272.500 1.197 0.370
Chain 1: 700 -8538.233 1.031 0.234
Chain 1: 800 -9185.837 0.911 0.234
Chain 1: 900 -8617.078 0.817 0.085
Chain 1: 1000 -8460.301 0.737 0.085
Chain 1: 1100 -8624.099 0.639 0.071 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8358.500 0.098 0.066
Chain 1: 1300 -8458.166 0.062 0.051
Chain 1: 1400 -8554.947 0.055 0.032
Chain 1: 1500 -8381.418 0.033 0.031
Chain 1: 1600 -8498.119 0.029 0.021
Chain 1: 1700 -8579.165 0.027 0.019
Chain 1: 1800 -8168.547 0.025 0.019
Chain 1: 1900 -8264.379 0.020 0.019
Chain 1: 2000 -8237.169 0.018 0.014
Chain 1: 2100 -8359.268 0.018 0.014
Chain 1: 2200 -8178.379 0.017 0.014
Chain 1: 2300 -8260.059 0.017 0.014
Chain 1: 2400 -8329.025 0.016 0.014
Chain 1: 2500 -8274.383 0.015 0.012
Chain 1: 2600 -8273.271 0.014 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004187 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 41.87 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8409040.632 1.000 1.000
Chain 1: 200 -1586781.118 2.650 4.299
Chain 1: 300 -890579.008 2.027 1.000
Chain 1: 400 -457375.069 1.757 1.000
Chain 1: 500 -357469.643 1.462 0.947
Chain 1: 600 -232610.665 1.307 0.947
Chain 1: 700 -118967.844 1.257 0.947
Chain 1: 800 -86232.083 1.147 0.947
Chain 1: 900 -66610.711 1.053 0.782
Chain 1: 1000 -51437.206 0.977 0.782
Chain 1: 1100 -38944.514 0.909 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38124.756 0.481 0.380
Chain 1: 1300 -26111.128 0.449 0.380
Chain 1: 1400 -25834.055 0.355 0.321
Chain 1: 1500 -22428.574 0.343 0.321
Chain 1: 1600 -21647.178 0.293 0.295
Chain 1: 1700 -20524.263 0.202 0.295
Chain 1: 1800 -20469.262 0.165 0.152
Chain 1: 1900 -20795.329 0.137 0.055
Chain 1: 2000 -19308.233 0.115 0.055
Chain 1: 2100 -19546.733 0.084 0.036
Chain 1: 2200 -19772.756 0.083 0.036
Chain 1: 2300 -19390.286 0.039 0.020
Chain 1: 2400 -19162.361 0.039 0.020
Chain 1: 2500 -18964.274 0.025 0.016
Chain 1: 2600 -18594.742 0.024 0.016
Chain 1: 2700 -18551.755 0.018 0.012
Chain 1: 2800 -18268.579 0.020 0.016
Chain 1: 2900 -18549.729 0.020 0.015
Chain 1: 3000 -18536.038 0.012 0.012
Chain 1: 3100 -18620.999 0.011 0.012
Chain 1: 3200 -18311.771 0.012 0.015
Chain 1: 3300 -18516.401 0.011 0.012
Chain 1: 3400 -17991.461 0.013 0.015
Chain 1: 3500 -18603.119 0.015 0.016
Chain 1: 3600 -17909.998 0.017 0.016
Chain 1: 3700 -18296.607 0.019 0.017
Chain 1: 3800 -17256.673 0.023 0.021
Chain 1: 3900 -17252.761 0.022 0.021
Chain 1: 4000 -17370.108 0.022 0.021
Chain 1: 4100 -17283.874 0.022 0.021
Chain 1: 4200 -17100.170 0.022 0.021
Chain 1: 4300 -17238.558 0.021 0.021
Chain 1: 4400 -17195.431 0.019 0.011
Chain 1: 4500 -17097.915 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003459 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.59 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12226.233 1.000 1.000
Chain 1: 200 -9190.244 0.665 1.000
Chain 1: 300 -7877.649 0.499 0.330
Chain 1: 400 -8062.505 0.380 0.330
Chain 1: 500 -7918.649 0.308 0.167
Chain 1: 600 -7838.929 0.258 0.167
Chain 1: 700 -7747.748 0.223 0.023
Chain 1: 800 -7792.209 0.196 0.023
Chain 1: 900 -7915.425 0.176 0.018
Chain 1: 1000 -7824.045 0.159 0.018
Chain 1: 1100 -7767.479 0.060 0.016
Chain 1: 1200 -7781.031 0.027 0.012
Chain 1: 1300 -7712.786 0.011 0.012
Chain 1: 1400 -7732.592 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001508 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.08 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -50642.285 1.000 1.000
Chain 1: 200 -16020.553 1.581 2.161
Chain 1: 300 -8622.720 1.340 1.000
Chain 1: 400 -8479.007 1.009 1.000
Chain 1: 500 -8013.811 0.819 0.858
Chain 1: 600 -8513.740 0.692 0.858
Chain 1: 700 -8080.100 0.601 0.059
Chain 1: 800 -8205.327 0.528 0.059
Chain 1: 900 -7663.767 0.477 0.059
Chain 1: 1000 -7716.286 0.430 0.059
Chain 1: 1100 -7693.045 0.330 0.058
Chain 1: 1200 -7654.698 0.115 0.054
Chain 1: 1300 -7658.931 0.029 0.017
Chain 1: 1400 -7845.763 0.030 0.024
Chain 1: 1500 -7537.875 0.028 0.024
Chain 1: 1600 -7580.669 0.023 0.015
Chain 1: 1700 -7470.660 0.019 0.015
Chain 1: 1800 -7532.251 0.018 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00353 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.3 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86379.824 1.000 1.000
Chain 1: 200 -13324.116 3.241 5.483
Chain 1: 300 -9695.986 2.286 1.000
Chain 1: 400 -10398.807 1.731 1.000
Chain 1: 500 -8637.627 1.426 0.374
Chain 1: 600 -8209.023 1.197 0.374
Chain 1: 700 -8525.639 1.031 0.204
Chain 1: 800 -9186.291 0.911 0.204
Chain 1: 900 -8499.533 0.819 0.081
Chain 1: 1000 -8248.704 0.740 0.081
Chain 1: 1100 -8565.315 0.644 0.072 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8065.479 0.102 0.068
Chain 1: 1300 -8403.461 0.068 0.062
Chain 1: 1400 -8406.449 0.062 0.052
Chain 1: 1500 -8281.126 0.043 0.040
Chain 1: 1600 -8387.847 0.039 0.037
Chain 1: 1700 -8472.860 0.036 0.037
Chain 1: 1800 -8064.839 0.034 0.037
Chain 1: 1900 -8161.361 0.027 0.030
Chain 1: 2000 -8133.581 0.024 0.015
Chain 1: 2100 -8254.554 0.022 0.015
Chain 1: 2200 -8082.616 0.018 0.015
Chain 1: 2300 -8198.805 0.015 0.014
Chain 1: 2400 -8210.327 0.016 0.014
Chain 1: 2500 -8171.940 0.014 0.013
Chain 1: 2600 -8171.803 0.013 0.012
Chain 1: 2700 -8086.965 0.013 0.012
Chain 1: 2800 -8051.651 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.005071 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 50.71 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8396723.407 1.000 1.000
Chain 1: 200 -1582476.164 2.653 4.306
Chain 1: 300 -891366.604 2.027 1.000
Chain 1: 400 -457853.272 1.757 1.000
Chain 1: 500 -358483.947 1.461 0.947
Chain 1: 600 -233246.257 1.307 0.947
Chain 1: 700 -119304.510 1.257 0.947
Chain 1: 800 -86434.728 1.147 0.947
Chain 1: 900 -66724.689 1.053 0.775
Chain 1: 1000 -51483.116 0.977 0.775
Chain 1: 1100 -38924.349 0.909 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38094.892 0.481 0.380
Chain 1: 1300 -26020.931 0.450 0.380
Chain 1: 1400 -25734.736 0.356 0.323
Chain 1: 1500 -22314.203 0.344 0.323
Chain 1: 1600 -21527.790 0.294 0.296
Chain 1: 1700 -20398.616 0.204 0.295
Chain 1: 1800 -20341.892 0.166 0.153
Chain 1: 1900 -20667.873 0.138 0.055
Chain 1: 2000 -19177.616 0.116 0.055
Chain 1: 2100 -19416.077 0.085 0.037
Chain 1: 2200 -19642.619 0.084 0.037
Chain 1: 2300 -19259.796 0.040 0.020
Chain 1: 2400 -19031.939 0.040 0.020
Chain 1: 2500 -18833.951 0.025 0.016
Chain 1: 2600 -18464.258 0.024 0.016
Chain 1: 2700 -18421.286 0.018 0.012
Chain 1: 2800 -18138.224 0.020 0.016
Chain 1: 2900 -18419.449 0.020 0.015
Chain 1: 3000 -18405.595 0.012 0.012
Chain 1: 3100 -18490.581 0.011 0.012
Chain 1: 3200 -18181.341 0.012 0.015
Chain 1: 3300 -18386.008 0.011 0.012
Chain 1: 3400 -17861.097 0.013 0.015
Chain 1: 3500 -18472.718 0.015 0.016
Chain 1: 3600 -17779.775 0.017 0.016
Chain 1: 3700 -18166.338 0.019 0.017
Chain 1: 3800 -17126.573 0.023 0.021
Chain 1: 3900 -17122.761 0.022 0.021
Chain 1: 4000 -17240.043 0.022 0.021
Chain 1: 4100 -17153.808 0.022 0.021
Chain 1: 4200 -16970.223 0.022 0.021
Chain 1: 4300 -17108.511 0.021 0.021
Chain 1: 4400 -17065.457 0.019 0.011
Chain 1: 4500 -16968.011 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004324 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 43.24 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12426.495 1.000 1.000
Chain 1: 200 -9273.436 0.670 1.000
Chain 1: 300 -8180.639 0.491 0.340
Chain 1: 400 -8293.671 0.372 0.340
Chain 1: 500 -8191.970 0.300 0.134
Chain 1: 600 -8107.363 0.252 0.134
Chain 1: 700 -8023.736 0.217 0.014
Chain 1: 800 -8056.819 0.191 0.014
Chain 1: 900 -8208.268 0.171 0.014
Chain 1: 1000 -8058.541 0.156 0.018
Chain 1: 1100 -8094.893 0.057 0.014
Chain 1: 1200 -8059.637 0.023 0.012
Chain 1: 1300 -7992.639 0.011 0.010
Chain 1: 1400 -8008.199 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001541 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.41 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56983.486 1.000 1.000
Chain 1: 200 -17433.478 1.634 2.269
Chain 1: 300 -8750.436 1.420 1.000
Chain 1: 400 -8222.079 1.081 1.000
Chain 1: 500 -8524.352 0.872 0.992
Chain 1: 600 -8379.677 0.730 0.992
Chain 1: 700 -8677.227 0.630 0.064
Chain 1: 800 -8150.956 0.560 0.065
Chain 1: 900 -7876.964 0.501 0.064
Chain 1: 1000 -8002.424 0.453 0.064
Chain 1: 1100 -8001.315 0.353 0.035
Chain 1: 1200 -7680.191 0.130 0.035
Chain 1: 1300 -7786.013 0.032 0.035
Chain 1: 1400 -7834.011 0.026 0.034
Chain 1: 1500 -7631.670 0.025 0.027
Chain 1: 1600 -7779.625 0.026 0.027
Chain 1: 1700 -7538.041 0.025 0.027
Chain 1: 1800 -7626.575 0.020 0.019
Chain 1: 1900 -7596.292 0.017 0.016
Chain 1: 2000 -7639.380 0.016 0.014
Chain 1: 2100 -7613.762 0.016 0.014
Chain 1: 2200 -7728.246 0.014 0.014
Chain 1: 2300 -7635.769 0.014 0.012
Chain 1: 2400 -7677.757 0.013 0.012
Chain 1: 2500 -7623.806 0.012 0.012
Chain 1: 2600 -7585.496 0.010 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003023 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.23 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85931.036 1.000 1.000
Chain 1: 200 -13519.142 3.178 5.356
Chain 1: 300 -9944.735 2.239 1.000
Chain 1: 400 -10931.294 1.701 1.000
Chain 1: 500 -8889.467 1.407 0.359
Chain 1: 600 -8838.040 1.174 0.359
Chain 1: 700 -8473.420 1.012 0.230
Chain 1: 800 -8917.178 0.892 0.230
Chain 1: 900 -8781.783 0.794 0.090
Chain 1: 1000 -8519.102 0.718 0.090
Chain 1: 1100 -8765.926 0.621 0.050 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8419.042 0.089 0.043
Chain 1: 1300 -8731.071 0.057 0.041
Chain 1: 1400 -8640.906 0.049 0.036
Chain 1: 1500 -8543.582 0.027 0.031
Chain 1: 1600 -8644.951 0.028 0.031
Chain 1: 1700 -8734.335 0.024 0.028
Chain 1: 1800 -8336.643 0.024 0.028
Chain 1: 1900 -8437.929 0.024 0.028
Chain 1: 2000 -8408.720 0.021 0.012
Chain 1: 2100 -8530.112 0.020 0.012
Chain 1: 2200 -8309.160 0.018 0.012
Chain 1: 2300 -8466.768 0.017 0.012
Chain 1: 2400 -8479.781 0.016 0.012
Chain 1: 2500 -8450.449 0.015 0.012
Chain 1: 2600 -8453.119 0.014 0.012
Chain 1: 2700 -8359.316 0.014 0.012
Chain 1: 2800 -8329.879 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002924 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.24 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8390054.204 1.000 1.000
Chain 1: 200 -1585936.319 2.645 4.290
Chain 1: 300 -891911.923 2.023 1.000
Chain 1: 400 -458546.449 1.753 1.000
Chain 1: 500 -358836.432 1.458 0.945
Chain 1: 600 -233580.914 1.305 0.945
Chain 1: 700 -119475.790 1.255 0.945
Chain 1: 800 -86629.009 1.145 0.945
Chain 1: 900 -66912.421 1.051 0.778
Chain 1: 1000 -51668.735 0.975 0.778
Chain 1: 1100 -39113.299 0.907 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38280.214 0.480 0.379
Chain 1: 1300 -26210.466 0.449 0.379
Chain 1: 1400 -25925.211 0.355 0.321
Chain 1: 1500 -22506.436 0.343 0.321
Chain 1: 1600 -21720.445 0.293 0.295
Chain 1: 1700 -20591.610 0.203 0.295
Chain 1: 1800 -20534.860 0.165 0.152
Chain 1: 1900 -20860.606 0.137 0.055
Chain 1: 2000 -19371.016 0.115 0.055
Chain 1: 2100 -19609.409 0.084 0.036
Chain 1: 2200 -19835.905 0.083 0.036
Chain 1: 2300 -19453.136 0.039 0.020
Chain 1: 2400 -19225.306 0.039 0.020
Chain 1: 2500 -19027.500 0.025 0.016
Chain 1: 2600 -18657.964 0.024 0.016
Chain 1: 2700 -18614.924 0.018 0.012
Chain 1: 2800 -18332.050 0.020 0.015
Chain 1: 2900 -18613.127 0.020 0.015
Chain 1: 3000 -18599.329 0.012 0.012
Chain 1: 3100 -18684.303 0.011 0.012
Chain 1: 3200 -18375.175 0.012 0.015
Chain 1: 3300 -18579.701 0.011 0.012
Chain 1: 3400 -18055.085 0.013 0.015
Chain 1: 3500 -18666.370 0.015 0.015
Chain 1: 3600 -17973.769 0.017 0.015
Chain 1: 3700 -18360.073 0.019 0.017
Chain 1: 3800 -17321.005 0.023 0.021
Chain 1: 3900 -17317.183 0.021 0.021
Chain 1: 4000 -17434.461 0.022 0.021
Chain 1: 4100 -17348.336 0.022 0.021
Chain 1: 4200 -17164.795 0.021 0.021
Chain 1: 4300 -17303.015 0.021 0.021
Chain 1: 4400 -17260.041 0.019 0.011
Chain 1: 4500 -17162.630 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001324 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.24 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12977.634 1.000 1.000
Chain 1: 200 -9798.814 0.662 1.000
Chain 1: 300 -8344.694 0.500 0.324
Chain 1: 400 -8584.570 0.382 0.324
Chain 1: 500 -8515.955 0.307 0.174
Chain 1: 600 -8289.441 0.260 0.174
Chain 1: 700 -8154.526 0.226 0.028
Chain 1: 800 -8164.411 0.197 0.028
Chain 1: 900 -8168.257 0.176 0.027
Chain 1: 1000 -8248.089 0.159 0.027
Chain 1: 1100 -8262.317 0.059 0.017
Chain 1: 1200 -8197.960 0.028 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001492 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.92 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -62357.862 1.000 1.000
Chain 1: 200 -18548.002 1.681 2.362
Chain 1: 300 -9233.529 1.457 1.009
Chain 1: 400 -8463.893 1.115 1.009
Chain 1: 500 -8561.501 0.895 1.000
Chain 1: 600 -9106.674 0.755 1.000
Chain 1: 700 -7670.691 0.674 0.187
Chain 1: 800 -8348.802 0.600 0.187
Chain 1: 900 -7521.737 0.546 0.110
Chain 1: 1000 -7765.489 0.494 0.110
Chain 1: 1100 -7864.054 0.396 0.091
Chain 1: 1200 -7898.305 0.160 0.081
Chain 1: 1300 -7877.065 0.059 0.060
Chain 1: 1400 -7824.997 0.051 0.031
Chain 1: 1500 -7471.723 0.054 0.047
Chain 1: 1600 -7713.135 0.051 0.031
Chain 1: 1700 -7507.226 0.035 0.031
Chain 1: 1800 -7608.634 0.029 0.027
Chain 1: 1900 -7494.617 0.019 0.015
Chain 1: 2000 -7498.401 0.016 0.013
Chain 1: 2100 -7481.589 0.015 0.013
Chain 1: 2200 -7716.522 0.018 0.015
Chain 1: 2300 -7598.449 0.019 0.016
Chain 1: 2400 -7539.307 0.019 0.016
Chain 1: 2500 -7550.452 0.015 0.015
Chain 1: 2600 -7493.607 0.012 0.013
Chain 1: 2700 -7487.625 0.009 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003051 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.51 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86428.371 1.000 1.000
Chain 1: 200 -14178.854 3.048 5.096
Chain 1: 300 -10375.107 2.154 1.000
Chain 1: 400 -12282.424 1.654 1.000
Chain 1: 500 -9070.649 1.394 0.367
Chain 1: 600 -9654.259 1.172 0.367
Chain 1: 700 -8872.878 1.017 0.354
Chain 1: 800 -9460.554 0.898 0.354
Chain 1: 900 -9065.881 0.803 0.155
Chain 1: 1000 -9328.016 0.725 0.155
Chain 1: 1100 -9085.514 0.628 0.088 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8645.532 0.124 0.062
Chain 1: 1300 -8964.422 0.090 0.060
Chain 1: 1400 -8828.145 0.076 0.051
Chain 1: 1500 -8872.903 0.042 0.044
Chain 1: 1600 -8919.788 0.036 0.036
Chain 1: 1700 -8971.944 0.028 0.028
Chain 1: 1800 -8527.828 0.027 0.028
Chain 1: 1900 -8627.652 0.024 0.027
Chain 1: 2000 -8648.101 0.021 0.015
Chain 1: 2100 -8734.108 0.019 0.012
Chain 1: 2200 -8512.608 0.017 0.012
Chain 1: 2300 -8730.716 0.016 0.012
Chain 1: 2400 -8525.354 0.017 0.012
Chain 1: 2500 -8599.061 0.017 0.012
Chain 1: 2600 -8509.097 0.018 0.012
Chain 1: 2700 -8542.866 0.017 0.012
Chain 1: 2800 -8493.569 0.013 0.011
Chain 1: 2900 -8608.734 0.013 0.011
Chain 1: 3000 -8518.862 0.014 0.011
Chain 1: 3100 -8485.500 0.013 0.011
Chain 1: 3200 -8456.710 0.011 0.011
Chain 1: 3300 -8719.093 0.011 0.011
Chain 1: 3400 -8764.385 0.010 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003263 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.63 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8412342.699 1.000 1.000
Chain 1: 200 -1583822.868 2.656 4.311
Chain 1: 300 -891653.250 2.029 1.000
Chain 1: 400 -458820.341 1.758 1.000
Chain 1: 500 -359010.098 1.462 0.943
Chain 1: 600 -234046.471 1.307 0.943
Chain 1: 700 -120106.948 1.256 0.943
Chain 1: 800 -87282.931 1.146 0.943
Chain 1: 900 -67602.392 1.051 0.776
Chain 1: 1000 -52387.388 0.975 0.776
Chain 1: 1100 -39846.710 0.906 0.534 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39029.409 0.477 0.376
Chain 1: 1300 -26949.792 0.445 0.376
Chain 1: 1400 -26670.752 0.351 0.315
Chain 1: 1500 -23247.657 0.338 0.315
Chain 1: 1600 -22462.847 0.288 0.291
Chain 1: 1700 -21331.122 0.199 0.290
Chain 1: 1800 -21274.755 0.161 0.147
Chain 1: 1900 -21601.749 0.134 0.053
Chain 1: 2000 -20108.608 0.112 0.053
Chain 1: 2100 -20347.195 0.082 0.035
Chain 1: 2200 -20574.702 0.081 0.035
Chain 1: 2300 -20190.795 0.038 0.019
Chain 1: 2400 -19962.515 0.038 0.019
Chain 1: 2500 -19764.698 0.024 0.015
Chain 1: 2600 -19393.767 0.023 0.015
Chain 1: 2700 -19350.465 0.018 0.012
Chain 1: 2800 -19066.948 0.019 0.015
Chain 1: 2900 -19348.707 0.019 0.015
Chain 1: 3000 -19334.824 0.011 0.012
Chain 1: 3100 -19419.904 0.011 0.011
Chain 1: 3200 -19109.960 0.011 0.015
Chain 1: 3300 -19315.210 0.010 0.011
Chain 1: 3400 -18789.011 0.012 0.015
Chain 1: 3500 -19402.560 0.014 0.015
Chain 1: 3600 -18707.136 0.016 0.015
Chain 1: 3700 -19095.488 0.018 0.016
Chain 1: 3800 -18051.897 0.022 0.020
Chain 1: 3900 -18047.984 0.021 0.020
Chain 1: 4000 -18165.290 0.021 0.020
Chain 1: 4100 -18078.845 0.021 0.020
Chain 1: 4200 -17894.409 0.021 0.020
Chain 1: 4300 -18033.272 0.020 0.020
Chain 1: 4400 -17989.499 0.018 0.010
Chain 1: 4500 -17891.959 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001298 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.98 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48731.051 1.000 1.000
Chain 1: 200 -14867.582 1.639 2.278
Chain 1: 300 -13771.824 1.119 1.000
Chain 1: 400 -15019.064 0.860 1.000
Chain 1: 500 -18201.589 0.723 0.175
Chain 1: 600 -12527.214 0.678 0.453
Chain 1: 700 -11944.881 0.588 0.175
Chain 1: 800 -17501.487 0.554 0.317
Chain 1: 900 -19248.872 0.503 0.175
Chain 1: 1000 -10939.815 0.528 0.317
Chain 1: 1100 -10983.678 0.429 0.175
Chain 1: 1200 -13891.052 0.222 0.175
Chain 1: 1300 -11840.951 0.231 0.175
Chain 1: 1400 -11636.830 0.225 0.175
Chain 1: 1500 -12228.698 0.212 0.173
Chain 1: 1600 -10417.014 0.184 0.173
Chain 1: 1700 -21829.515 0.232 0.174
Chain 1: 1800 -12734.784 0.271 0.174
Chain 1: 1900 -12543.083 0.264 0.174
Chain 1: 2000 -10962.822 0.202 0.173
Chain 1: 2100 -11361.349 0.205 0.173
Chain 1: 2200 -9669.418 0.202 0.173
Chain 1: 2300 -11847.565 0.203 0.174
Chain 1: 2400 -10572.366 0.213 0.174
Chain 1: 2500 -9602.808 0.219 0.174
Chain 1: 2600 -10060.911 0.206 0.144
Chain 1: 2700 -10942.006 0.162 0.121
Chain 1: 2800 -10130.359 0.098 0.101
Chain 1: 2900 -9505.235 0.103 0.101
Chain 1: 3000 -14284.712 0.122 0.101
Chain 1: 3100 -9894.374 0.163 0.121
Chain 1: 3200 -12268.634 0.165 0.121
Chain 1: 3300 -17690.370 0.177 0.121
Chain 1: 3400 -10734.729 0.230 0.194
Chain 1: 3500 -8968.757 0.240 0.197
Chain 1: 3600 -10020.565 0.245 0.197
Chain 1: 3700 -9408.169 0.244 0.197
Chain 1: 3800 -9080.898 0.240 0.197
Chain 1: 3900 -9878.764 0.241 0.197
Chain 1: 4000 -9265.272 0.214 0.194
Chain 1: 4100 -8741.642 0.176 0.105
Chain 1: 4200 -9244.411 0.162 0.081
Chain 1: 4300 -8601.206 0.139 0.075
Chain 1: 4400 -10994.556 0.096 0.075
Chain 1: 4500 -9074.610 0.097 0.075
Chain 1: 4600 -12278.445 0.113 0.075
Chain 1: 4700 -11917.299 0.109 0.075
Chain 1: 4800 -8611.917 0.144 0.081
Chain 1: 4900 -11510.136 0.161 0.212
Chain 1: 5000 -16406.673 0.184 0.218
Chain 1: 5100 -10443.425 0.235 0.252
Chain 1: 5200 -8657.537 0.251 0.252
Chain 1: 5300 -11504.443 0.268 0.252
Chain 1: 5400 -8661.225 0.279 0.261
Chain 1: 5500 -11894.618 0.285 0.272
Chain 1: 5600 -12971.822 0.267 0.272
Chain 1: 5700 -8913.371 0.310 0.298
Chain 1: 5800 -9511.694 0.278 0.272
Chain 1: 5900 -9381.156 0.254 0.272
Chain 1: 6000 -8344.051 0.236 0.247
Chain 1: 6100 -8739.245 0.184 0.206
Chain 1: 6200 -8331.621 0.168 0.124
Chain 1: 6300 -9792.937 0.158 0.124
Chain 1: 6400 -13215.008 0.151 0.124
Chain 1: 6500 -8411.064 0.181 0.124
Chain 1: 6600 -8415.930 0.173 0.124
Chain 1: 6700 -8624.045 0.130 0.063
Chain 1: 6800 -8318.775 0.127 0.049
Chain 1: 6900 -10834.086 0.149 0.124
Chain 1: 7000 -9457.804 0.151 0.146
Chain 1: 7100 -10760.060 0.159 0.146
Chain 1: 7200 -11731.952 0.162 0.146
Chain 1: 7300 -10375.351 0.160 0.131
Chain 1: 7400 -8809.386 0.152 0.131
Chain 1: 7500 -10825.605 0.114 0.131
Chain 1: 7600 -8386.301 0.143 0.146
Chain 1: 7700 -8399.837 0.141 0.146
Chain 1: 7800 -12773.956 0.171 0.178
Chain 1: 7900 -10820.006 0.166 0.178
Chain 1: 8000 -8364.248 0.181 0.181
Chain 1: 8100 -11813.499 0.198 0.186
Chain 1: 8200 -11841.962 0.190 0.186
Chain 1: 8300 -8526.922 0.216 0.291
Chain 1: 8400 -9976.225 0.212 0.291
Chain 1: 8500 -8535.496 0.211 0.291
Chain 1: 8600 -9427.937 0.191 0.181
Chain 1: 8700 -10894.115 0.204 0.181
Chain 1: 8800 -8364.355 0.200 0.181
Chain 1: 8900 -9254.277 0.192 0.169
Chain 1: 9000 -8572.366 0.170 0.145
Chain 1: 9100 -8274.587 0.145 0.135
Chain 1: 9200 -11513.088 0.173 0.145
Chain 1: 9300 -8275.077 0.173 0.145
Chain 1: 9400 -8388.793 0.160 0.135
Chain 1: 9500 -9171.200 0.151 0.096
Chain 1: 9600 -9415.343 0.145 0.096
Chain 1: 9700 -8822.626 0.138 0.085
Chain 1: 9800 -11174.520 0.129 0.085
Chain 1: 9900 -11029.296 0.120 0.080
Chain 1: 10000 -8117.130 0.148 0.085
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001425 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.25 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57198.885 1.000 1.000
Chain 1: 200 -17516.900 1.633 2.265
Chain 1: 300 -8731.047 1.424 1.006
Chain 1: 400 -8325.134 1.080 1.006
Chain 1: 500 -8280.385 0.865 1.000
Chain 1: 600 -8609.434 0.727 1.000
Chain 1: 700 -7814.451 0.638 0.102
Chain 1: 800 -8113.021 0.563 0.102
Chain 1: 900 -7792.685 0.505 0.049
Chain 1: 1000 -7823.765 0.455 0.049
Chain 1: 1100 -7709.259 0.356 0.041
Chain 1: 1200 -7606.319 0.131 0.038
Chain 1: 1300 -8017.921 0.036 0.038
Chain 1: 1400 -7784.493 0.034 0.037
Chain 1: 1500 -7584.398 0.036 0.037
Chain 1: 1600 -7725.281 0.034 0.030
Chain 1: 1700 -7484.300 0.027 0.030
Chain 1: 1800 -7624.717 0.025 0.026
Chain 1: 1900 -7539.706 0.022 0.018
Chain 1: 2000 -7585.621 0.022 0.018
Chain 1: 2100 -7566.127 0.021 0.018
Chain 1: 2200 -7683.796 0.021 0.018
Chain 1: 2300 -7520.451 0.018 0.018
Chain 1: 2400 -7628.852 0.017 0.018
Chain 1: 2500 -7544.587 0.015 0.015
Chain 1: 2600 -7530.104 0.013 0.014
Chain 1: 2700 -7519.576 0.010 0.011
Chain 1: 2800 -7558.985 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002963 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.63 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85598.217 1.000 1.000
Chain 1: 200 -13515.732 3.167 5.333
Chain 1: 300 -9859.557 2.235 1.000
Chain 1: 400 -10877.565 1.699 1.000
Chain 1: 500 -8850.090 1.405 0.371
Chain 1: 600 -8317.183 1.182 0.371
Chain 1: 700 -8388.755 1.014 0.229
Chain 1: 800 -8755.976 0.893 0.229
Chain 1: 900 -8609.781 0.795 0.094
Chain 1: 1000 -8430.040 0.718 0.094
Chain 1: 1100 -8469.671 0.618 0.064 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8200.006 0.088 0.042
Chain 1: 1300 -8485.380 0.055 0.034
Chain 1: 1400 -8538.939 0.046 0.033
Chain 1: 1500 -8427.309 0.024 0.021
Chain 1: 1600 -8537.389 0.019 0.017
Chain 1: 1700 -8611.107 0.019 0.017
Chain 1: 1800 -8189.746 0.020 0.017
Chain 1: 1900 -8289.392 0.020 0.013
Chain 1: 2000 -8263.684 0.018 0.013
Chain 1: 2100 -8388.723 0.019 0.013
Chain 1: 2200 -8194.619 0.018 0.013
Chain 1: 2300 -8284.157 0.016 0.013
Chain 1: 2400 -8353.211 0.016 0.013
Chain 1: 2500 -8299.379 0.015 0.012
Chain 1: 2600 -8300.346 0.014 0.011
Chain 1: 2700 -8217.272 0.014 0.011
Chain 1: 2800 -8177.707 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003113 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.13 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8378723.814 1.000 1.000
Chain 1: 200 -1579562.103 2.652 4.304
Chain 1: 300 -890945.904 2.026 1.000
Chain 1: 400 -458139.571 1.756 1.000
Chain 1: 500 -358836.062 1.460 0.945
Chain 1: 600 -233832.418 1.306 0.945
Chain 1: 700 -119675.445 1.255 0.945
Chain 1: 800 -86791.350 1.146 0.945
Chain 1: 900 -67058.417 1.051 0.773
Chain 1: 1000 -51792.269 0.976 0.773
Chain 1: 1100 -39207.781 0.908 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38376.864 0.479 0.379
Chain 1: 1300 -26268.002 0.448 0.379
Chain 1: 1400 -25981.295 0.355 0.321
Chain 1: 1500 -22551.637 0.342 0.321
Chain 1: 1600 -21763.234 0.292 0.295
Chain 1: 1700 -20629.007 0.203 0.294
Chain 1: 1800 -20571.340 0.165 0.152
Chain 1: 1900 -20897.558 0.137 0.055
Chain 1: 2000 -19404.323 0.115 0.055
Chain 1: 2100 -19642.917 0.084 0.036
Chain 1: 2200 -19870.193 0.083 0.036
Chain 1: 2300 -19486.655 0.039 0.020
Chain 1: 2400 -19258.636 0.039 0.020
Chain 1: 2500 -19060.968 0.025 0.016
Chain 1: 2600 -18690.807 0.024 0.016
Chain 1: 2700 -18647.602 0.018 0.012
Chain 1: 2800 -18364.592 0.020 0.015
Chain 1: 2900 -18645.939 0.020 0.015
Chain 1: 3000 -18632.066 0.012 0.012
Chain 1: 3100 -18717.112 0.011 0.012
Chain 1: 3200 -18407.653 0.012 0.015
Chain 1: 3300 -18612.453 0.011 0.012
Chain 1: 3400 -18087.289 0.013 0.015
Chain 1: 3500 -18699.449 0.015 0.015
Chain 1: 3600 -18005.763 0.017 0.015
Chain 1: 3700 -18392.919 0.018 0.017
Chain 1: 3800 -17352.155 0.023 0.021
Chain 1: 3900 -17348.330 0.021 0.021
Chain 1: 4000 -17465.580 0.022 0.021
Chain 1: 4100 -17379.381 0.022 0.021
Chain 1: 4200 -17195.480 0.021 0.021
Chain 1: 4300 -17333.941 0.021 0.021
Chain 1: 4400 -17290.682 0.019 0.011
Chain 1: 4500 -17193.229 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00164 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 16.4 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12397.993 1.000 1.000
Chain 1: 200 -9378.292 0.661 1.000
Chain 1: 300 -8070.103 0.495 0.322
Chain 1: 400 -8224.357 0.376 0.322
Chain 1: 500 -8085.356 0.304 0.162
Chain 1: 600 -8011.375 0.255 0.162
Chain 1: 700 -7865.627 0.221 0.019
Chain 1: 800 -7900.607 0.194 0.019
Chain 1: 900 -7936.090 0.173 0.019
Chain 1: 1000 -7905.694 0.156 0.019
Chain 1: 1100 -8002.684 0.057 0.017
Chain 1: 1200 -7907.802 0.026 0.012
Chain 1: 1300 -7850.499 0.011 0.012
Chain 1: 1400 -7873.156 0.009 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001377 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.77 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -45519.113 1.000 1.000
Chain 1: 200 -15533.273 1.465 1.930
Chain 1: 300 -8705.205 1.238 1.000
Chain 1: 400 -8711.614 0.929 1.000
Chain 1: 500 -8021.248 0.760 0.784
Chain 1: 600 -8607.645 0.645 0.784
Chain 1: 700 -8272.933 0.559 0.086
Chain 1: 800 -8087.995 0.492 0.086
Chain 1: 900 -8080.301 0.437 0.068
Chain 1: 1000 -7847.908 0.396 0.068
Chain 1: 1100 -7808.532 0.297 0.040
Chain 1: 1200 -7582.783 0.107 0.030
Chain 1: 1300 -7790.200 0.031 0.030
Chain 1: 1400 -7913.450 0.033 0.030
Chain 1: 1500 -7623.291 0.028 0.030
Chain 1: 1600 -7781.494 0.023 0.027
Chain 1: 1700 -7536.126 0.022 0.027
Chain 1: 1800 -7614.084 0.021 0.027
Chain 1: 1900 -7727.971 0.022 0.027
Chain 1: 2000 -7584.953 0.021 0.020
Chain 1: 2100 -7622.138 0.021 0.020
Chain 1: 2200 -7723.080 0.019 0.019
Chain 1: 2300 -7615.423 0.018 0.016
Chain 1: 2400 -7663.014 0.017 0.015
Chain 1: 2500 -7585.608 0.015 0.014
Chain 1: 2600 -7546.743 0.013 0.013
Chain 1: 2700 -7549.313 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00357 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.7 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87100.641 1.000 1.000
Chain 1: 200 -13532.860 3.218 5.436
Chain 1: 300 -9873.614 2.269 1.000
Chain 1: 400 -10708.515 1.721 1.000
Chain 1: 500 -8849.371 1.419 0.371
Chain 1: 600 -8681.626 1.186 0.371
Chain 1: 700 -8722.190 1.017 0.210
Chain 1: 800 -9030.907 0.894 0.210
Chain 1: 900 -8716.432 0.799 0.078
Chain 1: 1000 -8630.858 0.720 0.078
Chain 1: 1100 -8607.924 0.620 0.036 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8438.456 0.079 0.034
Chain 1: 1300 -8532.946 0.043 0.020
Chain 1: 1400 -8541.239 0.035 0.019
Chain 1: 1500 -8418.669 0.015 0.015
Chain 1: 1600 -8536.314 0.015 0.014
Chain 1: 1700 -8619.876 0.015 0.014
Chain 1: 1800 -8201.755 0.017 0.014
Chain 1: 1900 -8299.151 0.015 0.012
Chain 1: 2000 -8273.234 0.014 0.012
Chain 1: 2100 -8397.078 0.015 0.014
Chain 1: 2200 -8211.664 0.015 0.014
Chain 1: 2300 -8293.938 0.015 0.014
Chain 1: 2400 -8363.496 0.016 0.014
Chain 1: 2500 -8309.400 0.015 0.012
Chain 1: 2600 -8309.652 0.014 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003178 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.78 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8399801.904 1.000 1.000
Chain 1: 200 -1583577.699 2.652 4.304
Chain 1: 300 -892156.435 2.026 1.000
Chain 1: 400 -458911.816 1.756 1.000
Chain 1: 500 -359387.819 1.460 0.944
Chain 1: 600 -233892.694 1.306 0.944
Chain 1: 700 -119677.504 1.256 0.944
Chain 1: 800 -86757.935 1.146 0.944
Chain 1: 900 -67014.960 1.052 0.775
Chain 1: 1000 -51746.904 0.976 0.775
Chain 1: 1100 -39168.500 0.908 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38335.699 0.480 0.379
Chain 1: 1300 -26244.302 0.448 0.379
Chain 1: 1400 -25957.541 0.355 0.321
Chain 1: 1500 -22532.759 0.343 0.321
Chain 1: 1600 -21745.198 0.293 0.295
Chain 1: 1700 -20613.914 0.203 0.295
Chain 1: 1800 -20556.662 0.165 0.152
Chain 1: 1900 -20882.778 0.137 0.055
Chain 1: 2000 -19391.097 0.115 0.055
Chain 1: 2100 -19629.680 0.084 0.036
Chain 1: 2200 -19856.532 0.083 0.036
Chain 1: 2300 -19473.347 0.039 0.020
Chain 1: 2400 -19245.401 0.039 0.020
Chain 1: 2500 -19047.488 0.025 0.016
Chain 1: 2600 -18677.598 0.024 0.016
Chain 1: 2700 -18634.480 0.018 0.012
Chain 1: 2800 -18351.412 0.020 0.015
Chain 1: 2900 -18632.716 0.020 0.015
Chain 1: 3000 -18618.825 0.012 0.012
Chain 1: 3100 -18703.863 0.011 0.012
Chain 1: 3200 -18394.494 0.012 0.015
Chain 1: 3300 -18599.221 0.011 0.012
Chain 1: 3400 -18074.137 0.013 0.015
Chain 1: 3500 -18686.086 0.015 0.015
Chain 1: 3600 -17992.653 0.017 0.015
Chain 1: 3700 -18379.631 0.019 0.017
Chain 1: 3800 -17339.150 0.023 0.021
Chain 1: 3900 -17335.290 0.021 0.021
Chain 1: 4000 -17452.577 0.022 0.021
Chain 1: 4100 -17366.376 0.022 0.021
Chain 1: 4200 -17182.521 0.022 0.021
Chain 1: 4300 -17320.963 0.021 0.021
Chain 1: 4400 -17277.789 0.019 0.011
Chain 1: 4500 -17180.293 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001285 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.85 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12100.285 1.000 1.000
Chain 1: 200 -8888.814 0.681 1.000
Chain 1: 300 -7875.031 0.497 0.361
Chain 1: 400 -8029.545 0.377 0.361
Chain 1: 500 -7923.225 0.305 0.129
Chain 1: 600 -7784.051 0.257 0.129
Chain 1: 700 -7715.358 0.221 0.019
Chain 1: 800 -7720.425 0.194 0.019
Chain 1: 900 -7700.205 0.173 0.018
Chain 1: 1000 -7775.621 0.156 0.018
Chain 1: 1100 -7831.696 0.057 0.013
Chain 1: 1200 -7728.695 0.022 0.013
Chain 1: 1300 -7729.288 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001416 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.16 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56695.550 1.000 1.000
Chain 1: 200 -17032.271 1.664 2.329
Chain 1: 300 -8535.891 1.441 1.000
Chain 1: 400 -7816.584 1.104 1.000
Chain 1: 500 -8230.505 0.893 0.995
Chain 1: 600 -8003.823 0.749 0.995
Chain 1: 700 -7798.651 0.646 0.092
Chain 1: 800 -7974.857 0.568 0.092
Chain 1: 900 -7845.257 0.507 0.050
Chain 1: 1000 -7765.707 0.457 0.050
Chain 1: 1100 -7621.240 0.359 0.028
Chain 1: 1200 -7594.005 0.126 0.026
Chain 1: 1300 -7739.401 0.029 0.022
Chain 1: 1400 -7802.583 0.020 0.019
Chain 1: 1500 -7510.309 0.019 0.019
Chain 1: 1600 -7685.752 0.019 0.019
Chain 1: 1700 -7444.568 0.019 0.019
Chain 1: 1800 -7526.880 0.018 0.019
Chain 1: 1900 -7487.129 0.017 0.019
Chain 1: 2000 -7516.764 0.016 0.019
Chain 1: 2100 -7535.326 0.015 0.011
Chain 1: 2200 -7613.592 0.015 0.011
Chain 1: 2300 -7527.381 0.015 0.011
Chain 1: 2400 -7563.792 0.014 0.011
Chain 1: 2500 -7403.416 0.013 0.011
Chain 1: 2600 -7462.072 0.011 0.010
Chain 1: 2700 -7501.794 0.008 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004453 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 44.53 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86390.328 1.000 1.000
Chain 1: 200 -13148.776 3.285 5.570
Chain 1: 300 -9597.679 2.313 1.000
Chain 1: 400 -10538.916 1.757 1.000
Chain 1: 500 -8503.395 1.454 0.370
Chain 1: 600 -8136.447 1.219 0.370
Chain 1: 700 -8358.435 1.049 0.239
Chain 1: 800 -8557.104 0.920 0.239
Chain 1: 900 -8430.130 0.820 0.089
Chain 1: 1000 -8186.196 0.741 0.089
Chain 1: 1100 -8471.254 0.644 0.045 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8110.461 0.092 0.044
Chain 1: 1300 -8329.652 0.057 0.034
Chain 1: 1400 -8317.598 0.049 0.030
Chain 1: 1500 -8223.771 0.026 0.027
Chain 1: 1600 -8320.581 0.022 0.026
Chain 1: 1700 -8409.055 0.021 0.023
Chain 1: 1800 -8020.551 0.023 0.026
Chain 1: 1900 -8123.016 0.023 0.026
Chain 1: 2000 -8092.978 0.020 0.013
Chain 1: 2100 -8224.495 0.019 0.013
Chain 1: 2200 -8010.470 0.017 0.013
Chain 1: 2300 -8152.322 0.016 0.013
Chain 1: 2400 -8164.753 0.016 0.013
Chain 1: 2500 -8132.784 0.015 0.013
Chain 1: 2600 -8132.634 0.014 0.013
Chain 1: 2700 -8040.812 0.014 0.013
Chain 1: 2800 -8016.732 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003076 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.76 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8431057.830 1.000 1.000
Chain 1: 200 -1585790.998 2.658 4.317
Chain 1: 300 -889704.598 2.033 1.000
Chain 1: 400 -456830.848 1.762 1.000
Chain 1: 500 -356917.622 1.465 0.948
Chain 1: 600 -231969.646 1.311 0.948
Chain 1: 700 -118475.017 1.260 0.948
Chain 1: 800 -85793.410 1.151 0.948
Chain 1: 900 -66196.668 1.056 0.782
Chain 1: 1000 -51043.755 0.980 0.782
Chain 1: 1100 -38578.009 0.912 0.539 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37756.680 0.483 0.381
Chain 1: 1300 -25773.592 0.451 0.381
Chain 1: 1400 -25496.709 0.357 0.323
Chain 1: 1500 -22100.785 0.344 0.323
Chain 1: 1600 -21322.115 0.294 0.297
Chain 1: 1700 -20203.194 0.204 0.296
Chain 1: 1800 -20148.920 0.166 0.154
Chain 1: 1900 -20474.622 0.138 0.055
Chain 1: 2000 -18990.883 0.116 0.055
Chain 1: 2100 -19228.785 0.085 0.037
Chain 1: 2200 -19454.378 0.084 0.037
Chain 1: 2300 -19072.492 0.040 0.020
Chain 1: 2400 -18844.854 0.040 0.020
Chain 1: 2500 -18646.777 0.026 0.016
Chain 1: 2600 -18277.525 0.024 0.016
Chain 1: 2700 -18234.764 0.019 0.012
Chain 1: 2800 -17951.815 0.020 0.016
Chain 1: 2900 -18232.752 0.020 0.015
Chain 1: 3000 -18218.955 0.012 0.012
Chain 1: 3100 -18303.854 0.011 0.012
Chain 1: 3200 -17994.924 0.012 0.015
Chain 1: 3300 -18199.385 0.011 0.012
Chain 1: 3400 -17674.943 0.013 0.015
Chain 1: 3500 -18285.774 0.015 0.016
Chain 1: 3600 -17593.810 0.017 0.016
Chain 1: 3700 -17979.558 0.019 0.017
Chain 1: 3800 -16941.321 0.023 0.021
Chain 1: 3900 -16937.518 0.022 0.021
Chain 1: 4000 -17054.824 0.023 0.021
Chain 1: 4100 -16968.666 0.023 0.021
Chain 1: 4200 -16785.401 0.022 0.021
Chain 1: 4300 -16923.453 0.022 0.021
Chain 1: 4400 -16880.637 0.019 0.011
Chain 1: 4500 -16783.240 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001125 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 11.25 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49201.027 1.000 1.000
Chain 1: 200 -53813.066 0.543 1.000
Chain 1: 300 -21674.170 0.856 1.000
Chain 1: 400 -22026.059 0.646 1.000
Chain 1: 500 -23020.587 0.526 0.086
Chain 1: 600 -15249.809 0.523 0.510
Chain 1: 700 -16814.284 0.461 0.093
Chain 1: 800 -15426.778 0.415 0.093
Chain 1: 900 -14192.227 0.379 0.090
Chain 1: 1000 -12537.426 0.354 0.093
Chain 1: 1100 -20206.321 0.292 0.093
Chain 1: 1200 -11004.634 0.367 0.132
Chain 1: 1300 -10919.780 0.219 0.093
Chain 1: 1400 -14509.926 0.243 0.132
Chain 1: 1500 -11197.430 0.268 0.247
Chain 1: 1600 -9942.458 0.229 0.132
Chain 1: 1700 -12985.626 0.244 0.234
Chain 1: 1800 -10528.024 0.258 0.234
Chain 1: 1900 -16888.453 0.287 0.247
Chain 1: 2000 -14707.994 0.289 0.247
Chain 1: 2100 -9570.227 0.304 0.247
Chain 1: 2200 -13788.064 0.251 0.247
Chain 1: 2300 -9364.617 0.298 0.296
Chain 1: 2400 -9856.274 0.278 0.296
Chain 1: 2500 -11555.379 0.263 0.234
Chain 1: 2600 -11797.691 0.253 0.234
Chain 1: 2700 -9356.584 0.255 0.261
Chain 1: 2800 -10933.546 0.246 0.261
Chain 1: 2900 -9570.694 0.223 0.148
Chain 1: 3000 -8981.836 0.215 0.147
Chain 1: 3100 -8760.197 0.163 0.144
Chain 1: 3200 -10110.617 0.146 0.142
Chain 1: 3300 -10479.183 0.102 0.134
Chain 1: 3400 -9875.101 0.104 0.134
Chain 1: 3500 -9583.279 0.092 0.066
Chain 1: 3600 -11537.846 0.107 0.134
Chain 1: 3700 -12097.233 0.085 0.066
Chain 1: 3800 -9849.241 0.094 0.066
Chain 1: 3900 -9554.771 0.083 0.061
Chain 1: 4000 -9756.999 0.078 0.046
Chain 1: 4100 -8805.023 0.086 0.061
Chain 1: 4200 -8883.772 0.074 0.046
Chain 1: 4300 -9122.511 0.073 0.046
Chain 1: 4400 -12981.521 0.097 0.046
Chain 1: 4500 -9618.532 0.129 0.108
Chain 1: 4600 -9121.950 0.117 0.054
Chain 1: 4700 -8586.424 0.119 0.062
Chain 1: 4800 -8790.417 0.098 0.054
Chain 1: 4900 -12792.907 0.126 0.062
Chain 1: 5000 -14675.946 0.137 0.108
Chain 1: 5100 -8808.220 0.193 0.128
Chain 1: 5200 -8764.668 0.193 0.128
Chain 1: 5300 -9137.218 0.194 0.128
Chain 1: 5400 -8949.583 0.166 0.062
Chain 1: 5500 -13171.289 0.163 0.062
Chain 1: 5600 -15966.177 0.176 0.128
Chain 1: 5700 -12395.255 0.198 0.175
Chain 1: 5800 -8698.235 0.238 0.288
Chain 1: 5900 -15114.166 0.249 0.288
Chain 1: 6000 -8843.131 0.308 0.321
Chain 1: 6100 -9378.621 0.247 0.288
Chain 1: 6200 -8260.788 0.260 0.288
Chain 1: 6300 -8982.841 0.264 0.288
Chain 1: 6400 -9292.733 0.265 0.288
Chain 1: 6500 -8875.605 0.237 0.175
Chain 1: 6600 -11385.990 0.242 0.220
Chain 1: 6700 -8803.849 0.243 0.220
Chain 1: 6800 -12193.590 0.228 0.220
Chain 1: 6900 -12593.037 0.189 0.135
Chain 1: 7000 -12781.467 0.119 0.080
Chain 1: 7100 -13309.959 0.117 0.080
Chain 1: 7200 -10297.705 0.133 0.080
Chain 1: 7300 -10893.830 0.131 0.055
Chain 1: 7400 -11417.551 0.132 0.055
Chain 1: 7500 -8736.672 0.158 0.220
Chain 1: 7600 -8865.880 0.137 0.055
Chain 1: 7700 -8486.884 0.112 0.046
Chain 1: 7800 -13191.711 0.120 0.046
Chain 1: 7900 -8572.088 0.171 0.055
Chain 1: 8000 -8852.620 0.173 0.055
Chain 1: 8100 -9281.152 0.173 0.055
Chain 1: 8200 -9710.265 0.148 0.046
Chain 1: 8300 -9829.210 0.144 0.046
Chain 1: 8400 -11847.639 0.157 0.046
Chain 1: 8500 -8307.367 0.169 0.046
Chain 1: 8600 -9899.470 0.183 0.161
Chain 1: 8700 -8914.185 0.190 0.161
Chain 1: 8800 -8847.894 0.155 0.111
Chain 1: 8900 -9223.478 0.105 0.046
Chain 1: 9000 -9792.643 0.108 0.058
Chain 1: 9100 -9325.004 0.108 0.058
Chain 1: 9200 -11847.489 0.125 0.111
Chain 1: 9300 -8280.344 0.167 0.161
Chain 1: 9400 -8231.333 0.150 0.111
Chain 1: 9500 -8305.803 0.109 0.058
Chain 1: 9600 -9529.161 0.105 0.058
Chain 1: 9700 -8525.302 0.106 0.058
Chain 1: 9800 -8219.706 0.109 0.058
Chain 1: 9900 -9669.870 0.120 0.118
Chain 1: 10000 -8478.972 0.128 0.128
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001392 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.92 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61806.812 1.000 1.000
Chain 1: 200 -17969.050 1.720 2.440
Chain 1: 300 -8934.587 1.484 1.011
Chain 1: 400 -8921.501 1.113 1.011
Chain 1: 500 -8712.218 0.895 1.000
Chain 1: 600 -8974.331 0.751 1.000
Chain 1: 700 -7832.709 0.664 0.146
Chain 1: 800 -8322.844 0.589 0.146
Chain 1: 900 -8142.223 0.526 0.059
Chain 1: 1000 -7902.974 0.476 0.059
Chain 1: 1100 -7920.282 0.376 0.030
Chain 1: 1200 -7698.618 0.135 0.029
Chain 1: 1300 -7697.309 0.034 0.029
Chain 1: 1400 -8122.626 0.039 0.029
Chain 1: 1500 -7610.469 0.044 0.030
Chain 1: 1600 -7795.293 0.043 0.030
Chain 1: 1700 -7565.650 0.032 0.030
Chain 1: 1800 -7592.381 0.026 0.029
Chain 1: 1900 -7629.766 0.024 0.029
Chain 1: 2000 -7675.005 0.022 0.024
Chain 1: 2100 -7631.595 0.022 0.024
Chain 1: 2200 -7732.453 0.021 0.013
Chain 1: 2300 -7590.533 0.023 0.019
Chain 1: 2400 -7639.312 0.018 0.013
Chain 1: 2500 -7642.981 0.011 0.006 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.0032 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85915.486 1.000 1.000
Chain 1: 200 -13650.497 3.147 5.294
Chain 1: 300 -10009.368 2.219 1.000
Chain 1: 400 -10951.667 1.686 1.000
Chain 1: 500 -8982.507 1.393 0.364
Chain 1: 600 -8408.704 1.172 0.364
Chain 1: 700 -8571.241 1.007 0.219
Chain 1: 800 -8937.776 0.886 0.219
Chain 1: 900 -8975.169 0.788 0.086
Chain 1: 1000 -8732.598 0.712 0.086
Chain 1: 1100 -8805.949 0.613 0.068 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8327.946 0.089 0.057
Chain 1: 1300 -8684.181 0.057 0.041
Chain 1: 1400 -8693.292 0.049 0.041
Chain 1: 1500 -8564.234 0.028 0.028
Chain 1: 1600 -8672.718 0.023 0.019
Chain 1: 1700 -8750.532 0.022 0.015
Chain 1: 1800 -8329.858 0.023 0.015
Chain 1: 1900 -8429.017 0.023 0.015
Chain 1: 2000 -8403.117 0.021 0.013
Chain 1: 2100 -8527.742 0.022 0.015
Chain 1: 2200 -8336.209 0.018 0.015
Chain 1: 2300 -8423.630 0.015 0.013
Chain 1: 2400 -8492.888 0.016 0.013
Chain 1: 2500 -8438.992 0.015 0.012
Chain 1: 2600 -8439.705 0.014 0.010
Chain 1: 2700 -8356.715 0.014 0.010
Chain 1: 2800 -8317.581 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003354 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.54 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8404188.139 1.000 1.000
Chain 1: 200 -1580562.464 2.659 4.317
Chain 1: 300 -889544.977 2.031 1.000
Chain 1: 400 -457190.310 1.760 1.000
Chain 1: 500 -357709.230 1.464 0.946
Chain 1: 600 -232916.469 1.309 0.946
Chain 1: 700 -119277.302 1.258 0.946
Chain 1: 800 -86544.443 1.148 0.946
Chain 1: 900 -66905.964 1.053 0.777
Chain 1: 1000 -51718.611 0.977 0.777
Chain 1: 1100 -39211.940 0.909 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38391.106 0.479 0.378
Chain 1: 1300 -26353.336 0.447 0.378
Chain 1: 1400 -26074.148 0.354 0.319
Chain 1: 1500 -22663.314 0.341 0.319
Chain 1: 1600 -21881.156 0.291 0.294
Chain 1: 1700 -20754.915 0.201 0.294
Chain 1: 1800 -20699.449 0.164 0.151
Chain 1: 1900 -21025.656 0.136 0.054
Chain 1: 2000 -19537.079 0.114 0.054
Chain 1: 2100 -19775.247 0.084 0.036
Chain 1: 2200 -20001.875 0.083 0.036
Chain 1: 2300 -19618.941 0.039 0.020
Chain 1: 2400 -19391.028 0.039 0.020
Chain 1: 2500 -19193.149 0.025 0.016
Chain 1: 2600 -18823.097 0.023 0.016
Chain 1: 2700 -18780.075 0.018 0.012
Chain 1: 2800 -18496.971 0.019 0.015
Chain 1: 2900 -18778.204 0.019 0.015
Chain 1: 3000 -18764.351 0.012 0.012
Chain 1: 3100 -18849.373 0.011 0.012
Chain 1: 3200 -18539.964 0.012 0.015
Chain 1: 3300 -18744.784 0.011 0.012
Chain 1: 3400 -18219.623 0.012 0.015
Chain 1: 3500 -18831.645 0.015 0.015
Chain 1: 3600 -18138.114 0.016 0.015
Chain 1: 3700 -18525.061 0.018 0.017
Chain 1: 3800 -17484.506 0.023 0.021
Chain 1: 3900 -17480.679 0.021 0.021
Chain 1: 4000 -17597.939 0.022 0.021
Chain 1: 4100 -17511.696 0.022 0.021
Chain 1: 4200 -17327.934 0.021 0.021
Chain 1: 4300 -17466.332 0.021 0.021
Chain 1: 4400 -17423.085 0.018 0.011
Chain 1: 4500 -17325.652 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001279 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.79 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12734.887 1.000 1.000
Chain 1: 200 -9492.107 0.671 1.000
Chain 1: 300 -8137.715 0.503 0.342
Chain 1: 400 -8254.192 0.381 0.342
Chain 1: 500 -8214.129 0.305 0.166
Chain 1: 600 -8087.003 0.257 0.166
Chain 1: 700 -7993.854 0.222 0.016
Chain 1: 800 -7996.913 0.194 0.016
Chain 1: 900 -7906.465 0.174 0.014
Chain 1: 1000 -8103.277 0.159 0.016
Chain 1: 1100 -8137.564 0.059 0.014
Chain 1: 1200 -8035.077 0.027 0.013
Chain 1: 1300 -7966.906 0.011 0.012
Chain 1: 1400 -7985.995 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001555 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.55 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -47345.163 1.000 1.000
Chain 1: 200 -15874.353 1.491 1.982
Chain 1: 300 -8616.915 1.275 1.000
Chain 1: 400 -8729.219 0.959 1.000
Chain 1: 500 -8644.137 0.769 0.842
Chain 1: 600 -7815.598 0.659 0.842
Chain 1: 700 -7822.977 0.565 0.106
Chain 1: 800 -8294.785 0.501 0.106
Chain 1: 900 -7840.979 0.452 0.058
Chain 1: 1000 -7646.261 0.409 0.058
Chain 1: 1100 -7719.396 0.310 0.057
Chain 1: 1200 -7722.436 0.112 0.025
Chain 1: 1300 -7684.526 0.028 0.013
Chain 1: 1400 -7534.082 0.029 0.020
Chain 1: 1500 -7484.947 0.029 0.020
Chain 1: 1600 -7701.240 0.021 0.020
Chain 1: 1700 -7475.292 0.024 0.025
Chain 1: 1800 -7514.892 0.019 0.020
Chain 1: 1900 -7521.997 0.013 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.005976 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 59.76 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85810.559 1.000 1.000
Chain 1: 200 -13659.071 3.141 5.282
Chain 1: 300 -10013.020 2.215 1.000
Chain 1: 400 -10814.519 1.680 1.000
Chain 1: 500 -8995.076 1.385 0.364
Chain 1: 600 -8660.116 1.160 0.364
Chain 1: 700 -8463.281 0.998 0.202
Chain 1: 800 -8929.026 0.880 0.202
Chain 1: 900 -8812.989 0.783 0.074
Chain 1: 1000 -8592.361 0.708 0.074
Chain 1: 1100 -8814.079 0.610 0.052 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8452.804 0.086 0.043
Chain 1: 1300 -8695.261 0.053 0.039
Chain 1: 1400 -8712.360 0.045 0.028
Chain 1: 1500 -8558.675 0.027 0.026
Chain 1: 1600 -8673.303 0.024 0.025
Chain 1: 1700 -8749.979 0.023 0.025
Chain 1: 1800 -8327.687 0.023 0.025
Chain 1: 1900 -8428.248 0.023 0.025
Chain 1: 2000 -8402.622 0.020 0.018
Chain 1: 2100 -8527.841 0.019 0.015
Chain 1: 2200 -8332.429 0.017 0.015
Chain 1: 2300 -8423.006 0.016 0.013
Chain 1: 2400 -8491.969 0.016 0.013
Chain 1: 2500 -8438.171 0.015 0.012
Chain 1: 2600 -8439.307 0.014 0.011
Chain 1: 2700 -8356.135 0.014 0.011
Chain 1: 2800 -8316.342 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00363 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 36.3 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8379501.601 1.000 1.000
Chain 1: 200 -1582188.639 2.648 4.296
Chain 1: 300 -891323.595 2.024 1.000
Chain 1: 400 -458092.054 1.754 1.000
Chain 1: 500 -358428.457 1.459 0.946
Chain 1: 600 -233469.106 1.305 0.946
Chain 1: 700 -119582.187 1.255 0.946
Chain 1: 800 -86719.362 1.145 0.946
Chain 1: 900 -67037.261 1.051 0.775
Chain 1: 1000 -51816.494 0.975 0.775
Chain 1: 1100 -39269.695 0.907 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38446.873 0.479 0.379
Chain 1: 1300 -26382.091 0.448 0.379
Chain 1: 1400 -26099.998 0.354 0.320
Chain 1: 1500 -22680.453 0.341 0.320
Chain 1: 1600 -21894.925 0.291 0.294
Chain 1: 1700 -20766.301 0.202 0.294
Chain 1: 1800 -20710.011 0.164 0.151
Chain 1: 1900 -21036.222 0.136 0.054
Chain 1: 2000 -19545.767 0.114 0.054
Chain 1: 2100 -19784.379 0.084 0.036
Chain 1: 2200 -20010.913 0.083 0.036
Chain 1: 2300 -19628.030 0.039 0.020
Chain 1: 2400 -19400.042 0.039 0.020
Chain 1: 2500 -19201.993 0.025 0.016
Chain 1: 2600 -18832.100 0.023 0.016
Chain 1: 2700 -18789.117 0.018 0.012
Chain 1: 2800 -18505.811 0.019 0.015
Chain 1: 2900 -18787.218 0.019 0.015
Chain 1: 3000 -18773.479 0.012 0.012
Chain 1: 3100 -18858.395 0.011 0.012
Chain 1: 3200 -18549.039 0.012 0.015
Chain 1: 3300 -18753.833 0.011 0.012
Chain 1: 3400 -18228.574 0.012 0.015
Chain 1: 3500 -18840.646 0.015 0.015
Chain 1: 3600 -18147.203 0.016 0.015
Chain 1: 3700 -18534.048 0.018 0.017
Chain 1: 3800 -17493.431 0.023 0.021
Chain 1: 3900 -17489.578 0.021 0.021
Chain 1: 4000 -17606.917 0.022 0.021
Chain 1: 4100 -17520.549 0.022 0.021
Chain 1: 4200 -17336.814 0.021 0.021
Chain 1: 4300 -17475.234 0.021 0.021
Chain 1: 4400 -17432.013 0.018 0.011
Chain 1: 4500 -17334.537 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001174 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 11.74 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12589.909 1.000 1.000
Chain 1: 200 -9613.304 0.655 1.000
Chain 1: 300 -8343.210 0.487 0.310
Chain 1: 400 -8498.848 0.370 0.310
Chain 1: 500 -8363.411 0.299 0.152
Chain 1: 600 -8229.700 0.252 0.152
Chain 1: 700 -8178.559 0.217 0.018
Chain 1: 800 -8130.492 0.191 0.018
Chain 1: 900 -8051.474 0.171 0.016
Chain 1: 1000 -8239.011 0.156 0.018
Chain 1: 1100 -8268.287 0.056 0.016
Chain 1: 1200 -8163.213 0.026 0.016
Chain 1: 1300 -8101.494 0.012 0.013
Chain 1: 1400 -8124.329 0.010 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001402 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.02 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -62408.612 1.000 1.000
Chain 1: 200 -18180.159 1.716 2.433
Chain 1: 300 -9029.905 1.482 1.013
Chain 1: 400 -9712.045 1.129 1.013
Chain 1: 500 -7849.319 0.951 1.000
Chain 1: 600 -8740.057 0.809 1.000
Chain 1: 700 -7937.546 0.708 0.237
Chain 1: 800 -8324.863 0.625 0.237
Chain 1: 900 -7712.215 0.565 0.102
Chain 1: 1000 -7922.233 0.511 0.102
Chain 1: 1100 -7887.933 0.411 0.101
Chain 1: 1200 -7615.843 0.172 0.079
Chain 1: 1300 -7863.309 0.073 0.070
Chain 1: 1400 -7962.700 0.068 0.047
Chain 1: 1500 -7623.310 0.048 0.045
Chain 1: 1600 -7704.318 0.039 0.036
Chain 1: 1700 -7607.236 0.030 0.031
Chain 1: 1800 -7648.444 0.026 0.027
Chain 1: 1900 -7665.460 0.019 0.013
Chain 1: 2000 -7767.818 0.017 0.013
Chain 1: 2100 -7641.324 0.018 0.013
Chain 1: 2200 -7762.822 0.016 0.013
Chain 1: 2300 -7625.810 0.015 0.013
Chain 1: 2400 -7685.190 0.015 0.013
Chain 1: 2500 -7603.524 0.011 0.013
Chain 1: 2600 -7566.934 0.011 0.013
Chain 1: 2700 -7533.567 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003132 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.32 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86424.919 1.000 1.000
Chain 1: 200 -13774.153 3.137 5.274
Chain 1: 300 -10133.763 2.211 1.000
Chain 1: 400 -11041.188 1.679 1.000
Chain 1: 500 -9108.688 1.386 0.359
Chain 1: 600 -8797.299 1.161 0.359
Chain 1: 700 -8630.871 0.998 0.212
Chain 1: 800 -8938.617 0.877 0.212
Chain 1: 900 -8950.836 0.780 0.082
Chain 1: 1000 -8749.446 0.704 0.082
Chain 1: 1100 -8937.162 0.606 0.035 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8579.851 0.083 0.035
Chain 1: 1300 -8795.114 0.049 0.034
Chain 1: 1400 -8814.675 0.041 0.024
Chain 1: 1500 -8690.924 0.022 0.023
Chain 1: 1600 -8796.025 0.019 0.021
Chain 1: 1700 -8878.000 0.018 0.021
Chain 1: 1800 -8459.358 0.020 0.021
Chain 1: 1900 -8557.815 0.021 0.021
Chain 1: 2000 -8531.728 0.019 0.014
Chain 1: 2100 -8655.706 0.018 0.014
Chain 1: 2200 -8469.123 0.016 0.014
Chain 1: 2300 -8552.366 0.015 0.012
Chain 1: 2400 -8621.889 0.015 0.012
Chain 1: 2500 -8567.857 0.015 0.012
Chain 1: 2600 -8568.187 0.013 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003611 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 36.11 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8406744.835 1.000 1.000
Chain 1: 200 -1583435.516 2.655 4.309
Chain 1: 300 -891651.722 2.028 1.000
Chain 1: 400 -458827.839 1.757 1.000
Chain 1: 500 -359393.436 1.461 0.943
Chain 1: 600 -234010.624 1.307 0.943
Chain 1: 700 -119867.392 1.256 0.943
Chain 1: 800 -87004.870 1.146 0.943
Chain 1: 900 -67262.168 1.052 0.776
Chain 1: 1000 -51999.116 0.976 0.776
Chain 1: 1100 -39428.078 0.908 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38595.928 0.479 0.378
Chain 1: 1300 -26504.861 0.447 0.378
Chain 1: 1400 -26218.133 0.354 0.319
Chain 1: 1500 -22794.750 0.341 0.319
Chain 1: 1600 -22008.251 0.291 0.294
Chain 1: 1700 -20876.464 0.201 0.294
Chain 1: 1800 -20819.334 0.164 0.150
Chain 1: 1900 -21145.493 0.136 0.054
Chain 1: 2000 -19653.982 0.114 0.054
Chain 1: 2100 -19892.248 0.083 0.036
Chain 1: 2200 -20119.380 0.082 0.036
Chain 1: 2300 -19736.032 0.039 0.019
Chain 1: 2400 -19508.093 0.039 0.019
Chain 1: 2500 -19310.399 0.025 0.015
Chain 1: 2600 -18940.209 0.023 0.015
Chain 1: 2700 -18897.035 0.018 0.012
Chain 1: 2800 -18614.057 0.019 0.015
Chain 1: 2900 -18895.404 0.019 0.015
Chain 1: 3000 -18881.370 0.012 0.012
Chain 1: 3100 -18966.458 0.011 0.012
Chain 1: 3200 -18657.001 0.012 0.015
Chain 1: 3300 -18861.839 0.011 0.012
Chain 1: 3400 -18336.686 0.012 0.015
Chain 1: 3500 -18948.736 0.015 0.015
Chain 1: 3600 -18255.186 0.016 0.015
Chain 1: 3700 -18642.271 0.018 0.017
Chain 1: 3800 -17601.654 0.023 0.021
Chain 1: 3900 -17597.854 0.021 0.021
Chain 1: 4000 -17715.096 0.022 0.021
Chain 1: 4100 -17628.893 0.022 0.021
Chain 1: 4200 -17445.047 0.021 0.021
Chain 1: 4300 -17583.452 0.021 0.021
Chain 1: 4400 -17540.227 0.018 0.011
Chain 1: 4500 -17442.794 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001244 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.44 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48684.339 1.000 1.000
Chain 1: 200 -22173.441 1.098 1.196
Chain 1: 300 -24549.293 0.764 1.000
Chain 1: 400 -12683.231 0.807 1.000
Chain 1: 500 -12485.653 0.649 0.936
Chain 1: 600 -16172.524 0.579 0.936
Chain 1: 700 -13136.920 0.529 0.231
Chain 1: 800 -10979.961 0.487 0.231
Chain 1: 900 -13684.067 0.455 0.228
Chain 1: 1000 -12254.408 0.421 0.228
Chain 1: 1100 -10485.101 0.338 0.198
Chain 1: 1200 -12959.007 0.238 0.196
Chain 1: 1300 -10671.323 0.250 0.198
Chain 1: 1400 -12022.507 0.167 0.196
Chain 1: 1500 -11214.729 0.173 0.196
Chain 1: 1600 -18556.374 0.190 0.196
Chain 1: 1700 -11396.343 0.229 0.196
Chain 1: 1800 -9700.948 0.227 0.191
Chain 1: 1900 -10376.773 0.214 0.175
Chain 1: 2000 -14728.554 0.232 0.191
Chain 1: 2100 -9353.640 0.272 0.214
Chain 1: 2200 -12586.630 0.279 0.257
Chain 1: 2300 -9251.702 0.294 0.295
Chain 1: 2400 -10012.946 0.290 0.295
Chain 1: 2500 -9167.959 0.292 0.295
Chain 1: 2600 -10196.684 0.262 0.257
Chain 1: 2700 -8790.881 0.216 0.175
Chain 1: 2800 -10221.102 0.212 0.160
Chain 1: 2900 -9803.685 0.210 0.160
Chain 1: 3000 -8989.856 0.189 0.140
Chain 1: 3100 -9318.316 0.135 0.101
Chain 1: 3200 -9414.342 0.111 0.092
Chain 1: 3300 -10093.184 0.081 0.091
Chain 1: 3400 -11639.567 0.087 0.092
Chain 1: 3500 -9142.320 0.105 0.101
Chain 1: 3600 -10637.567 0.109 0.133
Chain 1: 3700 -8702.365 0.115 0.133
Chain 1: 3800 -10206.242 0.116 0.133
Chain 1: 3900 -9000.106 0.125 0.134
Chain 1: 4000 -8974.046 0.117 0.134
Chain 1: 4100 -12885.069 0.143 0.141
Chain 1: 4200 -11885.133 0.151 0.141
Chain 1: 4300 -9247.923 0.173 0.147
Chain 1: 4400 -12178.949 0.183 0.222
Chain 1: 4500 -9115.203 0.190 0.222
Chain 1: 4600 -10582.964 0.189 0.222
Chain 1: 4700 -10426.760 0.169 0.147
Chain 1: 4800 -8581.275 0.176 0.215
Chain 1: 4900 -12850.181 0.195 0.241
Chain 1: 5000 -11495.589 0.207 0.241
Chain 1: 5100 -8287.223 0.215 0.241
Chain 1: 5200 -8502.563 0.209 0.241
Chain 1: 5300 -12845.441 0.215 0.241
Chain 1: 5400 -8382.658 0.244 0.332
Chain 1: 5500 -8286.086 0.211 0.215
Chain 1: 5600 -13196.769 0.235 0.332
Chain 1: 5700 -12305.002 0.240 0.332
Chain 1: 5800 -8484.140 0.264 0.338
Chain 1: 5900 -12502.195 0.263 0.338
Chain 1: 6000 -11655.632 0.258 0.338
Chain 1: 6100 -8567.572 0.256 0.338
Chain 1: 6200 -8684.289 0.254 0.338
Chain 1: 6300 -8350.608 0.225 0.321
Chain 1: 6400 -10863.580 0.195 0.231
Chain 1: 6500 -12868.743 0.209 0.231
Chain 1: 6600 -8407.227 0.225 0.231
Chain 1: 6700 -10125.516 0.235 0.231
Chain 1: 6800 -9604.975 0.195 0.170
Chain 1: 6900 -8434.551 0.177 0.156
Chain 1: 7000 -8345.663 0.170 0.156
Chain 1: 7100 -8679.388 0.138 0.139
Chain 1: 7200 -9033.523 0.141 0.139
Chain 1: 7300 -8798.384 0.140 0.139
Chain 1: 7400 -8183.821 0.124 0.075
Chain 1: 7500 -10860.108 0.133 0.075
Chain 1: 7600 -8317.052 0.110 0.075
Chain 1: 7700 -10546.462 0.115 0.075
Chain 1: 7800 -8739.819 0.130 0.139
Chain 1: 7900 -8037.774 0.125 0.087
Chain 1: 8000 -8747.519 0.132 0.087
Chain 1: 8100 -10429.105 0.144 0.161
Chain 1: 8200 -8877.232 0.158 0.175
Chain 1: 8300 -11658.811 0.179 0.207
Chain 1: 8400 -9319.119 0.196 0.211
Chain 1: 8500 -8227.023 0.185 0.207
Chain 1: 8600 -8127.490 0.156 0.175
Chain 1: 8700 -8221.243 0.136 0.161
Chain 1: 8800 -8384.241 0.117 0.133
Chain 1: 8900 -8680.312 0.112 0.133
Chain 1: 9000 -10019.316 0.117 0.134
Chain 1: 9100 -8211.622 0.123 0.134
Chain 1: 9200 -8338.076 0.107 0.133
Chain 1: 9300 -9150.471 0.092 0.089
Chain 1: 9400 -10842.970 0.082 0.089
Chain 1: 9500 -8813.161 0.092 0.089
Chain 1: 9600 -8111.446 0.100 0.089
Chain 1: 9700 -8430.287 0.102 0.089
Chain 1: 9800 -12629.852 0.134 0.134
Chain 1: 9900 -8495.688 0.179 0.156
Chain 1: 10000 -8464.776 0.166 0.156
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001396 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.96 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56886.240 1.000 1.000
Chain 1: 200 -17179.018 1.656 2.311
Chain 1: 300 -8605.958 1.436 1.000
Chain 1: 400 -7963.633 1.097 1.000
Chain 1: 500 -8599.968 0.892 0.996
Chain 1: 600 -7908.239 0.758 0.996
Chain 1: 700 -8276.031 0.656 0.087
Chain 1: 800 -7971.366 0.579 0.087
Chain 1: 900 -7860.974 0.516 0.081
Chain 1: 1000 -7696.710 0.467 0.081
Chain 1: 1100 -7626.973 0.368 0.074
Chain 1: 1200 -7642.257 0.137 0.044
Chain 1: 1300 -7649.239 0.037 0.038
Chain 1: 1400 -7842.733 0.032 0.025
Chain 1: 1500 -7600.511 0.027 0.025
Chain 1: 1600 -7569.776 0.019 0.021
Chain 1: 1700 -7503.443 0.016 0.014
Chain 1: 1800 -7545.971 0.012 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003608 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 36.08 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85812.484 1.000 1.000
Chain 1: 200 -13232.533 3.242 5.485
Chain 1: 300 -9680.527 2.284 1.000
Chain 1: 400 -10539.413 1.733 1.000
Chain 1: 500 -8615.652 1.431 0.367
Chain 1: 600 -8491.660 1.195 0.367
Chain 1: 700 -8553.276 1.025 0.223
Chain 1: 800 -8980.950 0.903 0.223
Chain 1: 900 -8556.282 0.808 0.081
Chain 1: 1000 -8260.121 0.731 0.081
Chain 1: 1100 -8574.843 0.635 0.050 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8109.058 0.092 0.050
Chain 1: 1300 -8423.506 0.059 0.048
Chain 1: 1400 -8408.313 0.051 0.037
Chain 1: 1500 -8288.924 0.030 0.037
Chain 1: 1600 -8392.069 0.030 0.037
Chain 1: 1700 -8479.206 0.030 0.037
Chain 1: 1800 -8087.698 0.030 0.037
Chain 1: 1900 -8189.917 0.027 0.036
Chain 1: 2000 -8160.092 0.023 0.014
Chain 1: 2100 -8287.735 0.021 0.014
Chain 1: 2200 -8074.067 0.018 0.014
Chain 1: 2300 -8218.753 0.016 0.014
Chain 1: 2400 -8233.730 0.016 0.014
Chain 1: 2500 -8200.336 0.015 0.012
Chain 1: 2600 -8201.912 0.014 0.012
Chain 1: 2700 -8109.050 0.014 0.012
Chain 1: 2800 -8082.801 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002915 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.15 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8384806.404 1.000 1.000
Chain 1: 200 -1582033.615 2.650 4.300
Chain 1: 300 -890203.341 2.026 1.000
Chain 1: 400 -457557.993 1.756 1.000
Chain 1: 500 -358214.290 1.460 0.946
Chain 1: 600 -233128.865 1.306 0.946
Chain 1: 700 -119122.037 1.256 0.946
Chain 1: 800 -86316.922 1.147 0.946
Chain 1: 900 -66609.779 1.052 0.777
Chain 1: 1000 -51369.426 0.977 0.777
Chain 1: 1100 -38821.508 0.909 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37988.450 0.481 0.380
Chain 1: 1300 -25921.745 0.450 0.380
Chain 1: 1400 -25636.414 0.357 0.323
Chain 1: 1500 -22218.872 0.344 0.323
Chain 1: 1600 -21433.519 0.294 0.297
Chain 1: 1700 -20304.613 0.204 0.296
Chain 1: 1800 -20247.905 0.166 0.154
Chain 1: 1900 -20573.638 0.138 0.056
Chain 1: 2000 -19084.423 0.116 0.056
Chain 1: 2100 -19322.646 0.085 0.037
Chain 1: 2200 -19549.199 0.084 0.037
Chain 1: 2300 -19166.406 0.040 0.020
Chain 1: 2400 -18938.615 0.040 0.020
Chain 1: 2500 -18740.935 0.026 0.016
Chain 1: 2600 -18371.358 0.024 0.016
Chain 1: 2700 -18328.332 0.019 0.012
Chain 1: 2800 -18045.580 0.020 0.016
Chain 1: 2900 -18326.599 0.020 0.015
Chain 1: 3000 -18312.746 0.012 0.012
Chain 1: 3100 -18397.722 0.011 0.012
Chain 1: 3200 -18088.649 0.012 0.015
Chain 1: 3300 -18293.157 0.011 0.012
Chain 1: 3400 -17768.676 0.013 0.015
Chain 1: 3500 -18379.790 0.015 0.016
Chain 1: 3600 -17687.400 0.017 0.016
Chain 1: 3700 -18073.564 0.019 0.017
Chain 1: 3800 -17034.865 0.023 0.021
Chain 1: 3900 -17031.080 0.022 0.021
Chain 1: 4000 -17148.332 0.022 0.021
Chain 1: 4100 -17062.246 0.022 0.021
Chain 1: 4200 -16878.789 0.022 0.021
Chain 1: 4300 -17016.936 0.022 0.021
Chain 1: 4400 -16974.027 0.019 0.011
Chain 1: 4500 -16876.640 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00134 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.4 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12486.686 1.000 1.000
Chain 1: 200 -9077.238 0.688 1.000
Chain 1: 300 -7968.520 0.505 0.376
Chain 1: 400 -8089.227 0.382 0.376
Chain 1: 500 -8046.938 0.307 0.139
Chain 1: 600 -7886.360 0.259 0.139
Chain 1: 700 -7827.222 0.223 0.020
Chain 1: 800 -7845.322 0.196 0.020
Chain 1: 900 -7780.133 0.175 0.015
Chain 1: 1000 -7895.746 0.159 0.015
Chain 1: 1100 -7975.229 0.060 0.015
Chain 1: 1200 -7845.655 0.024 0.015
Chain 1: 1300 -7870.018 0.010 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001387 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.87 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -55435.307 1.000 1.000
Chain 1: 200 -17065.136 1.624 2.248
Chain 1: 300 -8614.142 1.410 1.000
Chain 1: 400 -8175.385 1.071 1.000
Chain 1: 500 -7867.433 0.864 0.981
Chain 1: 600 -8888.913 0.740 0.981
Chain 1: 700 -8086.341 0.648 0.115
Chain 1: 800 -7786.942 0.572 0.115
Chain 1: 900 -7879.785 0.510 0.099
Chain 1: 1000 -7674.468 0.461 0.099
Chain 1: 1100 -7619.164 0.362 0.054
Chain 1: 1200 -7500.993 0.139 0.039
Chain 1: 1300 -7571.641 0.042 0.038
Chain 1: 1400 -7886.452 0.040 0.038
Chain 1: 1500 -7549.039 0.041 0.038
Chain 1: 1600 -7707.031 0.031 0.027
Chain 1: 1700 -7440.642 0.025 0.027
Chain 1: 1800 -7518.053 0.022 0.020
Chain 1: 1900 -7520.511 0.021 0.020
Chain 1: 2000 -7547.716 0.019 0.016
Chain 1: 2100 -7537.452 0.018 0.016
Chain 1: 2200 -7629.924 0.018 0.012
Chain 1: 2300 -7512.384 0.018 0.016
Chain 1: 2400 -7578.352 0.015 0.012
Chain 1: 2500 -7514.276 0.012 0.010
Chain 1: 2600 -7479.079 0.010 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002903 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.03 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86304.057 1.000 1.000
Chain 1: 200 -13427.984 3.214 5.427
Chain 1: 300 -9803.965 2.266 1.000
Chain 1: 400 -10549.417 1.717 1.000
Chain 1: 500 -8780.411 1.414 0.370
Chain 1: 600 -8261.214 1.189 0.370
Chain 1: 700 -8661.915 1.025 0.201
Chain 1: 800 -9323.948 0.906 0.201
Chain 1: 900 -8633.700 0.814 0.080
Chain 1: 1000 -8372.068 0.736 0.080
Chain 1: 1100 -8640.064 0.639 0.071 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8262.436 0.101 0.071
Chain 1: 1300 -8438.410 0.066 0.063
Chain 1: 1400 -8489.864 0.060 0.046
Chain 1: 1500 -8367.411 0.041 0.046
Chain 1: 1600 -8475.431 0.036 0.031
Chain 1: 1700 -8568.422 0.032 0.031
Chain 1: 1800 -8154.798 0.030 0.031
Chain 1: 1900 -8250.869 0.024 0.021
Chain 1: 2000 -8224.143 0.021 0.015
Chain 1: 2100 -8346.489 0.019 0.015
Chain 1: 2200 -8166.282 0.017 0.015
Chain 1: 2300 -8246.046 0.016 0.013
Chain 1: 2400 -8315.652 0.016 0.013
Chain 1: 2500 -8260.958 0.015 0.012
Chain 1: 2600 -8260.254 0.014 0.011
Chain 1: 2700 -8177.572 0.014 0.010
Chain 1: 2800 -8141.477 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002571 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.71 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8418111.810 1.000 1.000
Chain 1: 200 -1588390.192 2.650 4.300
Chain 1: 300 -890727.042 2.028 1.000
Chain 1: 400 -456968.850 1.758 1.000
Chain 1: 500 -356637.230 1.463 0.949
Chain 1: 600 -231743.811 1.309 0.949
Chain 1: 700 -118562.514 1.258 0.949
Chain 1: 800 -85897.131 1.148 0.949
Chain 1: 900 -66368.680 1.054 0.783
Chain 1: 1000 -51265.435 0.978 0.783
Chain 1: 1100 -38831.163 0.910 0.539 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38022.196 0.482 0.380
Chain 1: 1300 -26068.790 0.449 0.380
Chain 1: 1400 -25797.771 0.355 0.320
Chain 1: 1500 -22407.199 0.342 0.320
Chain 1: 1600 -21630.524 0.292 0.295
Chain 1: 1700 -20515.064 0.202 0.294
Chain 1: 1800 -20461.941 0.164 0.151
Chain 1: 1900 -20788.025 0.137 0.054
Chain 1: 2000 -19304.777 0.115 0.054
Chain 1: 2100 -19543.038 0.084 0.036
Chain 1: 2200 -19768.377 0.083 0.036
Chain 1: 2300 -19386.583 0.039 0.020
Chain 1: 2400 -19158.773 0.039 0.020
Chain 1: 2500 -18960.325 0.025 0.016
Chain 1: 2600 -18591.038 0.023 0.016
Chain 1: 2700 -18548.284 0.018 0.012
Chain 1: 2800 -18264.855 0.020 0.016
Chain 1: 2900 -18546.019 0.020 0.015
Chain 1: 3000 -18532.380 0.012 0.012
Chain 1: 3100 -18617.253 0.011 0.012
Chain 1: 3200 -18308.143 0.012 0.015
Chain 1: 3300 -18512.764 0.011 0.012
Chain 1: 3400 -17987.747 0.013 0.015
Chain 1: 3500 -18599.310 0.015 0.016
Chain 1: 3600 -17906.460 0.017 0.016
Chain 1: 3700 -18292.760 0.019 0.017
Chain 1: 3800 -17253.029 0.023 0.021
Chain 1: 3900 -17249.125 0.022 0.021
Chain 1: 4000 -17366.518 0.022 0.021
Chain 1: 4100 -17280.177 0.022 0.021
Chain 1: 4200 -17096.636 0.022 0.021
Chain 1: 4300 -17234.952 0.021 0.021
Chain 1: 4400 -17191.882 0.019 0.011
Chain 1: 4500 -17094.370 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001332 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.32 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48684.148 1.000 1.000
Chain 1: 200 -20411.180 1.193 1.385
Chain 1: 300 -13521.784 0.965 1.000
Chain 1: 400 -13442.167 0.725 1.000
Chain 1: 500 -14669.392 0.597 0.510
Chain 1: 600 -13566.836 0.511 0.510
Chain 1: 700 -14694.591 0.449 0.084
Chain 1: 800 -14034.120 0.399 0.084
Chain 1: 900 -11215.006 0.382 0.084
Chain 1: 1000 -11297.898 0.345 0.084
Chain 1: 1100 -16441.278 0.276 0.084
Chain 1: 1200 -12584.626 0.168 0.084
Chain 1: 1300 -11349.367 0.128 0.084
Chain 1: 1400 -18080.982 0.165 0.109
Chain 1: 1500 -10534.016 0.228 0.251
Chain 1: 1600 -10224.758 0.223 0.251
Chain 1: 1700 -19231.532 0.262 0.306
Chain 1: 1800 -18283.999 0.263 0.306
Chain 1: 1900 -9181.224 0.337 0.313
Chain 1: 2000 -16314.090 0.380 0.372
Chain 1: 2100 -9282.582 0.424 0.437
Chain 1: 2200 -10018.858 0.401 0.437
Chain 1: 2300 -9144.862 0.399 0.437
Chain 1: 2400 -8946.214 0.364 0.437
Chain 1: 2500 -14232.769 0.330 0.371
Chain 1: 2600 -9921.548 0.370 0.435
Chain 1: 2700 -8851.964 0.336 0.371
Chain 1: 2800 -9610.691 0.338 0.371
Chain 1: 2900 -9076.026 0.245 0.121
Chain 1: 3000 -8698.275 0.206 0.096
Chain 1: 3100 -11865.097 0.157 0.096
Chain 1: 3200 -14180.456 0.166 0.121
Chain 1: 3300 -14300.101 0.157 0.121
Chain 1: 3400 -14094.252 0.156 0.121
Chain 1: 3500 -9355.370 0.170 0.121
Chain 1: 3600 -9409.509 0.127 0.079
Chain 1: 3700 -11469.825 0.133 0.079
Chain 1: 3800 -13720.811 0.141 0.163
Chain 1: 3900 -9027.145 0.187 0.164
Chain 1: 4000 -8402.341 0.190 0.164
Chain 1: 4100 -9084.941 0.171 0.163
Chain 1: 4200 -10397.312 0.167 0.126
Chain 1: 4300 -9619.890 0.175 0.126
Chain 1: 4400 -8867.265 0.182 0.126
Chain 1: 4500 -8713.776 0.133 0.085
Chain 1: 4600 -10873.563 0.152 0.126
Chain 1: 4700 -13308.698 0.152 0.126
Chain 1: 4800 -8700.445 0.189 0.126
Chain 1: 4900 -14565.466 0.177 0.126
Chain 1: 5000 -14890.854 0.172 0.126
Chain 1: 5100 -8311.999 0.244 0.183
Chain 1: 5200 -9106.287 0.240 0.183
Chain 1: 5300 -8399.753 0.240 0.183
Chain 1: 5400 -13235.360 0.268 0.199
Chain 1: 5500 -8597.732 0.320 0.365
Chain 1: 5600 -8969.595 0.305 0.365
Chain 1: 5700 -12726.119 0.316 0.365
Chain 1: 5800 -9034.943 0.304 0.365
Chain 1: 5900 -11304.991 0.284 0.295
Chain 1: 6000 -9806.532 0.297 0.295
Chain 1: 6100 -8347.685 0.235 0.201
Chain 1: 6200 -12680.473 0.260 0.295
Chain 1: 6300 -9557.597 0.285 0.327
Chain 1: 6400 -13599.897 0.278 0.297
Chain 1: 6500 -8739.501 0.280 0.297
Chain 1: 6600 -8319.645 0.280 0.297
Chain 1: 6700 -8308.269 0.251 0.297
Chain 1: 6800 -8143.616 0.212 0.201
Chain 1: 6900 -8153.815 0.192 0.175
Chain 1: 7000 -8486.895 0.181 0.175
Chain 1: 7100 -9752.786 0.176 0.130
Chain 1: 7200 -10436.944 0.149 0.066
Chain 1: 7300 -8804.240 0.135 0.066
Chain 1: 7400 -13235.033 0.138 0.066
Chain 1: 7500 -8590.416 0.137 0.066
Chain 1: 7600 -8183.240 0.137 0.066
Chain 1: 7700 -8188.657 0.137 0.066
Chain 1: 7800 -8932.632 0.143 0.083
Chain 1: 7900 -8085.107 0.153 0.105
Chain 1: 8000 -9244.929 0.162 0.125
Chain 1: 8100 -7979.398 0.165 0.125
Chain 1: 8200 -10701.872 0.184 0.159
Chain 1: 8300 -8791.572 0.187 0.159
Chain 1: 8400 -8723.605 0.154 0.125
Chain 1: 8500 -8037.549 0.109 0.105
Chain 1: 8600 -8364.152 0.108 0.105
Chain 1: 8700 -8347.659 0.108 0.105
Chain 1: 8800 -10377.195 0.119 0.125
Chain 1: 8900 -11630.743 0.119 0.125
Chain 1: 9000 -12074.487 0.110 0.108
Chain 1: 9100 -12175.323 0.095 0.085
Chain 1: 9200 -8062.754 0.121 0.085
Chain 1: 9300 -8227.760 0.101 0.039
Chain 1: 9400 -8443.853 0.103 0.039
Chain 1: 9500 -10358.517 0.113 0.039
Chain 1: 9600 -9356.914 0.120 0.107
Chain 1: 9700 -8274.408 0.133 0.108
Chain 1: 9800 -8462.015 0.115 0.107
Chain 1: 9900 -9007.863 0.111 0.061
Chain 1: 10000 -9732.356 0.114 0.074
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001435 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.35 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61490.162 1.000 1.000
Chain 1: 200 -17536.764 1.753 2.506
Chain 1: 300 -8686.536 1.508 1.019
Chain 1: 400 -8952.428 1.139 1.019
Chain 1: 500 -7753.519 0.942 1.000
Chain 1: 600 -8386.067 0.797 1.000
Chain 1: 700 -8241.638 0.686 0.155
Chain 1: 800 -8081.641 0.603 0.155
Chain 1: 900 -7790.803 0.540 0.075
Chain 1: 1000 -7769.135 0.486 0.075
Chain 1: 1100 -7818.804 0.387 0.037
Chain 1: 1200 -7528.978 0.140 0.037
Chain 1: 1300 -7716.329 0.041 0.030
Chain 1: 1400 -7860.320 0.039 0.024
Chain 1: 1500 -7607.249 0.027 0.024
Chain 1: 1600 -7503.371 0.021 0.020
Chain 1: 1700 -7488.570 0.020 0.020
Chain 1: 1800 -7530.539 0.018 0.018
Chain 1: 1900 -7575.411 0.015 0.014
Chain 1: 2000 -7567.533 0.015 0.014
Chain 1: 2100 -7581.782 0.014 0.014
Chain 1: 2200 -7656.550 0.012 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002555 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.55 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85181.582 1.000 1.000
Chain 1: 200 -13128.831 3.244 5.488
Chain 1: 300 -9590.950 2.286 1.000
Chain 1: 400 -10465.806 1.735 1.000
Chain 1: 500 -8515.145 1.434 0.369
Chain 1: 600 -8152.149 1.202 0.369
Chain 1: 700 -8395.137 1.035 0.229
Chain 1: 800 -8794.598 0.911 0.229
Chain 1: 900 -8431.523 0.815 0.084
Chain 1: 1000 -8212.368 0.736 0.084
Chain 1: 1100 -8459.628 0.639 0.045 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8125.893 0.094 0.045
Chain 1: 1300 -8342.621 0.060 0.043
Chain 1: 1400 -8330.475 0.052 0.041
Chain 1: 1500 -8228.096 0.030 0.029
Chain 1: 1600 -8322.291 0.027 0.029
Chain 1: 1700 -8416.944 0.025 0.027
Chain 1: 1800 -8030.506 0.025 0.027
Chain 1: 1900 -8132.900 0.022 0.026
Chain 1: 2000 -8102.686 0.020 0.013
Chain 1: 2100 -8238.164 0.018 0.013
Chain 1: 2200 -8021.897 0.017 0.013
Chain 1: 2300 -8163.211 0.016 0.013
Chain 1: 2400 -8173.601 0.016 0.013
Chain 1: 2500 -8141.981 0.015 0.013
Chain 1: 2600 -8139.714 0.014 0.013
Chain 1: 2700 -8049.267 0.014 0.013
Chain 1: 2800 -8027.811 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003449 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.49 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8402745.249 1.000 1.000
Chain 1: 200 -1584653.257 2.651 4.303
Chain 1: 300 -890796.376 2.027 1.000
Chain 1: 400 -457307.647 1.757 1.000
Chain 1: 500 -357522.332 1.462 0.948
Chain 1: 600 -232610.839 1.308 0.948
Chain 1: 700 -118835.764 1.258 0.948
Chain 1: 800 -86038.994 1.148 0.948
Chain 1: 900 -66384.951 1.053 0.779
Chain 1: 1000 -51177.360 0.978 0.779
Chain 1: 1100 -38653.045 0.910 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37824.367 0.482 0.381
Chain 1: 1300 -25789.002 0.451 0.381
Chain 1: 1400 -25506.340 0.357 0.324
Chain 1: 1500 -22096.169 0.345 0.324
Chain 1: 1600 -21312.423 0.295 0.297
Chain 1: 1700 -20187.948 0.204 0.296
Chain 1: 1800 -20132.190 0.167 0.154
Chain 1: 1900 -20457.716 0.139 0.056
Chain 1: 2000 -18970.876 0.117 0.056
Chain 1: 2100 -19209.042 0.086 0.037
Chain 1: 2200 -19434.976 0.085 0.037
Chain 1: 2300 -19052.844 0.040 0.020
Chain 1: 2400 -18825.186 0.040 0.020
Chain 1: 2500 -18627.168 0.026 0.016
Chain 1: 2600 -18257.969 0.024 0.016
Chain 1: 2700 -18215.152 0.019 0.012
Chain 1: 2800 -17932.238 0.020 0.016
Chain 1: 2900 -18213.213 0.020 0.015
Chain 1: 3000 -18199.503 0.012 0.012
Chain 1: 3100 -18284.361 0.011 0.012
Chain 1: 3200 -17975.439 0.012 0.015
Chain 1: 3300 -18179.859 0.011 0.012
Chain 1: 3400 -17655.461 0.013 0.015
Chain 1: 3500 -18266.259 0.015 0.016
Chain 1: 3600 -17574.412 0.017 0.016
Chain 1: 3700 -17960.091 0.019 0.017
Chain 1: 3800 -16922.016 0.023 0.021
Chain 1: 3900 -16918.232 0.022 0.021
Chain 1: 4000 -17035.538 0.023 0.021
Chain 1: 4100 -16949.372 0.023 0.021
Chain 1: 4200 -16766.152 0.022 0.021
Chain 1: 4300 -16904.176 0.022 0.021
Chain 1: 4400 -16861.389 0.019 0.011
Chain 1: 4500 -16764.027 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003446 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.46 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49363.255 1.000 1.000
Chain 1: 200 -16500.164 1.496 1.992
Chain 1: 300 -21316.688 1.073 1.000
Chain 1: 400 -29100.025 0.871 1.000
Chain 1: 500 -15283.485 0.878 0.904
Chain 1: 600 -15924.908 0.738 0.904
Chain 1: 700 -13350.631 0.660 0.267
Chain 1: 800 -15312.649 0.594 0.267
Chain 1: 900 -11040.510 0.571 0.267
Chain 1: 1000 -12940.639 0.528 0.267
Chain 1: 1100 -12088.933 0.435 0.226
Chain 1: 1200 -12251.162 0.238 0.193
Chain 1: 1300 -14817.827 0.232 0.173
Chain 1: 1400 -11024.605 0.240 0.173
Chain 1: 1500 -10401.370 0.156 0.147
Chain 1: 1600 -12463.644 0.168 0.165
Chain 1: 1700 -10358.346 0.169 0.165
Chain 1: 1800 -13021.455 0.177 0.173
Chain 1: 1900 -9864.957 0.170 0.173
Chain 1: 2000 -10181.602 0.159 0.173
Chain 1: 2100 -9555.925 0.158 0.173
Chain 1: 2200 -11353.877 0.173 0.173
Chain 1: 2300 -9507.413 0.175 0.194
Chain 1: 2400 -10579.781 0.150 0.165
Chain 1: 2500 -15282.426 0.175 0.194
Chain 1: 2600 -10394.880 0.206 0.203
Chain 1: 2700 -9939.616 0.190 0.194
Chain 1: 2800 -10837.574 0.178 0.158
Chain 1: 2900 -9691.799 0.158 0.118
Chain 1: 3000 -11797.372 0.172 0.158
Chain 1: 3100 -10135.094 0.182 0.164
Chain 1: 3200 -11216.835 0.176 0.164
Chain 1: 3300 -10643.862 0.162 0.118
Chain 1: 3400 -10401.833 0.154 0.118
Chain 1: 3500 -10607.871 0.125 0.096
Chain 1: 3600 -17667.987 0.118 0.096
Chain 1: 3700 -9836.241 0.193 0.118
Chain 1: 3800 -9228.366 0.192 0.118
Chain 1: 3900 -12403.980 0.205 0.164
Chain 1: 4000 -17053.725 0.215 0.164
Chain 1: 4100 -10374.428 0.263 0.256
Chain 1: 4200 -13458.209 0.276 0.256
Chain 1: 4300 -9904.930 0.306 0.273
Chain 1: 4400 -9251.915 0.311 0.273
Chain 1: 4500 -12480.861 0.335 0.273
Chain 1: 4600 -8953.397 0.335 0.273
Chain 1: 4700 -11792.995 0.279 0.259
Chain 1: 4800 -9279.685 0.300 0.271
Chain 1: 4900 -12038.179 0.297 0.271
Chain 1: 5000 -14870.066 0.289 0.259
Chain 1: 5100 -9308.403 0.284 0.259
Chain 1: 5200 -8942.618 0.265 0.259
Chain 1: 5300 -10587.995 0.245 0.241
Chain 1: 5400 -8732.847 0.259 0.241
Chain 1: 5500 -8713.097 0.233 0.229
Chain 1: 5600 -8669.364 0.194 0.212
Chain 1: 5700 -8975.746 0.174 0.190
Chain 1: 5800 -8920.643 0.147 0.155
Chain 1: 5900 -10046.775 0.136 0.112
Chain 1: 6000 -9278.091 0.125 0.083
Chain 1: 6100 -8684.743 0.072 0.068
Chain 1: 6200 -8803.501 0.069 0.068
Chain 1: 6300 -12181.931 0.081 0.068
Chain 1: 6400 -13386.881 0.069 0.068
Chain 1: 6500 -10107.893 0.101 0.083
Chain 1: 6600 -14498.569 0.131 0.090
Chain 1: 6700 -8863.891 0.191 0.112
Chain 1: 6800 -11581.326 0.214 0.235
Chain 1: 6900 -8609.814 0.237 0.277
Chain 1: 7000 -10730.147 0.249 0.277
Chain 1: 7100 -8442.722 0.269 0.277
Chain 1: 7200 -8529.405 0.269 0.277
Chain 1: 7300 -14248.533 0.281 0.303
Chain 1: 7400 -10570.459 0.307 0.324
Chain 1: 7500 -10248.986 0.278 0.303
Chain 1: 7600 -9055.272 0.261 0.271
Chain 1: 7700 -10871.688 0.214 0.235
Chain 1: 7800 -11180.034 0.193 0.198
Chain 1: 7900 -8475.380 0.191 0.198
Chain 1: 8000 -9617.513 0.183 0.167
Chain 1: 8100 -8459.365 0.169 0.137
Chain 1: 8200 -9350.897 0.178 0.137
Chain 1: 8300 -8486.507 0.148 0.132
Chain 1: 8400 -9125.862 0.120 0.119
Chain 1: 8500 -9901.815 0.125 0.119
Chain 1: 8600 -13442.457 0.138 0.119
Chain 1: 8700 -8648.283 0.177 0.119
Chain 1: 8800 -8625.281 0.174 0.119
Chain 1: 8900 -11597.627 0.168 0.119
Chain 1: 9000 -9314.073 0.180 0.137
Chain 1: 9100 -9939.825 0.173 0.102
Chain 1: 9200 -10218.002 0.166 0.102
Chain 1: 9300 -8601.470 0.175 0.188
Chain 1: 9400 -12605.583 0.200 0.245
Chain 1: 9500 -8440.953 0.241 0.256
Chain 1: 9600 -11020.456 0.238 0.245
Chain 1: 9700 -10957.523 0.183 0.234
Chain 1: 9800 -11236.520 0.186 0.234
Chain 1: 9900 -8483.212 0.192 0.234
Chain 1: 10000 -8808.237 0.172 0.188
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001428 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.28 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -63544.957 1.000 1.000
Chain 1: 200 -18484.735 1.719 2.438
Chain 1: 300 -8930.797 1.502 1.070
Chain 1: 400 -9221.750 1.135 1.070
Chain 1: 500 -8190.050 0.933 1.000
Chain 1: 600 -8530.952 0.784 1.000
Chain 1: 700 -8172.546 0.678 0.126
Chain 1: 800 -8333.433 0.596 0.126
Chain 1: 900 -7776.382 0.538 0.072
Chain 1: 1000 -8131.792 0.488 0.072
Chain 1: 1100 -7729.481 0.394 0.052
Chain 1: 1200 -7618.139 0.151 0.044
Chain 1: 1300 -7822.322 0.047 0.044
Chain 1: 1400 -7957.126 0.045 0.044
Chain 1: 1500 -7642.329 0.037 0.041
Chain 1: 1600 -7717.423 0.034 0.041
Chain 1: 1700 -7570.704 0.031 0.026
Chain 1: 1800 -7709.595 0.031 0.026
Chain 1: 1900 -7595.053 0.026 0.019
Chain 1: 2000 -7671.106 0.022 0.018
Chain 1: 2100 -7619.811 0.018 0.017
Chain 1: 2200 -7738.983 0.018 0.017
Chain 1: 2300 -7603.227 0.017 0.017
Chain 1: 2400 -7665.781 0.016 0.015
Chain 1: 2500 -7662.326 0.012 0.015
Chain 1: 2600 -7562.350 0.012 0.015
Chain 1: 2700 -7567.531 0.011 0.013
Chain 1: 2800 -7595.459 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00262 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 26.2 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86465.242 1.000 1.000
Chain 1: 200 -13776.194 3.138 5.276
Chain 1: 300 -10081.489 2.214 1.000
Chain 1: 400 -11149.370 1.685 1.000
Chain 1: 500 -9064.764 1.394 0.366
Chain 1: 600 -8572.557 1.171 0.366
Chain 1: 700 -8717.243 1.006 0.230
Chain 1: 800 -9115.448 0.886 0.230
Chain 1: 900 -8953.815 0.789 0.096
Chain 1: 1000 -8998.695 0.711 0.096
Chain 1: 1100 -8717.097 0.614 0.057 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8472.584 0.089 0.044
Chain 1: 1300 -8754.296 0.056 0.032
Chain 1: 1400 -8715.867 0.047 0.032
Chain 1: 1500 -8608.280 0.025 0.029
Chain 1: 1600 -8714.905 0.021 0.018
Chain 1: 1700 -8787.835 0.020 0.018
Chain 1: 1800 -8355.335 0.021 0.018
Chain 1: 1900 -8459.535 0.020 0.012
Chain 1: 2000 -8434.919 0.020 0.012
Chain 1: 2100 -8413.828 0.017 0.012
Chain 1: 2200 -8378.177 0.014 0.012
Chain 1: 2300 -8506.407 0.013 0.012
Chain 1: 2400 -8362.083 0.014 0.012
Chain 1: 2500 -8430.597 0.013 0.012
Chain 1: 2600 -8350.176 0.013 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002517 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.17 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8416667.328 1.000 1.000
Chain 1: 200 -1582445.503 2.659 4.319
Chain 1: 300 -889742.444 2.032 1.000
Chain 1: 400 -457578.155 1.760 1.000
Chain 1: 500 -357954.690 1.464 0.944
Chain 1: 600 -233093.747 1.309 0.944
Chain 1: 700 -119432.244 1.258 0.944
Chain 1: 800 -86691.680 1.148 0.944
Chain 1: 900 -67045.770 1.053 0.779
Chain 1: 1000 -51858.309 0.977 0.779
Chain 1: 1100 -39349.448 0.909 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38528.278 0.479 0.378
Chain 1: 1300 -26488.071 0.447 0.378
Chain 1: 1400 -26209.677 0.353 0.318
Chain 1: 1500 -22798.023 0.340 0.318
Chain 1: 1600 -22015.831 0.290 0.293
Chain 1: 1700 -20889.167 0.201 0.293
Chain 1: 1800 -20833.623 0.163 0.150
Chain 1: 1900 -21160.095 0.135 0.054
Chain 1: 2000 -19670.715 0.114 0.054
Chain 1: 2100 -19908.980 0.083 0.036
Chain 1: 2200 -20135.822 0.082 0.036
Chain 1: 2300 -19752.656 0.039 0.019
Chain 1: 2400 -19524.603 0.039 0.019
Chain 1: 2500 -19326.745 0.025 0.015
Chain 1: 2600 -18956.477 0.023 0.015
Chain 1: 2700 -18913.341 0.018 0.012
Chain 1: 2800 -18630.100 0.019 0.015
Chain 1: 2900 -18911.498 0.019 0.015
Chain 1: 3000 -18897.630 0.012 0.012
Chain 1: 3100 -18982.676 0.011 0.012
Chain 1: 3200 -18673.128 0.012 0.015
Chain 1: 3300 -18878.040 0.011 0.012
Chain 1: 3400 -18352.584 0.012 0.015
Chain 1: 3500 -18965.052 0.015 0.015
Chain 1: 3600 -18270.942 0.016 0.015
Chain 1: 3700 -18658.328 0.018 0.017
Chain 1: 3800 -17616.855 0.023 0.021
Chain 1: 3900 -17612.981 0.021 0.021
Chain 1: 4000 -17730.277 0.022 0.021
Chain 1: 4100 -17643.986 0.022 0.021
Chain 1: 4200 -17459.968 0.021 0.021
Chain 1: 4300 -17598.541 0.021 0.021
Chain 1: 4400 -17555.138 0.018 0.011
Chain 1: 4500 -17457.641 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00132 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.2 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49464.089 1.000 1.000
Chain 1: 200 -16505.990 1.498 1.997
Chain 1: 300 -25212.273 1.114 1.000
Chain 1: 400 -19062.743 0.916 1.000
Chain 1: 500 -15598.358 0.777 0.345
Chain 1: 600 -13302.465 0.677 0.345
Chain 1: 700 -15165.666 0.597 0.323
Chain 1: 800 -14104.268 0.532 0.323
Chain 1: 900 -12420.419 0.488 0.222
Chain 1: 1000 -10949.604 0.453 0.222
Chain 1: 1100 -22162.264 0.403 0.222
Chain 1: 1200 -11352.827 0.299 0.222
Chain 1: 1300 -13274.681 0.279 0.173
Chain 1: 1400 -16017.552 0.264 0.171
Chain 1: 1500 -11496.843 0.281 0.171
Chain 1: 1600 -13311.830 0.277 0.145
Chain 1: 1700 -10844.399 0.288 0.171
Chain 1: 1800 -12393.178 0.293 0.171
Chain 1: 1900 -10893.798 0.293 0.171
Chain 1: 2000 -19084.004 0.322 0.228
Chain 1: 2100 -12707.336 0.322 0.228
Chain 1: 2200 -10714.188 0.245 0.186
Chain 1: 2300 -10607.906 0.232 0.186
Chain 1: 2400 -11533.082 0.223 0.186
Chain 1: 2500 -10792.862 0.190 0.138
Chain 1: 2600 -10126.882 0.183 0.138
Chain 1: 2700 -9554.148 0.166 0.125
Chain 1: 2800 -10790.382 0.165 0.115
Chain 1: 2900 -11803.389 0.160 0.086
Chain 1: 3000 -15190.300 0.140 0.086
Chain 1: 3100 -10068.935 0.140 0.086
Chain 1: 3200 -14852.410 0.154 0.086
Chain 1: 3300 -10028.866 0.201 0.115
Chain 1: 3400 -13780.235 0.220 0.223
Chain 1: 3500 -9795.899 0.254 0.272
Chain 1: 3600 -9460.216 0.251 0.272
Chain 1: 3700 -10153.933 0.252 0.272
Chain 1: 3800 -10371.479 0.242 0.272
Chain 1: 3900 -12349.721 0.250 0.272
Chain 1: 4000 -13051.846 0.233 0.272
Chain 1: 4100 -9077.520 0.226 0.272
Chain 1: 4200 -9697.869 0.200 0.160
Chain 1: 4300 -16134.920 0.192 0.160
Chain 1: 4400 -9917.384 0.227 0.160
Chain 1: 4500 -12321.270 0.206 0.160
Chain 1: 4600 -9246.895 0.236 0.195
Chain 1: 4700 -13510.791 0.261 0.316
Chain 1: 4800 -9155.714 0.306 0.332
Chain 1: 4900 -9227.303 0.291 0.332
Chain 1: 5000 -15176.524 0.325 0.392
Chain 1: 5100 -8995.582 0.350 0.392
Chain 1: 5200 -9068.483 0.344 0.392
Chain 1: 5300 -9635.030 0.310 0.332
Chain 1: 5400 -8880.651 0.256 0.316
Chain 1: 5500 -9318.011 0.241 0.316
Chain 1: 5600 -9294.034 0.208 0.085
Chain 1: 5700 -9072.981 0.179 0.059
Chain 1: 5800 -9058.339 0.131 0.047
Chain 1: 5900 -9075.191 0.131 0.047
Chain 1: 6000 -11986.863 0.116 0.047
Chain 1: 6100 -9319.299 0.076 0.047
Chain 1: 6200 -16919.580 0.120 0.059
Chain 1: 6300 -10031.166 0.183 0.085
Chain 1: 6400 -10391.506 0.178 0.047
Chain 1: 6500 -15216.197 0.205 0.243
Chain 1: 6600 -9014.078 0.273 0.286
Chain 1: 6700 -8785.973 0.273 0.286
Chain 1: 6800 -9368.210 0.279 0.286
Chain 1: 6900 -11763.945 0.300 0.286
Chain 1: 7000 -10246.752 0.290 0.286
Chain 1: 7100 -8632.267 0.280 0.204
Chain 1: 7200 -8691.371 0.236 0.187
Chain 1: 7300 -11552.207 0.192 0.187
Chain 1: 7400 -12826.060 0.199 0.187
Chain 1: 7500 -9047.608 0.209 0.187
Chain 1: 7600 -9458.296 0.144 0.148
Chain 1: 7700 -9820.908 0.145 0.148
Chain 1: 7800 -11919.526 0.157 0.176
Chain 1: 7900 -9290.947 0.165 0.176
Chain 1: 8000 -9158.835 0.151 0.176
Chain 1: 8100 -8845.825 0.136 0.099
Chain 1: 8200 -11025.322 0.155 0.176
Chain 1: 8300 -8725.733 0.157 0.176
Chain 1: 8400 -14086.426 0.185 0.198
Chain 1: 8500 -9129.411 0.197 0.198
Chain 1: 8600 -12168.525 0.218 0.250
Chain 1: 8700 -8917.771 0.251 0.264
Chain 1: 8800 -8633.104 0.236 0.264
Chain 1: 8900 -9478.462 0.217 0.250
Chain 1: 9000 -9634.790 0.217 0.250
Chain 1: 9100 -9685.284 0.214 0.250
Chain 1: 9200 -9074.964 0.201 0.250
Chain 1: 9300 -12167.696 0.200 0.250
Chain 1: 9400 -10011.228 0.184 0.215
Chain 1: 9500 -8477.336 0.148 0.181
Chain 1: 9600 -9145.705 0.130 0.089
Chain 1: 9700 -9318.093 0.095 0.073
Chain 1: 9800 -9234.570 0.093 0.073
Chain 1: 9900 -10106.957 0.093 0.073
Chain 1: 10000 -12671.170 0.111 0.086
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001506 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.06 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58473.607 1.000 1.000
Chain 1: 200 -18226.131 1.604 2.208
Chain 1: 300 -8931.451 1.416 1.041
Chain 1: 400 -8150.662 1.086 1.041
Chain 1: 500 -9224.920 0.892 1.000
Chain 1: 600 -8819.011 0.751 1.000
Chain 1: 700 -8519.915 0.649 0.116
Chain 1: 800 -8482.489 0.568 0.116
Chain 1: 900 -7824.074 0.515 0.096
Chain 1: 1000 -8169.379 0.467 0.096
Chain 1: 1100 -7751.096 0.373 0.084
Chain 1: 1200 -7705.125 0.152 0.054
Chain 1: 1300 -7717.661 0.049 0.046
Chain 1: 1400 -7892.420 0.041 0.042
Chain 1: 1500 -7653.198 0.033 0.035
Chain 1: 1600 -7640.131 0.028 0.031
Chain 1: 1700 -7649.815 0.025 0.022
Chain 1: 1800 -7785.612 0.026 0.022
Chain 1: 1900 -7610.150 0.020 0.022
Chain 1: 2000 -7644.875 0.016 0.017
Chain 1: 2100 -7583.398 0.012 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00257 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.7 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86481.875 1.000 1.000
Chain 1: 200 -14015.385 3.085 5.170
Chain 1: 300 -10336.746 2.175 1.000
Chain 1: 400 -11177.721 1.650 1.000
Chain 1: 500 -9166.008 1.364 0.356
Chain 1: 600 -8743.465 1.145 0.356
Chain 1: 700 -9109.589 0.987 0.219
Chain 1: 800 -9297.325 0.866 0.219
Chain 1: 900 -9070.327 0.773 0.075
Chain 1: 1000 -9032.090 0.696 0.075
Chain 1: 1100 -8980.900 0.596 0.048 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8752.416 0.082 0.040
Chain 1: 1300 -9021.547 0.049 0.030
Chain 1: 1400 -8987.129 0.042 0.026
Chain 1: 1500 -8876.825 0.022 0.025
Chain 1: 1600 -8984.361 0.018 0.020
Chain 1: 1700 -9063.486 0.015 0.012
Chain 1: 1800 -8636.042 0.018 0.012
Chain 1: 1900 -8738.915 0.016 0.012
Chain 1: 2000 -8713.845 0.016 0.012
Chain 1: 2100 -8841.255 0.017 0.012
Chain 1: 2200 -8639.760 0.017 0.012
Chain 1: 2300 -8734.395 0.015 0.012
Chain 1: 2400 -8802.065 0.015 0.012
Chain 1: 2500 -8748.270 0.015 0.012
Chain 1: 2600 -8750.900 0.014 0.011
Chain 1: 2700 -8666.968 0.014 0.011
Chain 1: 2800 -8625.349 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002752 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 27.52 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8386467.698 1.000 1.000
Chain 1: 200 -1583637.767 2.648 4.296
Chain 1: 300 -891581.464 2.024 1.000
Chain 1: 400 -458702.372 1.754 1.000
Chain 1: 500 -359205.716 1.459 0.944
Chain 1: 600 -234098.363 1.305 0.944
Chain 1: 700 -120064.752 1.254 0.944
Chain 1: 800 -87171.499 1.144 0.944
Chain 1: 900 -67463.590 1.050 0.776
Chain 1: 1000 -52218.187 0.974 0.776
Chain 1: 1100 -39652.656 0.906 0.534 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38826.485 0.478 0.377
Chain 1: 1300 -26741.141 0.446 0.377
Chain 1: 1400 -26456.973 0.352 0.317
Chain 1: 1500 -23032.517 0.340 0.317
Chain 1: 1600 -22245.532 0.290 0.292
Chain 1: 1700 -21114.306 0.200 0.292
Chain 1: 1800 -21057.368 0.163 0.149
Chain 1: 1900 -21383.723 0.135 0.054
Chain 1: 2000 -19891.689 0.113 0.054
Chain 1: 2100 -20130.376 0.083 0.035
Chain 1: 2200 -20357.292 0.082 0.035
Chain 1: 2300 -19974.004 0.038 0.019
Chain 1: 2400 -19745.962 0.038 0.019
Chain 1: 2500 -19548.068 0.025 0.015
Chain 1: 2600 -19178.042 0.023 0.015
Chain 1: 2700 -19134.915 0.018 0.012
Chain 1: 2800 -18851.699 0.019 0.015
Chain 1: 2900 -19133.123 0.019 0.015
Chain 1: 3000 -19119.296 0.012 0.012
Chain 1: 3100 -19204.304 0.011 0.012
Chain 1: 3200 -18894.882 0.011 0.015
Chain 1: 3300 -19099.675 0.011 0.012
Chain 1: 3400 -18574.390 0.012 0.015
Chain 1: 3500 -19186.649 0.014 0.015
Chain 1: 3600 -18492.870 0.016 0.015
Chain 1: 3700 -18880.046 0.018 0.016
Chain 1: 3800 -17839.043 0.022 0.021
Chain 1: 3900 -17835.175 0.021 0.021
Chain 1: 4000 -17952.477 0.021 0.021
Chain 1: 4100 -17866.200 0.022 0.021
Chain 1: 4200 -17682.273 0.021 0.021
Chain 1: 4300 -17820.788 0.021 0.021
Chain 1: 4400 -17777.509 0.018 0.010
Chain 1: 4500 -17680.011 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001353 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.53 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12933.746 1.000 1.000
Chain 1: 200 -9867.847 0.655 1.000
Chain 1: 300 -8538.027 0.489 0.311
Chain 1: 400 -8725.753 0.372 0.311
Chain 1: 500 -8725.639 0.298 0.156
Chain 1: 600 -8491.727 0.253 0.156
Chain 1: 700 -8397.208 0.218 0.028
Chain 1: 800 -8420.756 0.191 0.028
Chain 1: 900 -8529.210 0.171 0.022
Chain 1: 1000 -8436.935 0.155 0.022
Chain 1: 1100 -8470.340 0.056 0.013
Chain 1: 1200 -8435.559 0.025 0.011
Chain 1: 1300 -8353.272 0.010 0.011
Chain 1: 1400 -8377.163 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001692 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 16.92 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -46933.900 1.000 1.000
Chain 1: 200 -16239.263 1.445 1.890
Chain 1: 300 -9019.353 1.230 1.000
Chain 1: 400 -8637.752 0.934 1.000
Chain 1: 500 -9203.629 0.759 0.800
Chain 1: 600 -8891.491 0.639 0.800
Chain 1: 700 -8430.913 0.555 0.061
Chain 1: 800 -8041.486 0.492 0.061
Chain 1: 900 -7724.678 0.442 0.055
Chain 1: 1000 -7905.180 0.400 0.055
Chain 1: 1100 -8116.387 0.302 0.048
Chain 1: 1200 -7927.067 0.116 0.044
Chain 1: 1300 -7902.219 0.036 0.041
Chain 1: 1400 -8003.198 0.033 0.035
Chain 1: 1500 -7565.542 0.033 0.035
Chain 1: 1600 -7769.015 0.032 0.026
Chain 1: 1700 -7519.589 0.030 0.026
Chain 1: 1800 -7764.944 0.028 0.026
Chain 1: 1900 -7658.071 0.025 0.026
Chain 1: 2000 -7730.200 0.024 0.026
Chain 1: 2100 -7669.471 0.022 0.024
Chain 1: 2200 -7806.765 0.021 0.018
Chain 1: 2300 -7610.561 0.024 0.026
Chain 1: 2400 -7689.018 0.023 0.026
Chain 1: 2500 -7474.567 0.020 0.026
Chain 1: 2600 -7592.343 0.019 0.018
Chain 1: 2700 -7496.800 0.017 0.016
Chain 1: 2800 -7705.729 0.017 0.016
Chain 1: 2900 -7426.860 0.019 0.018
Chain 1: 3000 -7590.749 0.020 0.022
Chain 1: 3100 -7584.631 0.020 0.022
Chain 1: 3200 -7798.089 0.021 0.026
Chain 1: 3300 -7497.531 0.022 0.027
Chain 1: 3400 -7758.709 0.025 0.027
Chain 1: 3500 -7494.497 0.025 0.027
Chain 1: 3600 -7551.756 0.024 0.027
Chain 1: 3700 -7509.544 0.024 0.027
Chain 1: 3800 -7508.834 0.021 0.027
Chain 1: 3900 -7462.295 0.018 0.022
Chain 1: 4000 -7453.324 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002609 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 26.09 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87239.991 1.000 1.000
Chain 1: 200 -14118.695 3.090 5.179
Chain 1: 300 -10439.158 2.177 1.000
Chain 1: 400 -11671.279 1.659 1.000
Chain 1: 500 -9380.680 1.376 0.352
Chain 1: 600 -9677.713 1.152 0.352
Chain 1: 700 -9093.363 0.997 0.244
Chain 1: 800 -9821.141 0.881 0.244
Chain 1: 900 -9296.507 0.790 0.106
Chain 1: 1000 -8836.587 0.716 0.106
Chain 1: 1100 -9201.343 0.620 0.074 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8743.345 0.107 0.064
Chain 1: 1300 -8992.592 0.075 0.056
Chain 1: 1400 -9131.711 0.066 0.052
Chain 1: 1500 -8962.551 0.043 0.052
Chain 1: 1600 -9084.905 0.041 0.052
Chain 1: 1700 -9155.400 0.036 0.040
Chain 1: 1800 -8725.940 0.033 0.040
Chain 1: 1900 -8829.603 0.029 0.028
Chain 1: 2000 -8804.739 0.024 0.019
Chain 1: 2100 -8935.335 0.021 0.015
Chain 1: 2200 -8732.004 0.018 0.015
Chain 1: 2300 -8827.163 0.017 0.015
Chain 1: 2400 -8892.888 0.016 0.013
Chain 1: 2500 -8838.444 0.015 0.012
Chain 1: 2600 -8842.138 0.013 0.011
Chain 1: 2700 -8757.633 0.014 0.011
Chain 1: 2800 -8714.974 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003173 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.73 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8410707.556 1.000 1.000
Chain 1: 200 -1585191.620 2.653 4.306
Chain 1: 300 -892063.173 2.028 1.000
Chain 1: 400 -458932.076 1.757 1.000
Chain 1: 500 -359093.198 1.461 0.944
Chain 1: 600 -233885.622 1.307 0.944
Chain 1: 700 -119967.772 1.256 0.944
Chain 1: 800 -87154.043 1.146 0.944
Chain 1: 900 -67470.462 1.051 0.777
Chain 1: 1000 -52253.690 0.975 0.777
Chain 1: 1100 -39717.594 0.906 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38894.372 0.478 0.377
Chain 1: 1300 -26834.329 0.445 0.377
Chain 1: 1400 -26552.543 0.352 0.316
Chain 1: 1500 -23136.169 0.339 0.316
Chain 1: 1600 -22351.901 0.289 0.292
Chain 1: 1700 -21223.480 0.199 0.291
Chain 1: 1800 -21167.312 0.162 0.148
Chain 1: 1900 -21493.672 0.134 0.053
Chain 1: 2000 -20003.504 0.113 0.053
Chain 1: 2100 -20241.787 0.082 0.035
Chain 1: 2200 -20468.678 0.081 0.035
Chain 1: 2300 -20085.516 0.038 0.019
Chain 1: 2400 -19857.516 0.038 0.019
Chain 1: 2500 -19659.622 0.024 0.015
Chain 1: 2600 -19289.392 0.023 0.015
Chain 1: 2700 -19246.289 0.018 0.012
Chain 1: 2800 -18963.077 0.019 0.015
Chain 1: 2900 -19244.514 0.019 0.015
Chain 1: 3000 -19230.596 0.012 0.012
Chain 1: 3100 -19315.620 0.011 0.011
Chain 1: 3200 -19006.110 0.011 0.015
Chain 1: 3300 -19211.017 0.010 0.011
Chain 1: 3400 -18685.616 0.012 0.015
Chain 1: 3500 -19297.950 0.014 0.015
Chain 1: 3600 -18604.115 0.016 0.015
Chain 1: 3700 -18991.304 0.018 0.016
Chain 1: 3800 -17950.144 0.022 0.020
Chain 1: 3900 -17946.303 0.021 0.020
Chain 1: 4000 -18063.601 0.021 0.020
Chain 1: 4100 -17977.292 0.021 0.020
Chain 1: 4200 -17793.376 0.021 0.020
Chain 1: 4300 -17931.866 0.021 0.020
Chain 1: 4400 -17888.535 0.018 0.010
Chain 1: 4500 -17791.073 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001435 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.35 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12111.165 1.000 1.000
Chain 1: 200 -9084.910 0.667 1.000
Chain 1: 300 -7837.684 0.497 0.333
Chain 1: 400 -8055.804 0.380 0.333
Chain 1: 500 -7922.353 0.307 0.159
Chain 1: 600 -7783.880 0.259 0.159
Chain 1: 700 -7700.596 0.224 0.027
Chain 1: 800 -7711.186 0.196 0.027
Chain 1: 900 -7626.869 0.175 0.018
Chain 1: 1000 -7805.288 0.160 0.023
Chain 1: 1100 -7839.024 0.060 0.018
Chain 1: 1200 -7734.615 0.028 0.017
Chain 1: 1300 -7683.025 0.013 0.013
Chain 1: 1400 -7698.430 0.011 0.011
Chain 1: 1500 -7784.747 0.010 0.011
Chain 1: 1600 -7753.354 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001435 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.35 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61196.333 1.000 1.000
Chain 1: 200 -17586.842 1.740 2.480
Chain 1: 300 -8749.756 1.497 1.010
Chain 1: 400 -8216.328 1.139 1.010
Chain 1: 500 -8262.154 0.912 1.000
Chain 1: 600 -8886.937 0.772 1.000
Chain 1: 700 -8089.824 0.676 0.099
Chain 1: 800 -8084.680 0.591 0.099
Chain 1: 900 -7939.516 0.528 0.070
Chain 1: 1000 -7738.711 0.477 0.070
Chain 1: 1100 -7796.636 0.378 0.065
Chain 1: 1200 -7633.316 0.132 0.026
Chain 1: 1300 -7662.383 0.032 0.021
Chain 1: 1400 -7765.558 0.027 0.018
Chain 1: 1500 -7634.584 0.028 0.018
Chain 1: 1600 -7771.530 0.022 0.018
Chain 1: 1700 -7509.677 0.016 0.018
Chain 1: 1800 -7567.312 0.017 0.018
Chain 1: 1900 -7616.496 0.016 0.017
Chain 1: 2000 -7661.000 0.014 0.013
Chain 1: 2100 -7661.435 0.013 0.013
Chain 1: 2200 -7693.614 0.011 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003134 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.34 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86604.829 1.000 1.000
Chain 1: 200 -13248.298 3.269 5.537
Chain 1: 300 -9644.311 2.304 1.000
Chain 1: 400 -10403.519 1.746 1.000
Chain 1: 500 -8570.438 1.440 0.374
Chain 1: 600 -8511.884 1.201 0.374
Chain 1: 700 -8486.549 1.030 0.214
Chain 1: 800 -9095.022 0.909 0.214
Chain 1: 900 -8398.066 0.817 0.083
Chain 1: 1000 -8289.961 0.737 0.083
Chain 1: 1100 -8452.954 0.639 0.073 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8216.570 0.088 0.067
Chain 1: 1300 -8363.027 0.053 0.029
Chain 1: 1400 -8365.060 0.045 0.019
Chain 1: 1500 -8229.893 0.026 0.018
Chain 1: 1600 -8341.412 0.026 0.018
Chain 1: 1700 -8427.703 0.027 0.018
Chain 1: 1800 -8026.910 0.025 0.018
Chain 1: 1900 -8125.873 0.018 0.016
Chain 1: 2000 -8097.169 0.017 0.016
Chain 1: 2100 -8217.047 0.017 0.015
Chain 1: 2200 -8008.048 0.016 0.015
Chain 1: 2300 -8158.009 0.016 0.015
Chain 1: 2400 -8038.222 0.018 0.015
Chain 1: 2500 -8101.411 0.017 0.015
Chain 1: 2600 -8123.136 0.016 0.015
Chain 1: 2700 -8042.103 0.016 0.015
Chain 1: 2800 -8015.843 0.011 0.012
Chain 1: 2900 -8071.243 0.011 0.010
Chain 1: 3000 -7955.278 0.012 0.015
Chain 1: 3100 -8093.285 0.012 0.015
Chain 1: 3200 -7973.104 0.011 0.015
Chain 1: 3300 -7994.735 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003348 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.48 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8438590.231 1.000 1.000
Chain 1: 200 -1587845.111 2.657 4.314
Chain 1: 300 -891765.627 2.032 1.000
Chain 1: 400 -457744.289 1.761 1.000
Chain 1: 500 -357877.277 1.464 0.948
Chain 1: 600 -232619.987 1.310 0.948
Chain 1: 700 -118905.275 1.260 0.948
Chain 1: 800 -86127.136 1.150 0.948
Chain 1: 900 -66467.221 1.055 0.781
Chain 1: 1000 -51268.558 0.979 0.781
Chain 1: 1100 -38756.033 0.911 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37931.468 0.482 0.381
Chain 1: 1300 -25903.735 0.450 0.381
Chain 1: 1400 -25622.334 0.357 0.323
Chain 1: 1500 -22214.584 0.344 0.323
Chain 1: 1600 -21432.311 0.294 0.296
Chain 1: 1700 -20308.405 0.204 0.296
Chain 1: 1800 -20252.994 0.166 0.153
Chain 1: 1900 -20578.851 0.138 0.055
Chain 1: 2000 -19091.927 0.116 0.055
Chain 1: 2100 -19330.083 0.085 0.036
Chain 1: 2200 -19556.170 0.084 0.036
Chain 1: 2300 -19173.811 0.040 0.020
Chain 1: 2400 -18946.061 0.040 0.020
Chain 1: 2500 -18747.993 0.025 0.016
Chain 1: 2600 -18378.458 0.024 0.016
Chain 1: 2700 -18335.566 0.019 0.012
Chain 1: 2800 -18052.500 0.020 0.016
Chain 1: 2900 -18333.639 0.020 0.015
Chain 1: 3000 -18319.775 0.012 0.012
Chain 1: 3100 -18404.748 0.011 0.012
Chain 1: 3200 -18095.583 0.012 0.015
Chain 1: 3300 -18300.219 0.011 0.012
Chain 1: 3400 -17775.425 0.013 0.015
Chain 1: 3500 -18386.775 0.015 0.016
Chain 1: 3600 -17694.168 0.017 0.016
Chain 1: 3700 -18080.447 0.019 0.017
Chain 1: 3800 -17041.174 0.023 0.021
Chain 1: 3900 -17037.358 0.022 0.021
Chain 1: 4000 -17154.658 0.022 0.021
Chain 1: 4100 -17068.431 0.022 0.021
Chain 1: 4200 -16884.961 0.022 0.021
Chain 1: 4300 -17023.161 0.022 0.021
Chain 1: 4400 -16980.181 0.019 0.011
Chain 1: 4500 -16882.751 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00133 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.3 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48714.700 1.000 1.000
Chain 1: 200 -22810.345 1.068 1.136
Chain 1: 300 -19955.981 0.760 1.000
Chain 1: 400 -14708.066 0.659 1.000
Chain 1: 500 -16272.123 0.546 0.357
Chain 1: 600 -11457.356 0.525 0.420
Chain 1: 700 -15415.812 0.487 0.357
Chain 1: 800 -21441.138 0.461 0.357
Chain 1: 900 -15775.026 0.450 0.357
Chain 1: 1000 -12809.670 0.428 0.357
Chain 1: 1100 -14363.747 0.339 0.281
Chain 1: 1200 -10334.446 0.264 0.281
Chain 1: 1300 -13910.363 0.276 0.281
Chain 1: 1400 -10544.766 0.272 0.281
Chain 1: 1500 -11736.386 0.272 0.281
Chain 1: 1600 -10701.490 0.240 0.257
Chain 1: 1700 -11125.937 0.218 0.257
Chain 1: 1800 -12739.146 0.203 0.231
Chain 1: 1900 -9931.311 0.195 0.231
Chain 1: 2000 -18369.765 0.218 0.257
Chain 1: 2100 -9539.609 0.300 0.283
Chain 1: 2200 -10066.023 0.266 0.257
Chain 1: 2300 -11468.094 0.252 0.127
Chain 1: 2400 -10542.558 0.229 0.122
Chain 1: 2500 -13855.697 0.243 0.127
Chain 1: 2600 -10792.019 0.262 0.239
Chain 1: 2700 -10434.205 0.261 0.239
Chain 1: 2800 -9966.664 0.253 0.239
Chain 1: 2900 -13236.947 0.250 0.239
Chain 1: 3000 -9550.463 0.243 0.239
Chain 1: 3100 -8735.342 0.159 0.122
Chain 1: 3200 -9794.718 0.165 0.122
Chain 1: 3300 -15800.256 0.191 0.239
Chain 1: 3400 -11066.705 0.225 0.247
Chain 1: 3500 -9476.577 0.218 0.247
Chain 1: 3600 -10203.199 0.196 0.168
Chain 1: 3700 -9023.249 0.206 0.168
Chain 1: 3800 -8663.340 0.205 0.168
Chain 1: 3900 -8883.156 0.183 0.131
Chain 1: 4000 -10397.083 0.159 0.131
Chain 1: 4100 -9688.219 0.157 0.131
Chain 1: 4200 -11532.512 0.162 0.146
Chain 1: 4300 -10065.997 0.139 0.146
Chain 1: 4400 -10422.483 0.099 0.131
Chain 1: 4500 -8978.606 0.099 0.131
Chain 1: 4600 -10234.208 0.104 0.131
Chain 1: 4700 -8801.998 0.107 0.146
Chain 1: 4800 -8587.812 0.105 0.146
Chain 1: 4900 -13609.098 0.140 0.146
Chain 1: 5000 -9434.270 0.170 0.160
Chain 1: 5100 -8799.082 0.169 0.160
Chain 1: 5200 -10445.048 0.169 0.158
Chain 1: 5300 -8527.016 0.177 0.161
Chain 1: 5400 -9033.825 0.179 0.161
Chain 1: 5500 -8599.948 0.168 0.158
Chain 1: 5600 -12411.793 0.187 0.163
Chain 1: 5700 -12399.961 0.171 0.158
Chain 1: 5800 -8576.402 0.213 0.225
Chain 1: 5900 -8849.671 0.179 0.158
Chain 1: 6000 -13199.835 0.168 0.158
Chain 1: 6100 -8791.356 0.210 0.225
Chain 1: 6200 -9564.306 0.203 0.225
Chain 1: 6300 -9446.431 0.182 0.081
Chain 1: 6400 -9977.144 0.181 0.081
Chain 1: 6500 -8671.359 0.191 0.151
Chain 1: 6600 -8835.333 0.162 0.081
Chain 1: 6700 -12697.364 0.193 0.151
Chain 1: 6800 -11387.798 0.160 0.115
Chain 1: 6900 -9142.615 0.181 0.151
Chain 1: 7000 -9639.862 0.153 0.115
Chain 1: 7100 -8368.894 0.118 0.115
Chain 1: 7200 -11158.758 0.135 0.151
Chain 1: 7300 -8551.944 0.165 0.152
Chain 1: 7400 -9387.871 0.168 0.152
Chain 1: 7500 -8902.661 0.159 0.152
Chain 1: 7600 -8835.466 0.157 0.152
Chain 1: 7700 -9211.093 0.131 0.115
Chain 1: 7800 -13001.372 0.149 0.152
Chain 1: 7900 -9560.167 0.160 0.152
Chain 1: 8000 -10527.037 0.164 0.152
Chain 1: 8100 -8229.707 0.177 0.250
Chain 1: 8200 -10255.651 0.172 0.198
Chain 1: 8300 -10010.160 0.144 0.092
Chain 1: 8400 -8936.742 0.147 0.120
Chain 1: 8500 -12763.033 0.171 0.198
Chain 1: 8600 -10505.558 0.192 0.215
Chain 1: 8700 -8464.073 0.212 0.241
Chain 1: 8800 -9183.773 0.191 0.215
Chain 1: 8900 -10783.141 0.170 0.198
Chain 1: 9000 -8838.597 0.182 0.215
Chain 1: 9100 -8540.996 0.158 0.198
Chain 1: 9200 -12457.976 0.170 0.215
Chain 1: 9300 -10027.888 0.191 0.220
Chain 1: 9400 -8712.617 0.195 0.220
Chain 1: 9500 -8385.034 0.168 0.215
Chain 1: 9600 -8489.165 0.148 0.151
Chain 1: 9700 -8413.836 0.125 0.148
Chain 1: 9800 -9403.146 0.128 0.148
Chain 1: 9900 -10042.102 0.119 0.105
Chain 1: 10000 -8677.586 0.113 0.105
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001396 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.96 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61810.940 1.000 1.000
Chain 1: 200 -17839.070 1.732 2.465
Chain 1: 300 -8833.914 1.495 1.019
Chain 1: 400 -9371.501 1.135 1.019
Chain 1: 500 -8333.299 0.933 1.000
Chain 1: 600 -8677.151 0.784 1.000
Chain 1: 700 -7912.795 0.686 0.125
Chain 1: 800 -7821.824 0.602 0.125
Chain 1: 900 -7943.674 0.537 0.097
Chain 1: 1000 -7771.068 0.485 0.097
Chain 1: 1100 -7740.434 0.386 0.057
Chain 1: 1200 -7777.197 0.140 0.040
Chain 1: 1300 -7742.440 0.038 0.022
Chain 1: 1400 -7939.378 0.035 0.022
Chain 1: 1500 -7596.618 0.027 0.022
Chain 1: 1600 -7746.793 0.025 0.019
Chain 1: 1700 -7533.960 0.018 0.019
Chain 1: 1800 -7635.076 0.018 0.019
Chain 1: 1900 -7473.991 0.019 0.022
Chain 1: 2000 -7576.272 0.018 0.019
Chain 1: 2100 -7617.009 0.018 0.019
Chain 1: 2200 -7699.285 0.019 0.019
Chain 1: 2300 -7600.589 0.019 0.019
Chain 1: 2400 -7649.086 0.018 0.014
Chain 1: 2500 -7567.774 0.014 0.013
Chain 1: 2600 -7518.817 0.013 0.013
Chain 1: 2700 -7515.003 0.010 0.011
Chain 1: 2800 -7576.485 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002645 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 26.45 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85734.496 1.000 1.000
Chain 1: 200 -13508.040 3.173 5.347
Chain 1: 300 -9947.293 2.235 1.000
Chain 1: 400 -10737.770 1.695 1.000
Chain 1: 500 -8770.391 1.401 0.358
Chain 1: 600 -8537.897 1.172 0.358
Chain 1: 700 -8651.727 1.006 0.224
Chain 1: 800 -9012.489 0.885 0.224
Chain 1: 900 -8757.982 0.790 0.074
Chain 1: 1000 -8542.260 0.714 0.074
Chain 1: 1100 -8769.372 0.616 0.040 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8420.502 0.086 0.040
Chain 1: 1300 -8484.112 0.051 0.029
Chain 1: 1400 -8588.507 0.045 0.027
Chain 1: 1500 -8510.622 0.023 0.026
Chain 1: 1600 -8517.528 0.020 0.025
Chain 1: 1700 -8440.978 0.020 0.025
Chain 1: 1800 -8327.189 0.017 0.014
Chain 1: 1900 -8447.145 0.016 0.014
Chain 1: 2000 -8407.311 0.014 0.012
Chain 1: 2100 -8532.034 0.013 0.012
Chain 1: 2200 -8314.891 0.011 0.012
Chain 1: 2300 -8468.636 0.012 0.014
Chain 1: 2400 -8477.878 0.011 0.014
Chain 1: 2500 -8453.170 0.011 0.014
Chain 1: 2600 -8454.877 0.010 0.014
Chain 1: 2700 -8360.600 0.011 0.014
Chain 1: 2800 -8331.850 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002916 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.16 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8368925.283 1.000 1.000
Chain 1: 200 -1575186.275 2.656 4.313
Chain 1: 300 -888945.757 2.028 1.000
Chain 1: 400 -456969.090 1.758 1.000
Chain 1: 500 -358107.344 1.461 0.945
Chain 1: 600 -233195.080 1.307 0.945
Chain 1: 700 -119376.095 1.256 0.945
Chain 1: 800 -86575.118 1.147 0.945
Chain 1: 900 -66885.652 1.052 0.772
Chain 1: 1000 -51654.444 0.976 0.772
Chain 1: 1100 -39106.385 0.908 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38277.867 0.479 0.379
Chain 1: 1300 -26207.122 0.448 0.379
Chain 1: 1400 -25922.044 0.355 0.321
Chain 1: 1500 -22502.738 0.342 0.321
Chain 1: 1600 -21717.180 0.292 0.295
Chain 1: 1700 -20587.576 0.203 0.294
Chain 1: 1800 -20530.920 0.165 0.152
Chain 1: 1900 -20856.777 0.137 0.055
Chain 1: 2000 -19367.019 0.115 0.055
Chain 1: 2100 -19605.261 0.084 0.036
Chain 1: 2200 -19831.884 0.083 0.036
Chain 1: 2300 -19449.045 0.039 0.020
Chain 1: 2400 -19221.250 0.039 0.020
Chain 1: 2500 -19023.476 0.025 0.016
Chain 1: 2600 -18653.801 0.024 0.016
Chain 1: 2700 -18610.848 0.018 0.012
Chain 1: 2800 -18327.992 0.020 0.015
Chain 1: 2900 -18609.094 0.020 0.015
Chain 1: 3000 -18595.214 0.012 0.012
Chain 1: 3100 -18680.186 0.011 0.012
Chain 1: 3200 -18371.056 0.012 0.015
Chain 1: 3300 -18575.650 0.011 0.012
Chain 1: 3400 -18051.029 0.013 0.015
Chain 1: 3500 -18662.320 0.015 0.015
Chain 1: 3600 -17969.768 0.017 0.015
Chain 1: 3700 -18356.052 0.019 0.017
Chain 1: 3800 -17317.028 0.023 0.021
Chain 1: 3900 -17313.267 0.021 0.021
Chain 1: 4000 -17430.497 0.022 0.021
Chain 1: 4100 -17344.362 0.022 0.021
Chain 1: 4200 -17160.918 0.022 0.021
Chain 1: 4300 -17299.075 0.021 0.021
Chain 1: 4400 -17256.124 0.019 0.011
Chain 1: 4500 -17158.751 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001311 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.11 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12745.384 1.000 1.000
Chain 1: 200 -9610.338 0.663 1.000
Chain 1: 300 -8256.876 0.497 0.326
Chain 1: 400 -8444.929 0.378 0.326
Chain 1: 500 -8310.290 0.306 0.164
Chain 1: 600 -8163.584 0.258 0.164
Chain 1: 700 -8249.135 0.222 0.022
Chain 1: 800 -8104.855 0.197 0.022
Chain 1: 900 -8199.024 0.176 0.018
Chain 1: 1000 -8096.241 0.160 0.018
Chain 1: 1100 -8149.811 0.061 0.018
Chain 1: 1200 -8070.369 0.029 0.016
Chain 1: 1300 -8060.421 0.013 0.013
Chain 1: 1400 -8043.216 0.011 0.011
Chain 1: 1500 -8131.523 0.010 0.011
Chain 1: 1600 -8080.676 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001435 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.35 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57374.131 1.000 1.000
Chain 1: 200 -17813.108 1.610 2.221
Chain 1: 300 -8936.056 1.405 1.000
Chain 1: 400 -8204.771 1.076 1.000
Chain 1: 500 -8884.151 0.876 0.993
Chain 1: 600 -8755.105 0.732 0.993
Chain 1: 700 -8386.800 0.634 0.089
Chain 1: 800 -8138.766 0.559 0.089
Chain 1: 900 -7886.087 0.500 0.076
Chain 1: 1000 -8050.669 0.452 0.076
Chain 1: 1100 -7932.488 0.354 0.044
Chain 1: 1200 -7595.616 0.136 0.044
Chain 1: 1300 -7795.701 0.039 0.032
Chain 1: 1400 -8112.796 0.034 0.032
Chain 1: 1500 -7684.442 0.032 0.032
Chain 1: 1600 -7850.485 0.033 0.032
Chain 1: 1700 -7524.997 0.033 0.032
Chain 1: 1800 -7684.162 0.032 0.032
Chain 1: 1900 -7756.460 0.029 0.026
Chain 1: 2000 -7797.498 0.028 0.026
Chain 1: 2100 -7554.703 0.030 0.032
Chain 1: 2200 -8028.391 0.031 0.032
Chain 1: 2300 -7652.474 0.033 0.039
Chain 1: 2400 -7738.747 0.031 0.032
Chain 1: 2500 -7663.929 0.026 0.021
Chain 1: 2600 -7570.340 0.025 0.021
Chain 1: 2700 -7563.616 0.021 0.012
Chain 1: 2800 -7529.541 0.019 0.011
Chain 1: 2900 -7432.051 0.020 0.012
Chain 1: 3000 -7588.962 0.021 0.013
Chain 1: 3100 -7580.117 0.018 0.012
Chain 1: 3200 -7784.103 0.015 0.012
Chain 1: 3300 -7506.221 0.014 0.012
Chain 1: 3400 -7734.369 0.016 0.013
Chain 1: 3500 -7485.453 0.018 0.021
Chain 1: 3600 -7553.716 0.018 0.021
Chain 1: 3700 -7502.223 0.018 0.021
Chain 1: 3800 -7500.012 0.018 0.021
Chain 1: 3900 -7465.424 0.017 0.021
Chain 1: 4000 -7457.875 0.015 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002612 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 26.12 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86167.725 1.000 1.000
Chain 1: 200 -13871.895 3.106 5.212
Chain 1: 300 -10153.194 2.193 1.000
Chain 1: 400 -11467.565 1.673 1.000
Chain 1: 500 -9172.785 1.389 0.366
Chain 1: 600 -8795.748 1.164 0.366
Chain 1: 700 -9052.854 1.002 0.250
Chain 1: 800 -9468.806 0.882 0.250
Chain 1: 900 -8952.929 0.791 0.115
Chain 1: 1000 -9011.325 0.712 0.115
Chain 1: 1100 -8791.705 0.615 0.058 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8525.912 0.097 0.044
Chain 1: 1300 -8795.817 0.063 0.043
Chain 1: 1400 -8783.908 0.052 0.031
Chain 1: 1500 -8659.396 0.028 0.031
Chain 1: 1600 -8773.398 0.025 0.028
Chain 1: 1700 -8832.527 0.023 0.025
Chain 1: 1800 -8394.954 0.024 0.025
Chain 1: 1900 -8498.736 0.019 0.014
Chain 1: 2000 -8477.119 0.019 0.014
Chain 1: 2100 -8458.544 0.017 0.013
Chain 1: 2200 -8417.050 0.014 0.012
Chain 1: 2300 -8551.857 0.013 0.012
Chain 1: 2400 -8396.904 0.014 0.013
Chain 1: 2500 -8468.058 0.014 0.012
Chain 1: 2600 -8380.657 0.013 0.010
Chain 1: 2700 -8418.168 0.013 0.010
Chain 1: 2800 -8376.142 0.008 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00266 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 26.6 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8372644.597 1.000 1.000
Chain 1: 200 -1582069.919 2.646 4.292
Chain 1: 300 -891219.528 2.022 1.000
Chain 1: 400 -458022.957 1.753 1.000
Chain 1: 500 -358468.170 1.458 0.946
Chain 1: 600 -233592.454 1.304 0.946
Chain 1: 700 -119756.999 1.254 0.946
Chain 1: 800 -86932.827 1.144 0.946
Chain 1: 900 -67268.871 1.050 0.775
Chain 1: 1000 -52056.832 0.974 0.775
Chain 1: 1100 -39512.240 0.906 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38694.966 0.478 0.378
Chain 1: 1300 -26618.284 0.446 0.378
Chain 1: 1400 -26337.517 0.353 0.317
Chain 1: 1500 -22914.629 0.340 0.317
Chain 1: 1600 -22129.101 0.290 0.292
Chain 1: 1700 -20998.181 0.200 0.292
Chain 1: 1800 -20941.747 0.163 0.149
Chain 1: 1900 -21268.364 0.135 0.054
Chain 1: 2000 -19776.154 0.114 0.054
Chain 1: 2100 -20014.864 0.083 0.035
Chain 1: 2200 -20241.906 0.082 0.035
Chain 1: 2300 -19858.461 0.039 0.019
Chain 1: 2400 -19630.341 0.039 0.019
Chain 1: 2500 -19432.440 0.025 0.015
Chain 1: 2600 -19062.054 0.023 0.015
Chain 1: 2700 -19018.918 0.018 0.012
Chain 1: 2800 -18735.520 0.019 0.015
Chain 1: 2900 -19017.087 0.019 0.015
Chain 1: 3000 -19003.292 0.012 0.012
Chain 1: 3100 -19088.302 0.011 0.012
Chain 1: 3200 -18778.653 0.011 0.015
Chain 1: 3300 -18983.689 0.011 0.012
Chain 1: 3400 -18457.995 0.012 0.015
Chain 1: 3500 -19070.773 0.014 0.015
Chain 1: 3600 -18376.387 0.016 0.015
Chain 1: 3700 -18763.947 0.018 0.016
Chain 1: 3800 -17721.950 0.023 0.021
Chain 1: 3900 -17718.086 0.021 0.021
Chain 1: 4000 -17835.385 0.022 0.021
Chain 1: 4100 -17748.979 0.022 0.021
Chain 1: 4200 -17564.950 0.021 0.021
Chain 1: 4300 -17703.561 0.021 0.021
Chain 1: 4400 -17660.085 0.018 0.010
Chain 1: 4500 -17562.586 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00122 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.2 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12328.037 1.000 1.000
Chain 1: 200 -9293.095 0.663 1.000
Chain 1: 300 -8165.281 0.488 0.327
Chain 1: 400 -8161.753 0.366 0.327
Chain 1: 500 -8053.324 0.296 0.138
Chain 1: 600 -7966.074 0.248 0.138
Chain 1: 700 -7869.665 0.215 0.013
Chain 1: 800 -7914.398 0.188 0.013
Chain 1: 900 -8039.114 0.169 0.013
Chain 1: 1000 -7968.687 0.153 0.013
Chain 1: 1100 -7968.620 0.053 0.012
Chain 1: 1200 -7890.818 0.022 0.011
Chain 1: 1300 -7842.246 0.008 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001574 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.74 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56903.337 1.000 1.000
Chain 1: 200 -17411.117 1.634 2.268
Chain 1: 300 -8682.600 1.425 1.005
Chain 1: 400 -8295.532 1.080 1.005
Chain 1: 500 -8338.189 0.865 1.000
Chain 1: 600 -8238.064 0.723 1.000
Chain 1: 700 -7896.944 0.626 0.047
Chain 1: 800 -8154.742 0.552 0.047
Chain 1: 900 -7833.396 0.495 0.043
Chain 1: 1000 -7637.172 0.448 0.043
Chain 1: 1100 -7666.683 0.348 0.041
Chain 1: 1200 -7578.262 0.123 0.032
Chain 1: 1300 -7742.022 0.024 0.026
Chain 1: 1400 -7965.650 0.022 0.026
Chain 1: 1500 -7548.153 0.027 0.028
Chain 1: 1600 -7708.382 0.028 0.028
Chain 1: 1700 -7450.246 0.027 0.028
Chain 1: 1800 -7556.337 0.026 0.026
Chain 1: 1900 -7520.676 0.022 0.021
Chain 1: 2000 -7558.988 0.020 0.021
Chain 1: 2100 -7525.701 0.020 0.021
Chain 1: 2200 -7647.316 0.020 0.021
Chain 1: 2300 -7544.005 0.020 0.016
Chain 1: 2400 -7583.450 0.017 0.014
Chain 1: 2500 -7503.146 0.013 0.014
Chain 1: 2600 -7480.846 0.011 0.011
Chain 1: 2700 -7505.059 0.008 0.005 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002468 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 24.68 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86704.523 1.000 1.000
Chain 1: 200 -13465.586 3.219 5.439
Chain 1: 300 -9850.545 2.269 1.000
Chain 1: 400 -10571.749 1.719 1.000
Chain 1: 500 -8814.131 1.415 0.367
Chain 1: 600 -8678.802 1.182 0.367
Chain 1: 700 -8579.546 1.014 0.199
Chain 1: 800 -9166.440 0.896 0.199
Chain 1: 900 -8644.797 0.803 0.068
Chain 1: 1000 -8465.465 0.725 0.068
Chain 1: 1100 -8593.199 0.626 0.064 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8189.643 0.087 0.060
Chain 1: 1300 -8535.994 0.055 0.049
Chain 1: 1400 -8541.053 0.048 0.041
Chain 1: 1500 -8413.315 0.029 0.021
Chain 1: 1600 -8521.311 0.029 0.021
Chain 1: 1700 -8605.857 0.029 0.021
Chain 1: 1800 -8195.752 0.027 0.021
Chain 1: 1900 -8291.714 0.023 0.015
Chain 1: 2000 -8264.378 0.021 0.015
Chain 1: 2100 -8386.068 0.021 0.015
Chain 1: 2200 -8226.241 0.018 0.015
Chain 1: 2300 -8289.006 0.014 0.013
Chain 1: 2400 -8356.265 0.015 0.013
Chain 1: 2500 -8302.065 0.014 0.012
Chain 1: 2600 -8300.406 0.013 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00366 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 36.6 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8398415.930 1.000 1.000
Chain 1: 200 -1583483.207 2.652 4.304
Chain 1: 300 -890534.891 2.027 1.000
Chain 1: 400 -457191.009 1.757 1.000
Chain 1: 500 -357779.715 1.462 0.948
Chain 1: 600 -232831.173 1.307 0.948
Chain 1: 700 -119143.615 1.257 0.948
Chain 1: 800 -86364.580 1.147 0.948
Chain 1: 900 -66719.788 1.052 0.778
Chain 1: 1000 -51522.946 0.977 0.778
Chain 1: 1100 -39004.675 0.909 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38182.006 0.481 0.380
Chain 1: 1300 -26145.080 0.449 0.380
Chain 1: 1400 -25864.427 0.355 0.321
Chain 1: 1500 -22453.440 0.343 0.321
Chain 1: 1600 -21670.285 0.292 0.295
Chain 1: 1700 -20544.896 0.203 0.294
Chain 1: 1800 -20489.203 0.165 0.152
Chain 1: 1900 -20815.307 0.137 0.055
Chain 1: 2000 -19327.006 0.115 0.055
Chain 1: 2100 -19565.397 0.084 0.036
Chain 1: 2200 -19791.707 0.083 0.036
Chain 1: 2300 -19409.047 0.039 0.020
Chain 1: 2400 -19181.177 0.039 0.020
Chain 1: 2500 -18983.129 0.025 0.016
Chain 1: 2600 -18613.508 0.024 0.016
Chain 1: 2700 -18570.513 0.018 0.012
Chain 1: 2800 -18287.386 0.020 0.015
Chain 1: 2900 -18568.605 0.020 0.015
Chain 1: 3000 -18554.798 0.012 0.012
Chain 1: 3100 -18639.763 0.011 0.012
Chain 1: 3200 -18330.528 0.012 0.015
Chain 1: 3300 -18535.192 0.011 0.012
Chain 1: 3400 -18010.239 0.013 0.015
Chain 1: 3500 -18621.915 0.015 0.015
Chain 1: 3600 -17928.856 0.017 0.015
Chain 1: 3700 -18315.477 0.019 0.017
Chain 1: 3800 -17275.562 0.023 0.021
Chain 1: 3900 -17271.698 0.022 0.021
Chain 1: 4000 -17389.013 0.022 0.021
Chain 1: 4100 -17302.797 0.022 0.021
Chain 1: 4200 -17119.106 0.022 0.021
Chain 1: 4300 -17257.463 0.021 0.021
Chain 1: 4400 -17214.368 0.019 0.011
Chain 1: 4500 -17116.893 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001593 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.93 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49254.700 1.000 1.000
Chain 1: 200 -18304.646 1.345 1.691
Chain 1: 300 -14122.998 0.996 1.000
Chain 1: 400 -25122.250 0.856 1.000
Chain 1: 500 -24851.685 0.687 0.438
Chain 1: 600 -24704.540 0.574 0.438
Chain 1: 700 -11984.466 0.643 0.438
Chain 1: 800 -14398.945 0.584 0.438
Chain 1: 900 -11810.915 0.543 0.296
Chain 1: 1000 -10614.968 0.500 0.296
Chain 1: 1100 -12256.804 0.414 0.219
Chain 1: 1200 -17330.885 0.274 0.219
Chain 1: 1300 -11522.307 0.295 0.219
Chain 1: 1400 -11865.408 0.254 0.168
Chain 1: 1500 -21323.314 0.297 0.219
Chain 1: 1600 -12435.133 0.368 0.293
Chain 1: 1700 -17851.184 0.292 0.293
Chain 1: 1800 -14964.547 0.295 0.293
Chain 1: 1900 -11254.674 0.306 0.303
Chain 1: 2000 -10571.149 0.301 0.303
Chain 1: 2100 -11834.064 0.298 0.303
Chain 1: 2200 -10343.965 0.283 0.303
Chain 1: 2300 -14491.167 0.261 0.286
Chain 1: 2400 -12754.829 0.272 0.286
Chain 1: 2500 -9487.722 0.262 0.286
Chain 1: 2600 -10809.535 0.203 0.193
Chain 1: 2700 -10765.234 0.173 0.144
Chain 1: 2800 -14483.529 0.179 0.144
Chain 1: 2900 -15880.555 0.155 0.136
Chain 1: 3000 -16162.622 0.151 0.136
Chain 1: 3100 -9469.838 0.211 0.144
Chain 1: 3200 -15397.939 0.235 0.257
Chain 1: 3300 -9929.436 0.261 0.257
Chain 1: 3400 -10087.670 0.249 0.257
Chain 1: 3500 -10934.383 0.222 0.122
Chain 1: 3600 -9117.664 0.230 0.199
Chain 1: 3700 -9640.861 0.235 0.199
Chain 1: 3800 -10152.213 0.214 0.088
Chain 1: 3900 -8989.008 0.219 0.129
Chain 1: 4000 -8935.069 0.217 0.129
Chain 1: 4100 -9188.514 0.150 0.077
Chain 1: 4200 -14657.511 0.148 0.077
Chain 1: 4300 -15106.768 0.096 0.054
Chain 1: 4400 -10090.913 0.144 0.077
Chain 1: 4500 -10172.897 0.137 0.054
Chain 1: 4600 -12722.808 0.138 0.054
Chain 1: 4700 -14089.047 0.142 0.097
Chain 1: 4800 -9257.514 0.189 0.129
Chain 1: 4900 -9006.035 0.179 0.097
Chain 1: 5000 -18656.439 0.230 0.200
Chain 1: 5100 -10909.605 0.298 0.373
Chain 1: 5200 -9467.836 0.276 0.200
Chain 1: 5300 -16449.205 0.316 0.424
Chain 1: 5400 -8926.626 0.350 0.424
Chain 1: 5500 -9071.882 0.351 0.424
Chain 1: 5600 -9386.805 0.334 0.424
Chain 1: 5700 -10286.954 0.333 0.424
Chain 1: 5800 -10010.973 0.284 0.152
Chain 1: 5900 -8644.825 0.297 0.158
Chain 1: 6000 -8491.870 0.247 0.152
Chain 1: 6100 -9773.776 0.189 0.131
Chain 1: 6200 -9049.323 0.182 0.088
Chain 1: 6300 -9372.284 0.143 0.080
Chain 1: 6400 -12140.846 0.081 0.080
Chain 1: 6500 -9206.178 0.112 0.088
Chain 1: 6600 -15502.408 0.149 0.131
Chain 1: 6700 -10110.461 0.194 0.158
Chain 1: 6800 -10333.477 0.193 0.158
Chain 1: 6900 -11243.392 0.185 0.131
Chain 1: 7000 -12399.595 0.193 0.131
Chain 1: 7100 -9418.616 0.211 0.228
Chain 1: 7200 -8646.104 0.212 0.228
Chain 1: 7300 -10394.012 0.226 0.228
Chain 1: 7400 -12290.419 0.218 0.168
Chain 1: 7500 -8436.445 0.232 0.168
Chain 1: 7600 -9076.622 0.198 0.154
Chain 1: 7700 -9269.227 0.147 0.093
Chain 1: 7800 -12661.545 0.172 0.154
Chain 1: 7900 -9279.218 0.200 0.168
Chain 1: 8000 -9818.153 0.196 0.168
Chain 1: 8100 -9439.987 0.169 0.154
Chain 1: 8200 -9698.266 0.162 0.154
Chain 1: 8300 -8484.485 0.160 0.143
Chain 1: 8400 -12234.871 0.175 0.143
Chain 1: 8500 -8493.623 0.174 0.143
Chain 1: 8600 -10598.137 0.186 0.199
Chain 1: 8700 -8391.973 0.211 0.263
Chain 1: 8800 -8741.504 0.188 0.199
Chain 1: 8900 -13406.869 0.186 0.199
Chain 1: 9000 -10283.470 0.211 0.263
Chain 1: 9100 -8392.811 0.230 0.263
Chain 1: 9200 -9290.184 0.237 0.263
Chain 1: 9300 -8961.017 0.226 0.263
Chain 1: 9400 -8600.884 0.199 0.225
Chain 1: 9500 -8340.505 0.158 0.199
Chain 1: 9600 -9312.541 0.149 0.104
Chain 1: 9700 -8751.436 0.129 0.097
Chain 1: 9800 -12257.296 0.154 0.104
Chain 1: 9900 -9398.392 0.149 0.104
Chain 1: 10000 -8693.135 0.127 0.097
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001534 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.34 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58573.190 1.000 1.000
Chain 1: 200 -18106.485 1.617 2.235
Chain 1: 300 -8835.165 1.428 1.049
Chain 1: 400 -8094.572 1.094 1.049
Chain 1: 500 -8267.156 0.879 1.000
Chain 1: 600 -9081.307 0.748 1.000
Chain 1: 700 -8806.646 0.645 0.091
Chain 1: 800 -8052.625 0.576 0.094
Chain 1: 900 -7924.427 0.514 0.091
Chain 1: 1000 -7723.949 0.465 0.091
Chain 1: 1100 -7572.248 0.367 0.090
Chain 1: 1200 -7530.622 0.144 0.031
Chain 1: 1300 -7698.945 0.042 0.026
Chain 1: 1400 -7714.489 0.033 0.022
Chain 1: 1500 -7570.319 0.033 0.022
Chain 1: 1600 -7730.174 0.026 0.021
Chain 1: 1700 -7597.570 0.024 0.020
Chain 1: 1800 -7693.538 0.016 0.019
Chain 1: 1900 -7569.403 0.016 0.019
Chain 1: 2000 -7639.202 0.014 0.017
Chain 1: 2100 -7533.106 0.014 0.016
Chain 1: 2200 -7714.259 0.016 0.017
Chain 1: 2300 -7518.871 0.016 0.017
Chain 1: 2400 -7569.117 0.017 0.017
Chain 1: 2500 -7545.420 0.015 0.016
Chain 1: 2600 -7531.942 0.013 0.014
Chain 1: 2700 -7453.807 0.012 0.012
Chain 1: 2800 -7479.848 0.011 0.010
Chain 1: 2900 -7340.791 0.012 0.010
Chain 1: 3000 -7490.925 0.013 0.014
Chain 1: 3100 -7490.927 0.011 0.010
Chain 1: 3200 -7717.858 0.012 0.010
Chain 1: 3300 -7422.964 0.013 0.010
Chain 1: 3400 -7665.491 0.016 0.019
Chain 1: 3500 -7405.245 0.019 0.020
Chain 1: 3600 -7471.771 0.020 0.020
Chain 1: 3700 -7421.248 0.019 0.020
Chain 1: 3800 -7422.663 0.019 0.020
Chain 1: 3900 -7379.692 0.018 0.020
Chain 1: 4000 -7372.297 0.016 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003662 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 36.62 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86421.936 1.000 1.000
Chain 1: 200 -13894.755 3.110 5.220
Chain 1: 300 -10177.694 2.195 1.000
Chain 1: 400 -11314.719 1.671 1.000
Chain 1: 500 -9164.866 1.384 0.365
Chain 1: 600 -9554.415 1.160 0.365
Chain 1: 700 -8945.148 1.004 0.235
Chain 1: 800 -8484.538 0.885 0.235
Chain 1: 900 -8566.604 0.788 0.100
Chain 1: 1000 -8821.752 0.712 0.100
Chain 1: 1100 -8951.483 0.614 0.068 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8538.175 0.096 0.054
Chain 1: 1300 -8852.504 0.064 0.048
Chain 1: 1400 -8784.929 0.054 0.041
Chain 1: 1500 -8681.784 0.032 0.036
Chain 1: 1600 -8792.645 0.029 0.029
Chain 1: 1700 -8855.832 0.023 0.014
Chain 1: 1800 -8416.311 0.023 0.014
Chain 1: 1900 -8521.376 0.023 0.014
Chain 1: 2000 -8500.896 0.020 0.013
Chain 1: 2100 -8491.988 0.019 0.012
Chain 1: 2200 -8438.876 0.015 0.012
Chain 1: 2300 -8574.752 0.013 0.012
Chain 1: 2400 -8419.859 0.014 0.012
Chain 1: 2500 -8491.130 0.014 0.012
Chain 1: 2600 -8403.773 0.013 0.010
Chain 1: 2700 -8441.213 0.013 0.010
Chain 1: 2800 -8399.181 0.008 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00373 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 37.3 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8400046.355 1.000 1.000
Chain 1: 200 -1583613.947 2.652 4.304
Chain 1: 300 -890443.582 2.028 1.000
Chain 1: 400 -458053.459 1.757 1.000
Chain 1: 500 -358564.274 1.461 0.944
Chain 1: 600 -233545.228 1.307 0.944
Chain 1: 700 -119727.859 1.256 0.944
Chain 1: 800 -86935.368 1.146 0.944
Chain 1: 900 -67263.297 1.051 0.778
Chain 1: 1000 -52051.755 0.975 0.778
Chain 1: 1100 -39518.990 0.907 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38697.149 0.479 0.377
Chain 1: 1300 -26633.884 0.446 0.377
Chain 1: 1400 -26352.839 0.353 0.317
Chain 1: 1500 -22935.304 0.340 0.317
Chain 1: 1600 -22151.319 0.290 0.292
Chain 1: 1700 -21022.087 0.200 0.292
Chain 1: 1800 -20965.921 0.163 0.149
Chain 1: 1900 -21292.557 0.135 0.054
Chain 1: 2000 -19801.450 0.113 0.054
Chain 1: 2100 -20039.855 0.083 0.035
Chain 1: 2200 -20267.000 0.082 0.035
Chain 1: 2300 -19883.477 0.038 0.019
Chain 1: 2400 -19655.380 0.039 0.019
Chain 1: 2500 -19457.541 0.025 0.015
Chain 1: 2600 -19087.109 0.023 0.015
Chain 1: 2700 -19043.886 0.018 0.012
Chain 1: 2800 -18760.615 0.019 0.015
Chain 1: 2900 -19042.115 0.019 0.015
Chain 1: 3000 -19028.178 0.012 0.012
Chain 1: 3100 -19113.274 0.011 0.012
Chain 1: 3200 -18803.623 0.011 0.015
Chain 1: 3300 -19008.617 0.011 0.012
Chain 1: 3400 -18482.983 0.012 0.015
Chain 1: 3500 -19095.760 0.014 0.015
Chain 1: 3600 -18401.262 0.016 0.015
Chain 1: 3700 -18788.969 0.018 0.016
Chain 1: 3800 -17746.891 0.022 0.021
Chain 1: 3900 -17743.019 0.021 0.021
Chain 1: 4000 -17860.298 0.022 0.021
Chain 1: 4100 -17773.996 0.022 0.021
Chain 1: 4200 -17589.836 0.021 0.021
Chain 1: 4300 -17728.498 0.021 0.021
Chain 1: 4400 -17685.002 0.018 0.010
Chain 1: 4500 -17587.495 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001599 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.99 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48934.270 1.000 1.000
Chain 1: 200 -18291.271 1.338 1.675
Chain 1: 300 -19281.089 0.909 1.000
Chain 1: 400 -24572.222 0.735 1.000
Chain 1: 500 -12777.974 0.773 0.923
Chain 1: 600 -17840.029 0.691 0.923
Chain 1: 700 -15249.780 0.617 0.284
Chain 1: 800 -13586.099 0.555 0.284
Chain 1: 900 -14897.411 0.503 0.215
Chain 1: 1000 -12817.055 0.469 0.215
Chain 1: 1100 -19251.038 0.403 0.215
Chain 1: 1200 -11199.697 0.307 0.215
Chain 1: 1300 -14146.817 0.323 0.215
Chain 1: 1400 -11982.185 0.319 0.208
Chain 1: 1500 -11468.221 0.231 0.181
Chain 1: 1600 -12342.965 0.210 0.170
Chain 1: 1700 -10572.404 0.210 0.167
Chain 1: 1800 -10426.671 0.199 0.167
Chain 1: 1900 -22596.073 0.244 0.181
Chain 1: 2000 -11127.964 0.331 0.208
Chain 1: 2100 -9550.162 0.314 0.181
Chain 1: 2200 -11213.023 0.257 0.167
Chain 1: 2300 -9640.507 0.252 0.165
Chain 1: 2400 -11271.738 0.249 0.163
Chain 1: 2500 -10421.317 0.252 0.163
Chain 1: 2600 -12408.337 0.261 0.163
Chain 1: 2700 -9741.147 0.272 0.163
Chain 1: 2800 -9491.822 0.273 0.163
Chain 1: 2900 -13536.339 0.249 0.163
Chain 1: 3000 -12239.146 0.157 0.160
Chain 1: 3100 -9102.272 0.175 0.160
Chain 1: 3200 -10143.674 0.170 0.160
Chain 1: 3300 -9507.089 0.161 0.145
Chain 1: 3400 -15663.744 0.185 0.160
Chain 1: 3500 -13626.027 0.192 0.160
Chain 1: 3600 -11019.071 0.200 0.237
Chain 1: 3700 -9200.180 0.192 0.198
Chain 1: 3800 -16610.089 0.234 0.237
Chain 1: 3900 -9295.208 0.283 0.237
Chain 1: 4000 -8853.526 0.277 0.237
Chain 1: 4100 -8861.328 0.243 0.198
Chain 1: 4200 -9169.179 0.236 0.198
Chain 1: 4300 -17723.117 0.278 0.237
Chain 1: 4400 -9805.389 0.319 0.237
Chain 1: 4500 -9064.487 0.312 0.237
Chain 1: 4600 -9675.052 0.295 0.198
Chain 1: 4700 -8589.093 0.288 0.126
Chain 1: 4800 -9940.655 0.257 0.126
Chain 1: 4900 -9563.051 0.182 0.082
Chain 1: 5000 -8704.128 0.187 0.099
Chain 1: 5100 -8613.299 0.188 0.099
Chain 1: 5200 -10468.696 0.202 0.126
Chain 1: 5300 -10217.708 0.157 0.099
Chain 1: 5400 -9063.094 0.089 0.099
Chain 1: 5500 -8813.533 0.083 0.099
Chain 1: 5600 -8646.062 0.079 0.099
Chain 1: 5700 -12859.175 0.099 0.099
Chain 1: 5800 -9062.926 0.127 0.099
Chain 1: 5900 -13309.729 0.155 0.127
Chain 1: 6000 -10321.863 0.174 0.177
Chain 1: 6100 -9291.502 0.184 0.177
Chain 1: 6200 -10267.297 0.176 0.127
Chain 1: 6300 -8513.227 0.194 0.206
Chain 1: 6400 -11579.730 0.208 0.265
Chain 1: 6500 -9634.279 0.225 0.265
Chain 1: 6600 -8835.253 0.232 0.265
Chain 1: 6700 -9896.763 0.210 0.206
Chain 1: 6800 -9693.978 0.171 0.202
Chain 1: 6900 -11015.394 0.151 0.120
Chain 1: 7000 -10091.259 0.131 0.111
Chain 1: 7100 -8557.727 0.138 0.120
Chain 1: 7200 -10710.471 0.148 0.179
Chain 1: 7300 -8544.105 0.153 0.179
Chain 1: 7400 -8972.496 0.131 0.120
Chain 1: 7500 -8555.710 0.116 0.107
Chain 1: 7600 -8695.283 0.109 0.107
Chain 1: 7700 -8452.719 0.101 0.092
Chain 1: 7800 -10316.500 0.117 0.120
Chain 1: 7900 -8386.894 0.128 0.179
Chain 1: 8000 -12023.227 0.149 0.181
Chain 1: 8100 -8416.910 0.174 0.201
Chain 1: 8200 -8238.847 0.156 0.181
Chain 1: 8300 -11579.603 0.159 0.181
Chain 1: 8400 -8599.068 0.189 0.230
Chain 1: 8500 -8194.971 0.189 0.230
Chain 1: 8600 -8297.388 0.189 0.230
Chain 1: 8700 -9736.454 0.201 0.230
Chain 1: 8800 -8271.090 0.200 0.230
Chain 1: 8900 -13343.935 0.215 0.289
Chain 1: 9000 -8235.792 0.247 0.289
Chain 1: 9100 -8280.810 0.205 0.177
Chain 1: 9200 -9039.777 0.211 0.177
Chain 1: 9300 -9190.145 0.184 0.148
Chain 1: 9400 -9656.672 0.154 0.084
Chain 1: 9500 -12239.669 0.170 0.148
Chain 1: 9600 -9686.166 0.195 0.177
Chain 1: 9700 -8334.197 0.197 0.177
Chain 1: 9800 -11178.870 0.205 0.211
Chain 1: 9900 -10616.512 0.172 0.162
Chain 1: 10000 -10657.461 0.110 0.084
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001391 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.91 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57149.516 1.000 1.000
Chain 1: 200 -17640.071 1.620 2.240
Chain 1: 300 -8847.601 1.411 1.000
Chain 1: 400 -8310.726 1.075 1.000
Chain 1: 500 -8618.855 0.867 0.994
Chain 1: 600 -8709.247 0.724 0.994
Chain 1: 700 -8016.038 0.633 0.086
Chain 1: 800 -8278.087 0.558 0.086
Chain 1: 900 -8014.030 0.499 0.065
Chain 1: 1000 -7910.543 0.451 0.065
Chain 1: 1100 -7781.795 0.352 0.036
Chain 1: 1200 -7703.786 0.130 0.033
Chain 1: 1300 -7925.142 0.033 0.032
Chain 1: 1400 -7871.883 0.027 0.028
Chain 1: 1500 -7648.338 0.027 0.028
Chain 1: 1600 -7865.857 0.028 0.028
Chain 1: 1700 -7589.591 0.023 0.028
Chain 1: 1800 -7647.891 0.021 0.028
Chain 1: 1900 -7654.212 0.018 0.017
Chain 1: 2000 -7618.918 0.017 0.017
Chain 1: 2100 -7670.553 0.016 0.010
Chain 1: 2200 -7778.022 0.016 0.014
Chain 1: 2300 -7607.917 0.016 0.014
Chain 1: 2400 -7697.532 0.016 0.014
Chain 1: 2500 -7684.738 0.013 0.012
Chain 1: 2600 -7582.443 0.012 0.012
Chain 1: 2700 -7625.174 0.009 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003486 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.86 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86165.000 1.000 1.000
Chain 1: 200 -13683.160 3.149 5.297
Chain 1: 300 -9996.451 2.222 1.000
Chain 1: 400 -11020.307 1.690 1.000
Chain 1: 500 -8990.624 1.397 0.369
Chain 1: 600 -8726.094 1.169 0.369
Chain 1: 700 -8363.735 1.008 0.226
Chain 1: 800 -8707.303 0.887 0.226
Chain 1: 900 -8713.634 0.789 0.093
Chain 1: 1000 -8763.144 0.710 0.093
Chain 1: 1100 -8694.800 0.611 0.043 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8418.854 0.085 0.039
Chain 1: 1300 -8667.979 0.051 0.033
Chain 1: 1400 -8684.980 0.042 0.030
Chain 1: 1500 -8532.973 0.021 0.029
Chain 1: 1600 -8647.137 0.019 0.018
Chain 1: 1700 -8718.186 0.016 0.013
Chain 1: 1800 -8288.986 0.017 0.013
Chain 1: 1900 -8392.671 0.018 0.013
Chain 1: 2000 -8367.766 0.018 0.013
Chain 1: 2100 -8498.916 0.019 0.015
Chain 1: 2200 -8295.327 0.018 0.015
Chain 1: 2300 -8390.420 0.016 0.013
Chain 1: 2400 -8455.980 0.017 0.013
Chain 1: 2500 -8401.393 0.015 0.012
Chain 1: 2600 -8405.191 0.014 0.011
Chain 1: 2700 -8320.690 0.014 0.011
Chain 1: 2800 -8277.885 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004226 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 42.26 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8388378.433 1.000 1.000
Chain 1: 200 -1580353.255 2.654 4.308
Chain 1: 300 -891557.341 2.027 1.000
Chain 1: 400 -458846.188 1.756 1.000
Chain 1: 500 -359543.580 1.460 0.943
Chain 1: 600 -234219.028 1.306 0.943
Chain 1: 700 -119939.170 1.255 0.943
Chain 1: 800 -87031.399 1.146 0.943
Chain 1: 900 -67263.590 1.051 0.773
Chain 1: 1000 -51980.347 0.975 0.773
Chain 1: 1100 -39387.856 0.907 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38553.115 0.479 0.378
Chain 1: 1300 -26436.843 0.447 0.378
Chain 1: 1400 -26148.728 0.354 0.320
Chain 1: 1500 -22717.682 0.342 0.320
Chain 1: 1600 -21929.019 0.292 0.294
Chain 1: 1700 -20794.062 0.202 0.294
Chain 1: 1800 -20736.120 0.164 0.151
Chain 1: 1900 -21062.469 0.136 0.055
Chain 1: 2000 -19568.579 0.115 0.055
Chain 1: 2100 -19807.216 0.084 0.036
Chain 1: 2200 -20034.665 0.083 0.036
Chain 1: 2300 -19650.898 0.039 0.020
Chain 1: 2400 -19422.816 0.039 0.020
Chain 1: 2500 -19225.176 0.025 0.015
Chain 1: 2600 -18854.888 0.023 0.015
Chain 1: 2700 -18811.575 0.018 0.012
Chain 1: 2800 -18528.595 0.019 0.015
Chain 1: 2900 -18809.954 0.019 0.015
Chain 1: 3000 -18796.010 0.012 0.012
Chain 1: 3100 -18881.138 0.011 0.012
Chain 1: 3200 -18571.565 0.012 0.015
Chain 1: 3300 -18776.423 0.011 0.012
Chain 1: 3400 -18251.127 0.012 0.015
Chain 1: 3500 -18863.513 0.015 0.015
Chain 1: 3600 -18169.430 0.016 0.015
Chain 1: 3700 -18556.948 0.018 0.017
Chain 1: 3800 -17515.623 0.023 0.021
Chain 1: 3900 -17511.767 0.021 0.021
Chain 1: 4000 -17629.016 0.022 0.021
Chain 1: 4100 -17542.838 0.022 0.021
Chain 1: 4200 -17358.759 0.021 0.021
Chain 1: 4300 -17497.338 0.021 0.021
Chain 1: 4400 -17453.990 0.018 0.011
Chain 1: 4500 -17356.486 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001346 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.46 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12733.313 1.000 1.000
Chain 1: 200 -9689.158 0.657 1.000
Chain 1: 300 -8400.411 0.489 0.314
Chain 1: 400 -8578.024 0.372 0.314
Chain 1: 500 -8559.261 0.298 0.153
Chain 1: 600 -8341.935 0.253 0.153
Chain 1: 700 -8253.809 0.218 0.026
Chain 1: 800 -8275.455 0.191 0.026
Chain 1: 900 -8394.629 0.172 0.021
Chain 1: 1000 -8288.485 0.156 0.021
Chain 1: 1100 -8330.028 0.056 0.014
Chain 1: 1200 -8288.231 0.025 0.013
Chain 1: 1300 -8211.520 0.011 0.011
Chain 1: 1400 -8243.818 0.009 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001375 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.75 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -62381.886 1.000 1.000
Chain 1: 200 -18343.654 1.700 2.401
Chain 1: 300 -9085.266 1.473 1.019
Chain 1: 400 -9400.062 1.113 1.019
Chain 1: 500 -8506.980 0.912 1.000
Chain 1: 600 -9097.691 0.771 1.000
Chain 1: 700 -8615.616 0.668 0.105
Chain 1: 800 -8686.505 0.586 0.105
Chain 1: 900 -8101.087 0.529 0.072
Chain 1: 1000 -7798.652 0.480 0.072
Chain 1: 1100 -7798.494 0.380 0.065
Chain 1: 1200 -7725.497 0.141 0.056
Chain 1: 1300 -7623.957 0.040 0.039
Chain 1: 1400 -7906.029 0.040 0.039
Chain 1: 1500 -7629.075 0.033 0.036
Chain 1: 1600 -7867.372 0.030 0.036
Chain 1: 1700 -7563.989 0.028 0.036
Chain 1: 1800 -7650.481 0.029 0.036
Chain 1: 1900 -7624.057 0.022 0.030
Chain 1: 2000 -7692.789 0.019 0.013
Chain 1: 2100 -7530.443 0.021 0.022
Chain 1: 2200 -7936.106 0.025 0.030
Chain 1: 2300 -7546.713 0.029 0.036
Chain 1: 2400 -7730.852 0.028 0.030
Chain 1: 2500 -7639.450 0.025 0.024
Chain 1: 2600 -7561.147 0.023 0.022
Chain 1: 2700 -7482.326 0.020 0.012
Chain 1: 2800 -7647.715 0.021 0.022
Chain 1: 2900 -7412.453 0.024 0.022
Chain 1: 3000 -7567.823 0.025 0.022
Chain 1: 3100 -7554.221 0.024 0.022
Chain 1: 3200 -7768.387 0.021 0.022
Chain 1: 3300 -7479.687 0.020 0.022
Chain 1: 3400 -7721.715 0.021 0.022
Chain 1: 3500 -7468.316 0.023 0.028
Chain 1: 3600 -7533.759 0.023 0.028
Chain 1: 3700 -7484.642 0.022 0.028
Chain 1: 3800 -7483.869 0.020 0.028
Chain 1: 3900 -7444.808 0.017 0.021
Chain 1: 4000 -7436.453 0.015 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003239 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.39 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86282.003 1.000 1.000
Chain 1: 200 -13930.844 3.097 5.194
Chain 1: 300 -10277.566 2.183 1.000
Chain 1: 400 -11422.471 1.662 1.000
Chain 1: 500 -9230.935 1.377 0.355
Chain 1: 600 -8769.009 1.157 0.355
Chain 1: 700 -8837.029 0.992 0.237
Chain 1: 800 -9044.414 0.871 0.237
Chain 1: 900 -9077.322 0.775 0.100
Chain 1: 1000 -8859.135 0.700 0.100
Chain 1: 1100 -9070.560 0.602 0.053 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8685.574 0.087 0.044
Chain 1: 1300 -8929.917 0.054 0.027
Chain 1: 1400 -8964.798 0.045 0.025
Chain 1: 1500 -8811.198 0.023 0.023
Chain 1: 1600 -8922.536 0.019 0.023
Chain 1: 1700 -9001.035 0.019 0.023
Chain 1: 1800 -8575.525 0.022 0.023
Chain 1: 1900 -8677.309 0.022 0.023
Chain 1: 2000 -8652.199 0.020 0.017
Chain 1: 2100 -8778.341 0.019 0.014
Chain 1: 2200 -8578.811 0.017 0.014
Chain 1: 2300 -8672.412 0.016 0.012
Chain 1: 2400 -8740.785 0.016 0.012
Chain 1: 2500 -8687.075 0.015 0.012
Chain 1: 2600 -8689.043 0.014 0.011
Chain 1: 2700 -8605.449 0.014 0.011
Chain 1: 2800 -8564.551 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003501 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.01 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8418844.202 1.000 1.000
Chain 1: 200 -1582815.434 2.659 4.319
Chain 1: 300 -891212.491 2.032 1.000
Chain 1: 400 -458311.149 1.760 1.000
Chain 1: 500 -358868.786 1.463 0.945
Chain 1: 600 -233836.426 1.309 0.945
Chain 1: 700 -119888.446 1.257 0.945
Chain 1: 800 -87063.817 1.147 0.945
Chain 1: 900 -67357.653 1.052 0.776
Chain 1: 1000 -52121.077 0.976 0.776
Chain 1: 1100 -39569.107 0.908 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38739.972 0.478 0.377
Chain 1: 1300 -26661.179 0.446 0.377
Chain 1: 1400 -26376.834 0.353 0.317
Chain 1: 1500 -22955.968 0.340 0.317
Chain 1: 1600 -22170.602 0.290 0.293
Chain 1: 1700 -21040.013 0.200 0.292
Chain 1: 1800 -20983.265 0.163 0.149
Chain 1: 1900 -21309.534 0.135 0.054
Chain 1: 2000 -19818.507 0.113 0.054
Chain 1: 2100 -20056.723 0.083 0.035
Chain 1: 2200 -20283.786 0.082 0.035
Chain 1: 2300 -19900.482 0.038 0.019
Chain 1: 2400 -19672.538 0.039 0.019
Chain 1: 2500 -19474.776 0.025 0.015
Chain 1: 2600 -19104.517 0.023 0.015
Chain 1: 2700 -19061.388 0.018 0.012
Chain 1: 2800 -18778.322 0.019 0.015
Chain 1: 2900 -19059.667 0.019 0.015
Chain 1: 3000 -19045.733 0.012 0.012
Chain 1: 3100 -19130.782 0.011 0.012
Chain 1: 3200 -18821.273 0.011 0.015
Chain 1: 3300 -19026.162 0.011 0.012
Chain 1: 3400 -18500.872 0.012 0.015
Chain 1: 3500 -19113.087 0.014 0.015
Chain 1: 3600 -18419.351 0.016 0.015
Chain 1: 3700 -18806.521 0.018 0.016
Chain 1: 3800 -17765.593 0.022 0.021
Chain 1: 3900 -17761.781 0.021 0.021
Chain 1: 4000 -17879.029 0.022 0.021
Chain 1: 4100 -17792.781 0.022 0.021
Chain 1: 4200 -17608.915 0.021 0.021
Chain 1: 4300 -17747.358 0.021 0.021
Chain 1: 4400 -17704.070 0.018 0.010
Chain 1: 4500 -17606.635 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001571 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.71 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12728.572 1.000 1.000
Chain 1: 200 -9525.019 0.668 1.000
Chain 1: 300 -8077.499 0.505 0.336
Chain 1: 400 -8313.237 0.386 0.336
Chain 1: 500 -8172.686 0.312 0.179
Chain 1: 600 -8245.811 0.262 0.179
Chain 1: 700 -7910.779 0.230 0.042
Chain 1: 800 -7996.441 0.203 0.042
Chain 1: 900 -7839.194 0.183 0.028
Chain 1: 1000 -8066.633 0.167 0.028
Chain 1: 1100 -7950.829 0.069 0.028
Chain 1: 1200 -7932.255 0.035 0.020
Chain 1: 1300 -7882.176 0.018 0.017
Chain 1: 1400 -7903.883 0.015 0.015
Chain 1: 1500 -7996.414 0.015 0.012
Chain 1: 1600 -7914.943 0.015 0.012
Chain 1: 1700 -7882.395 0.011 0.011
Chain 1: 1800 -7855.252 0.010 0.010
Chain 1: 1900 -7882.396 0.009 0.006 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001377 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.77 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61978.673 1.000 1.000
Chain 1: 200 -18209.155 1.702 2.404
Chain 1: 300 -9024.271 1.474 1.018
Chain 1: 400 -8472.404 1.122 1.018
Chain 1: 500 -8647.845 0.901 1.000
Chain 1: 600 -8195.521 0.760 1.000
Chain 1: 700 -8465.488 0.656 0.065
Chain 1: 800 -8303.441 0.577 0.065
Chain 1: 900 -7970.299 0.517 0.055
Chain 1: 1000 -7705.704 0.469 0.055
Chain 1: 1100 -7834.143 0.371 0.042
Chain 1: 1200 -7695.887 0.132 0.034
Chain 1: 1300 -7573.887 0.032 0.032
Chain 1: 1400 -7780.316 0.028 0.027
Chain 1: 1500 -7530.390 0.029 0.032
Chain 1: 1600 -7808.208 0.027 0.032
Chain 1: 1700 -7425.870 0.029 0.033
Chain 1: 1800 -7577.488 0.029 0.033
Chain 1: 1900 -7559.485 0.025 0.027
Chain 1: 2000 -7636.179 0.023 0.020
Chain 1: 2100 -7547.440 0.023 0.020
Chain 1: 2200 -7709.819 0.023 0.021
Chain 1: 2300 -7559.155 0.023 0.021
Chain 1: 2400 -7560.796 0.021 0.020
Chain 1: 2500 -7554.094 0.017 0.020
Chain 1: 2600 -7485.472 0.015 0.012
Chain 1: 2700 -7476.747 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00252 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.2 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86915.837 1.000 1.000
Chain 1: 200 -13855.536 3.137 5.273
Chain 1: 300 -10096.711 2.215 1.000
Chain 1: 400 -11691.669 1.695 1.000
Chain 1: 500 -8722.035 1.424 0.372
Chain 1: 600 -9005.781 1.192 0.372
Chain 1: 700 -8532.483 1.030 0.340
Chain 1: 800 -9296.308 0.911 0.340
Chain 1: 900 -8986.340 0.814 0.136
Chain 1: 1000 -8413.457 0.739 0.136
Chain 1: 1100 -8885.283 0.645 0.082 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8376.924 0.123 0.068
Chain 1: 1300 -8710.308 0.090 0.061
Chain 1: 1400 -8700.011 0.077 0.055
Chain 1: 1500 -8550.401 0.044 0.053
Chain 1: 1600 -8664.599 0.042 0.053
Chain 1: 1700 -8712.258 0.037 0.038
Chain 1: 1800 -8257.525 0.035 0.038
Chain 1: 1900 -8369.333 0.033 0.038
Chain 1: 2000 -8374.008 0.026 0.017
Chain 1: 2100 -8484.361 0.022 0.013
Chain 1: 2200 -8263.380 0.018 0.013
Chain 1: 2300 -8404.623 0.016 0.013
Chain 1: 2400 -8270.644 0.018 0.016
Chain 1: 2500 -8342.814 0.017 0.013
Chain 1: 2600 -8253.090 0.017 0.013
Chain 1: 2700 -8286.004 0.017 0.013
Chain 1: 2800 -8237.395 0.012 0.013
Chain 1: 2900 -8349.989 0.012 0.013
Chain 1: 3000 -8278.242 0.012 0.013
Chain 1: 3100 -8229.484 0.012 0.011
Chain 1: 3200 -8202.317 0.009 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004138 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 41.38 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8422611.169 1.000 1.000
Chain 1: 200 -1585286.027 2.656 4.313
Chain 1: 300 -889888.457 2.031 1.000
Chain 1: 400 -456933.081 1.760 1.000
Chain 1: 500 -357122.374 1.464 0.948
Chain 1: 600 -232257.265 1.310 0.948
Chain 1: 700 -119074.899 1.259 0.948
Chain 1: 800 -86436.970 1.148 0.948
Chain 1: 900 -66895.070 1.053 0.781
Chain 1: 1000 -51786.379 0.977 0.781
Chain 1: 1100 -39342.672 0.909 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38538.519 0.480 0.378
Chain 1: 1300 -26555.036 0.447 0.378
Chain 1: 1400 -26284.637 0.353 0.316
Chain 1: 1500 -22886.872 0.340 0.316
Chain 1: 1600 -22109.555 0.289 0.292
Chain 1: 1700 -20989.224 0.200 0.292
Chain 1: 1800 -20935.650 0.162 0.148
Chain 1: 1900 -21262.564 0.135 0.053
Chain 1: 2000 -19775.698 0.113 0.053
Chain 1: 2100 -20013.970 0.082 0.035
Chain 1: 2200 -20240.408 0.081 0.035
Chain 1: 2300 -19857.479 0.038 0.019
Chain 1: 2400 -19629.376 0.038 0.019
Chain 1: 2500 -19431.146 0.025 0.015
Chain 1: 2600 -19060.750 0.023 0.015
Chain 1: 2700 -19017.678 0.018 0.012
Chain 1: 2800 -18734.051 0.019 0.015
Chain 1: 2900 -19015.591 0.019 0.015
Chain 1: 3000 -19001.797 0.012 0.012
Chain 1: 3100 -19086.841 0.011 0.012
Chain 1: 3200 -18777.088 0.011 0.015
Chain 1: 3300 -18982.224 0.011 0.012
Chain 1: 3400 -18456.218 0.012 0.015
Chain 1: 3500 -19069.307 0.014 0.015
Chain 1: 3600 -18374.447 0.016 0.015
Chain 1: 3700 -18762.300 0.018 0.016
Chain 1: 3800 -17719.489 0.023 0.021
Chain 1: 3900 -17715.550 0.021 0.021
Chain 1: 4000 -17832.910 0.022 0.021
Chain 1: 4100 -17746.455 0.022 0.021
Chain 1: 4200 -17562.244 0.021 0.021
Chain 1: 4300 -17701.004 0.021 0.021
Chain 1: 4400 -17657.364 0.018 0.010
Chain 1: 4500 -17559.814 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001673 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 16.73 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12382.537 1.000 1.000
Chain 1: 200 -9285.391 0.667 1.000
Chain 1: 300 -8050.047 0.496 0.334
Chain 1: 400 -8221.065 0.377 0.334
Chain 1: 500 -8167.828 0.303 0.153
Chain 1: 600 -7995.388 0.256 0.153
Chain 1: 700 -7907.836 0.221 0.022
Chain 1: 800 -7915.826 0.193 0.022
Chain 1: 900 -7830.396 0.173 0.021
Chain 1: 1000 -8016.859 0.158 0.022
Chain 1: 1100 -8050.774 0.059 0.021
Chain 1: 1200 -7939.765 0.027 0.014
Chain 1: 1300 -7885.101 0.012 0.011
Chain 1: 1400 -7904.557 0.010 0.011
Chain 1: 1500 -7991.492 0.011 0.011
Chain 1: 1600 -7955.913 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001806 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 18.06 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61528.184 1.000 1.000
Chain 1: 200 -17893.656 1.719 2.439
Chain 1: 300 -8841.550 1.487 1.024
Chain 1: 400 -9416.211 1.131 1.024
Chain 1: 500 -8563.708 0.925 1.000
Chain 1: 600 -8655.318 0.772 1.000
Chain 1: 700 -7772.309 0.678 0.114
Chain 1: 800 -8865.258 0.609 0.123
Chain 1: 900 -7938.205 0.554 0.117
Chain 1: 1000 -7694.677 0.502 0.117
Chain 1: 1100 -7792.031 0.403 0.114
Chain 1: 1200 -7786.123 0.159 0.100
Chain 1: 1300 -7531.814 0.060 0.061
Chain 1: 1400 -7797.247 0.058 0.034
Chain 1: 1500 -7578.369 0.051 0.034
Chain 1: 1600 -7743.644 0.052 0.034
Chain 1: 1700 -7484.682 0.044 0.034
Chain 1: 1800 -7565.946 0.033 0.032
Chain 1: 1900 -7565.759 0.021 0.029
Chain 1: 2000 -7596.027 0.018 0.021
Chain 1: 2100 -7568.935 0.017 0.021
Chain 1: 2200 -7682.973 0.019 0.021
Chain 1: 2300 -7573.276 0.017 0.015
Chain 1: 2400 -7626.211 0.014 0.014
Chain 1: 2500 -7542.374 0.012 0.011
Chain 1: 2600 -7490.003 0.011 0.011
Chain 1: 2700 -7487.102 0.007 0.007 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003651 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 36.51 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85550.852 1.000 1.000
Chain 1: 200 -13511.171 3.166 5.332
Chain 1: 300 -9890.249 2.233 1.000
Chain 1: 400 -10637.004 1.692 1.000
Chain 1: 500 -8863.242 1.394 0.366
Chain 1: 600 -8359.410 1.171 0.366
Chain 1: 700 -8386.872 1.005 0.200
Chain 1: 800 -8847.744 0.885 0.200
Chain 1: 900 -8630.065 0.790 0.070
Chain 1: 1000 -8392.517 0.714 0.070
Chain 1: 1100 -8760.372 0.618 0.060 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8344.238 0.090 0.052
Chain 1: 1300 -8475.252 0.055 0.050
Chain 1: 1400 -8562.581 0.049 0.042
Chain 1: 1500 -8454.519 0.030 0.028
Chain 1: 1600 -8559.980 0.025 0.025
Chain 1: 1700 -8649.001 0.026 0.025
Chain 1: 1800 -8237.752 0.026 0.025
Chain 1: 1900 -8333.835 0.024 0.015
Chain 1: 2000 -8306.584 0.022 0.013
Chain 1: 2100 -8428.581 0.019 0.013
Chain 1: 2200 -8248.605 0.016 0.013
Chain 1: 2300 -8329.669 0.016 0.012
Chain 1: 2400 -8398.444 0.015 0.012
Chain 1: 2500 -8343.867 0.015 0.012
Chain 1: 2600 -8342.672 0.014 0.010
Chain 1: 2700 -8259.958 0.014 0.010
Chain 1: 2800 -8224.602 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003411 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.11 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8372237.706 1.000 1.000
Chain 1: 200 -1578553.894 2.652 4.304
Chain 1: 300 -890764.230 2.025 1.000
Chain 1: 400 -457926.652 1.755 1.000
Chain 1: 500 -358801.920 1.459 0.945
Chain 1: 600 -233745.434 1.305 0.945
Chain 1: 700 -119633.431 1.255 0.945
Chain 1: 800 -86737.934 1.146 0.945
Chain 1: 900 -67005.359 1.051 0.772
Chain 1: 1000 -51745.773 0.975 0.772
Chain 1: 1100 -39165.047 0.908 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38335.874 0.479 0.379
Chain 1: 1300 -26235.504 0.448 0.379
Chain 1: 1400 -25949.004 0.355 0.321
Chain 1: 1500 -22521.316 0.342 0.321
Chain 1: 1600 -21733.127 0.293 0.295
Chain 1: 1700 -20600.190 0.203 0.294
Chain 1: 1800 -20542.932 0.165 0.152
Chain 1: 1900 -20868.890 0.137 0.055
Chain 1: 2000 -19376.774 0.115 0.055
Chain 1: 2100 -19615.296 0.084 0.036
Chain 1: 2200 -19842.216 0.083 0.036
Chain 1: 2300 -19459.067 0.039 0.020
Chain 1: 2400 -19231.139 0.039 0.020
Chain 1: 2500 -19033.349 0.025 0.016
Chain 1: 2600 -18663.398 0.024 0.016
Chain 1: 2700 -18620.363 0.018 0.012
Chain 1: 2800 -18337.301 0.020 0.015
Chain 1: 2900 -18618.620 0.020 0.015
Chain 1: 3000 -18604.782 0.012 0.012
Chain 1: 3100 -18689.748 0.011 0.012
Chain 1: 3200 -18380.426 0.012 0.015
Chain 1: 3300 -18585.161 0.011 0.012
Chain 1: 3400 -18060.167 0.013 0.015
Chain 1: 3500 -18671.991 0.015 0.015
Chain 1: 3600 -17978.821 0.017 0.015
Chain 1: 3700 -18365.530 0.019 0.017
Chain 1: 3800 -17325.479 0.023 0.021
Chain 1: 3900 -17321.682 0.021 0.021
Chain 1: 4000 -17438.941 0.022 0.021
Chain 1: 4100 -17352.696 0.022 0.021
Chain 1: 4200 -17169.050 0.022 0.021
Chain 1: 4300 -17307.354 0.021 0.021
Chain 1: 4400 -17264.216 0.019 0.011
Chain 1: 4500 -17166.818 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00149 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.9 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49572.401 1.000 1.000
Chain 1: 200 -17598.358 1.408 1.817
Chain 1: 300 -14294.679 1.016 1.000
Chain 1: 400 -12561.523 0.796 1.000
Chain 1: 500 -13881.521 0.656 0.231
Chain 1: 600 -12087.608 0.572 0.231
Chain 1: 700 -11531.190 0.497 0.148
Chain 1: 800 -12445.358 0.444 0.148
Chain 1: 900 -14324.624 0.409 0.138
Chain 1: 1000 -13277.937 0.376 0.138
Chain 1: 1100 -15171.201 0.289 0.131
Chain 1: 1200 -12291.937 0.130 0.131
Chain 1: 1300 -11037.864 0.119 0.125
Chain 1: 1400 -17442.338 0.142 0.125
Chain 1: 1500 -10591.038 0.197 0.131
Chain 1: 1600 -11137.086 0.187 0.125
Chain 1: 1700 -11188.343 0.182 0.125
Chain 1: 1800 -11836.674 0.181 0.125
Chain 1: 1900 -24198.236 0.218 0.125
Chain 1: 2000 -11807.210 0.316 0.234
Chain 1: 2100 -11111.108 0.309 0.234
Chain 1: 2200 -10798.723 0.289 0.114
Chain 1: 2300 -17945.254 0.317 0.367
Chain 1: 2400 -10267.689 0.355 0.398
Chain 1: 2500 -16798.871 0.330 0.389
Chain 1: 2600 -11061.748 0.376 0.398
Chain 1: 2700 -12045.329 0.384 0.398
Chain 1: 2800 -11808.331 0.381 0.398
Chain 1: 2900 -9685.075 0.352 0.389
Chain 1: 3000 -9614.292 0.247 0.219
Chain 1: 3100 -16035.248 0.281 0.389
Chain 1: 3200 -11490.948 0.318 0.395
Chain 1: 3300 -10662.288 0.286 0.389
Chain 1: 3400 -14912.541 0.239 0.285
Chain 1: 3500 -15437.436 0.204 0.219
Chain 1: 3600 -9780.878 0.210 0.219
Chain 1: 3700 -9569.301 0.204 0.219
Chain 1: 3800 -9564.399 0.202 0.219
Chain 1: 3900 -13845.116 0.211 0.285
Chain 1: 4000 -9501.504 0.256 0.309
Chain 1: 4100 -9325.193 0.218 0.285
Chain 1: 4200 -11495.997 0.197 0.189
Chain 1: 4300 -10063.683 0.204 0.189
Chain 1: 4400 -9115.847 0.186 0.142
Chain 1: 4500 -9166.025 0.183 0.142
Chain 1: 4600 -13201.281 0.155 0.142
Chain 1: 4700 -10425.384 0.180 0.189
Chain 1: 4800 -9214.674 0.193 0.189
Chain 1: 4900 -9280.779 0.163 0.142
Chain 1: 5000 -17221.477 0.163 0.142
Chain 1: 5100 -12952.320 0.194 0.189
Chain 1: 5200 -9268.661 0.215 0.266
Chain 1: 5300 -10257.427 0.210 0.266
Chain 1: 5400 -11364.707 0.210 0.266
Chain 1: 5500 -12973.910 0.222 0.266
Chain 1: 5600 -16474.122 0.212 0.212
Chain 1: 5700 -9835.900 0.253 0.212
Chain 1: 5800 -18814.122 0.288 0.330
Chain 1: 5900 -14211.715 0.319 0.330
Chain 1: 6000 -11743.086 0.294 0.324
Chain 1: 6100 -15010.605 0.283 0.218
Chain 1: 6200 -9186.710 0.307 0.218
Chain 1: 6300 -15650.840 0.338 0.324
Chain 1: 6400 -9495.021 0.394 0.413
Chain 1: 6500 -9076.379 0.386 0.413
Chain 1: 6600 -9007.379 0.365 0.413
Chain 1: 6700 -12805.236 0.327 0.324
Chain 1: 6800 -8942.518 0.323 0.324
Chain 1: 6900 -13276.032 0.323 0.326
Chain 1: 7000 -9610.328 0.340 0.381
Chain 1: 7100 -9574.445 0.319 0.381
Chain 1: 7200 -12334.507 0.278 0.326
Chain 1: 7300 -12311.373 0.237 0.297
Chain 1: 7400 -8666.078 0.214 0.297
Chain 1: 7500 -9470.182 0.218 0.297
Chain 1: 7600 -9045.240 0.222 0.297
Chain 1: 7700 -9139.362 0.193 0.224
Chain 1: 7800 -12232.190 0.175 0.224
Chain 1: 7900 -8698.317 0.183 0.224
Chain 1: 8000 -9679.981 0.155 0.101
Chain 1: 8100 -9378.247 0.158 0.101
Chain 1: 8200 -13047.922 0.164 0.101
Chain 1: 8300 -8786.595 0.212 0.253
Chain 1: 8400 -9057.914 0.173 0.101
Chain 1: 8500 -9319.408 0.167 0.101
Chain 1: 8600 -9287.445 0.163 0.101
Chain 1: 8700 -9049.243 0.165 0.101
Chain 1: 8800 -9271.419 0.142 0.032
Chain 1: 8900 -9874.521 0.107 0.032
Chain 1: 9000 -8576.148 0.112 0.032
Chain 1: 9100 -14169.868 0.149 0.061
Chain 1: 9200 -8794.866 0.182 0.061
Chain 1: 9300 -8879.276 0.134 0.030
Chain 1: 9400 -9529.307 0.138 0.061
Chain 1: 9500 -8689.593 0.145 0.068
Chain 1: 9600 -11616.124 0.169 0.097
Chain 1: 9700 -8859.623 0.198 0.151
Chain 1: 9800 -9667.799 0.204 0.151
Chain 1: 9900 -11571.933 0.214 0.165
Chain 1: 10000 -8589.986 0.234 0.252
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001851 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 18.51 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -59023.376 1.000 1.000
Chain 1: 200 -18372.161 1.606 2.213
Chain 1: 300 -8959.756 1.421 1.051
Chain 1: 400 -8119.710 1.092 1.051
Chain 1: 500 -8176.313 0.875 1.000
Chain 1: 600 -8270.803 0.731 1.000
Chain 1: 700 -7768.138 0.636 0.103
Chain 1: 800 -8315.760 0.564 0.103
Chain 1: 900 -7848.737 0.508 0.066
Chain 1: 1000 -7831.688 0.458 0.066
Chain 1: 1100 -7696.978 0.359 0.065
Chain 1: 1200 -7626.022 0.139 0.060
Chain 1: 1300 -7544.025 0.035 0.018
Chain 1: 1400 -7577.948 0.025 0.011
Chain 1: 1500 -7480.344 0.026 0.013
Chain 1: 1600 -7662.615 0.027 0.018
Chain 1: 1700 -7592.538 0.022 0.013
Chain 1: 1800 -7540.547 0.016 0.011
Chain 1: 1900 -7488.503 0.010 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003825 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 38.25 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86174.310 1.000 1.000
Chain 1: 200 -14180.535 3.038 5.077
Chain 1: 300 -10420.224 2.146 1.000
Chain 1: 400 -12049.341 1.643 1.000
Chain 1: 500 -9301.482 1.374 0.361
Chain 1: 600 -8863.620 1.153 0.361
Chain 1: 700 -9202.603 0.994 0.295
Chain 1: 800 -10021.514 0.880 0.295
Chain 1: 900 -9270.892 0.791 0.135
Chain 1: 1000 -8782.099 0.717 0.135
Chain 1: 1100 -9255.914 0.622 0.082 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8719.046 0.121 0.081
Chain 1: 1300 -9040.700 0.088 0.062
Chain 1: 1400 -8906.594 0.076 0.056
Chain 1: 1500 -8900.781 0.047 0.051
Chain 1: 1600 -8995.442 0.043 0.051
Chain 1: 1700 -9048.528 0.040 0.051
Chain 1: 1800 -8594.551 0.037 0.051
Chain 1: 1900 -8705.322 0.030 0.036
Chain 1: 2000 -8709.865 0.025 0.015
Chain 1: 2100 -8655.246 0.020 0.013
Chain 1: 2200 -8627.662 0.014 0.011
Chain 1: 2300 -8808.674 0.013 0.011
Chain 1: 2400 -8601.336 0.014 0.011
Chain 1: 2500 -8673.783 0.014 0.011
Chain 1: 2600 -8585.345 0.014 0.010
Chain 1: 2700 -8622.539 0.014 0.010
Chain 1: 2800 -8573.847 0.010 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003136 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.36 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8366860.229 1.000 1.000
Chain 1: 200 -1583276.693 2.642 4.285
Chain 1: 300 -892033.721 2.020 1.000
Chain 1: 400 -458712.523 1.751 1.000
Chain 1: 500 -359289.732 1.456 0.945
Chain 1: 600 -234149.050 1.303 0.945
Chain 1: 700 -120185.920 1.252 0.945
Chain 1: 800 -87339.812 1.142 0.945
Chain 1: 900 -67648.394 1.048 0.775
Chain 1: 1000 -52427.529 0.972 0.775
Chain 1: 1100 -39871.353 0.904 0.534 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39053.095 0.477 0.376
Chain 1: 1300 -26960.844 0.445 0.376
Chain 1: 1400 -26679.576 0.351 0.315
Chain 1: 1500 -23252.837 0.338 0.315
Chain 1: 1600 -22466.488 0.288 0.291
Chain 1: 1700 -21333.542 0.199 0.290
Chain 1: 1800 -21276.806 0.161 0.147
Chain 1: 1900 -21603.673 0.134 0.053
Chain 1: 2000 -20109.833 0.112 0.053
Chain 1: 2100 -20348.640 0.082 0.035
Chain 1: 2200 -20576.091 0.081 0.035
Chain 1: 2300 -20192.208 0.038 0.019
Chain 1: 2400 -19963.942 0.038 0.019
Chain 1: 2500 -19766.137 0.024 0.015
Chain 1: 2600 -19395.446 0.023 0.015
Chain 1: 2700 -19352.169 0.018 0.012
Chain 1: 2800 -19068.695 0.019 0.015
Chain 1: 2900 -19350.407 0.019 0.015
Chain 1: 3000 -19336.541 0.011 0.012
Chain 1: 3100 -19421.626 0.011 0.011
Chain 1: 3200 -19111.804 0.011 0.015
Chain 1: 3300 -19316.947 0.010 0.011
Chain 1: 3400 -18790.958 0.012 0.015
Chain 1: 3500 -19404.267 0.014 0.015
Chain 1: 3600 -18709.153 0.016 0.015
Chain 1: 3700 -19097.276 0.018 0.016
Chain 1: 3800 -18054.232 0.022 0.020
Chain 1: 3900 -18050.337 0.021 0.020
Chain 1: 4000 -18167.628 0.021 0.020
Chain 1: 4100 -18081.216 0.021 0.020
Chain 1: 4200 -17896.898 0.021 0.020
Chain 1: 4300 -18035.690 0.020 0.020
Chain 1: 4400 -17992.018 0.018 0.010
Chain 1: 4500 -17894.484 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001862 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 18.62 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48856.766 1.000 1.000
Chain 1: 200 -22633.214 1.079 1.159
Chain 1: 300 -16272.758 0.850 1.000
Chain 1: 400 -23327.939 0.713 1.000
Chain 1: 500 -12074.688 0.757 0.932
Chain 1: 600 -30662.019 0.732 0.932
Chain 1: 700 -13340.661 0.813 0.932
Chain 1: 800 -10701.548 0.742 0.932
Chain 1: 900 -12830.452 0.678 0.606
Chain 1: 1000 -18546.235 0.641 0.606
Chain 1: 1100 -14370.894 0.570 0.391 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -12864.001 0.466 0.308
Chain 1: 1300 -11310.570 0.440 0.302
Chain 1: 1400 -12070.319 0.417 0.291
Chain 1: 1500 -10320.729 0.340 0.247
Chain 1: 1600 -9897.401 0.284 0.170
Chain 1: 1700 -11481.793 0.168 0.166
Chain 1: 1800 -12487.177 0.151 0.138
Chain 1: 1900 -10310.256 0.156 0.138
Chain 1: 2000 -10316.663 0.125 0.137
Chain 1: 2100 -9490.606 0.105 0.117
Chain 1: 2200 -10549.406 0.103 0.100
Chain 1: 2300 -9508.348 0.100 0.100
Chain 1: 2400 -11534.454 0.112 0.109
Chain 1: 2500 -9610.781 0.115 0.109
Chain 1: 2600 -9935.494 0.114 0.109
Chain 1: 2700 -9261.934 0.107 0.100
Chain 1: 2800 -14335.944 0.134 0.109
Chain 1: 2900 -9201.114 0.169 0.109
Chain 1: 3000 -17662.339 0.217 0.176
Chain 1: 3100 -8629.262 0.313 0.200
Chain 1: 3200 -9074.226 0.308 0.200
Chain 1: 3300 -13359.563 0.329 0.321
Chain 1: 3400 -17651.044 0.336 0.321
Chain 1: 3500 -9069.741 0.410 0.354
Chain 1: 3600 -9789.113 0.414 0.354
Chain 1: 3700 -13219.155 0.433 0.354
Chain 1: 3800 -9190.568 0.441 0.438
Chain 1: 3900 -10560.637 0.399 0.321
Chain 1: 4000 -9600.107 0.361 0.259
Chain 1: 4100 -9620.070 0.256 0.243
Chain 1: 4200 -15303.625 0.288 0.259
Chain 1: 4300 -10004.516 0.309 0.259
Chain 1: 4400 -13090.809 0.309 0.259
Chain 1: 4500 -9326.885 0.254 0.259
Chain 1: 4600 -10195.152 0.256 0.259
Chain 1: 4700 -9517.941 0.237 0.236
Chain 1: 4800 -8676.834 0.203 0.130
Chain 1: 4900 -13604.093 0.226 0.236
Chain 1: 5000 -9758.993 0.255 0.362
Chain 1: 5100 -8823.967 0.266 0.362
Chain 1: 5200 -11167.216 0.249 0.236
Chain 1: 5300 -12835.605 0.209 0.210
Chain 1: 5400 -10385.059 0.209 0.210
Chain 1: 5500 -8520.300 0.191 0.210
Chain 1: 5600 -13347.501 0.219 0.219
Chain 1: 5700 -14720.934 0.221 0.219
Chain 1: 5800 -8760.244 0.279 0.236
Chain 1: 5900 -10458.576 0.259 0.219
Chain 1: 6000 -10217.493 0.222 0.210
Chain 1: 6100 -9213.124 0.223 0.210
Chain 1: 6200 -8664.776 0.208 0.162
Chain 1: 6300 -10336.109 0.211 0.162
Chain 1: 6400 -12026.594 0.201 0.162
Chain 1: 6500 -9783.014 0.203 0.162
Chain 1: 6600 -9301.737 0.172 0.141
Chain 1: 6700 -9550.804 0.165 0.141
Chain 1: 6800 -10604.348 0.107 0.109
Chain 1: 6900 -11483.043 0.098 0.099
Chain 1: 7000 -12219.650 0.102 0.099
Chain 1: 7100 -8900.328 0.128 0.099
Chain 1: 7200 -8516.454 0.126 0.099
Chain 1: 7300 -9327.485 0.119 0.087
Chain 1: 7400 -8523.503 0.114 0.087
Chain 1: 7500 -10473.191 0.110 0.087
Chain 1: 7600 -8487.770 0.128 0.094
Chain 1: 7700 -8592.674 0.127 0.094
Chain 1: 7800 -9364.094 0.125 0.087
Chain 1: 7900 -11596.925 0.137 0.094
Chain 1: 8000 -9893.592 0.148 0.172
Chain 1: 8100 -8647.764 0.125 0.144
Chain 1: 8200 -8939.720 0.124 0.144
Chain 1: 8300 -11732.616 0.139 0.172
Chain 1: 8400 -8478.419 0.168 0.186
Chain 1: 8500 -9468.273 0.160 0.172
Chain 1: 8600 -8679.913 0.145 0.144
Chain 1: 8700 -8562.717 0.145 0.144
Chain 1: 8800 -8401.689 0.139 0.144
Chain 1: 8900 -8663.556 0.123 0.105
Chain 1: 9000 -10267.817 0.121 0.105
Chain 1: 9100 -8102.169 0.134 0.105
Chain 1: 9200 -11417.267 0.159 0.156
Chain 1: 9300 -10291.236 0.147 0.109
Chain 1: 9400 -8276.954 0.133 0.109
Chain 1: 9500 -10525.494 0.143 0.156
Chain 1: 9600 -8572.674 0.157 0.214
Chain 1: 9700 -9653.511 0.167 0.214
Chain 1: 9800 -11961.690 0.184 0.214
Chain 1: 9900 -9540.008 0.207 0.228
Chain 1: 10000 -11160.758 0.206 0.228
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001811 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 18.11 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58116.557 1.000 1.000
Chain 1: 200 -17638.084 1.647 2.295
Chain 1: 300 -8656.613 1.444 1.038
Chain 1: 400 -8133.751 1.099 1.038
Chain 1: 500 -8469.759 0.887 1.000
Chain 1: 600 -8653.647 0.743 1.000
Chain 1: 700 -7771.562 0.653 0.114
Chain 1: 800 -8120.220 0.577 0.114
Chain 1: 900 -7905.905 0.516 0.064
Chain 1: 1000 -8031.576 0.466 0.064
Chain 1: 1100 -7633.655 0.371 0.052
Chain 1: 1200 -7547.003 0.143 0.043
Chain 1: 1300 -7620.798 0.040 0.040
Chain 1: 1400 -7616.349 0.033 0.027
Chain 1: 1500 -7579.923 0.030 0.021
Chain 1: 1600 -7633.265 0.028 0.016
Chain 1: 1700 -7524.302 0.019 0.014
Chain 1: 1800 -7586.258 0.015 0.011
Chain 1: 1900 -7588.762 0.012 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002574 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.74 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86230.003 1.000 1.000
Chain 1: 200 -13484.881 3.197 5.395
Chain 1: 300 -9911.775 2.252 1.000
Chain 1: 400 -10880.786 1.711 1.000
Chain 1: 500 -8863.651 1.414 0.360
Chain 1: 600 -8732.023 1.181 0.360
Chain 1: 700 -8402.673 1.018 0.228
Chain 1: 800 -8796.872 0.896 0.228
Chain 1: 900 -8793.519 0.797 0.089
Chain 1: 1000 -8473.032 0.721 0.089
Chain 1: 1100 -8776.992 0.624 0.045 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8428.442 0.089 0.041
Chain 1: 1300 -8621.439 0.055 0.039
Chain 1: 1400 -8620.547 0.046 0.038
Chain 1: 1500 -8519.299 0.025 0.035
Chain 1: 1600 -8620.163 0.024 0.035
Chain 1: 1700 -8708.538 0.022 0.022
Chain 1: 1800 -8310.531 0.022 0.022
Chain 1: 1900 -8411.698 0.023 0.022
Chain 1: 2000 -8382.459 0.020 0.012
Chain 1: 2100 -8503.812 0.018 0.012
Chain 1: 2200 -8282.641 0.016 0.012
Chain 1: 2300 -8440.513 0.016 0.012
Chain 1: 2400 -8453.330 0.016 0.012
Chain 1: 2500 -8424.234 0.015 0.012
Chain 1: 2600 -8426.895 0.014 0.012
Chain 1: 2700 -8333.082 0.014 0.012
Chain 1: 2800 -8303.592 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003413 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.13 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8399760.364 1.000 1.000
Chain 1: 200 -1584384.950 2.651 4.302
Chain 1: 300 -889912.558 2.027 1.000
Chain 1: 400 -457089.280 1.757 1.000
Chain 1: 500 -357651.662 1.461 0.947
Chain 1: 600 -232552.925 1.307 0.947
Chain 1: 700 -119019.005 1.257 0.947
Chain 1: 800 -86278.325 1.147 0.947
Chain 1: 900 -66656.723 1.053 0.780
Chain 1: 1000 -51480.962 0.977 0.780
Chain 1: 1100 -38983.884 0.909 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38160.902 0.481 0.379
Chain 1: 1300 -26148.446 0.449 0.379
Chain 1: 1400 -25868.117 0.355 0.321
Chain 1: 1500 -22463.867 0.342 0.321
Chain 1: 1600 -21682.488 0.292 0.295
Chain 1: 1700 -20560.381 0.202 0.294
Chain 1: 1800 -20505.346 0.165 0.152
Chain 1: 1900 -20831.206 0.137 0.055
Chain 1: 2000 -19345.218 0.115 0.055
Chain 1: 2100 -19583.347 0.084 0.036
Chain 1: 2200 -19809.255 0.083 0.036
Chain 1: 2300 -19427.033 0.039 0.020
Chain 1: 2400 -19199.304 0.039 0.020
Chain 1: 2500 -19001.191 0.025 0.016
Chain 1: 2600 -18631.844 0.024 0.016
Chain 1: 2700 -18589.000 0.018 0.012
Chain 1: 2800 -18305.947 0.020 0.015
Chain 1: 2900 -18587.008 0.020 0.015
Chain 1: 3000 -18573.217 0.012 0.012
Chain 1: 3100 -18658.151 0.011 0.012
Chain 1: 3200 -18349.116 0.012 0.015
Chain 1: 3300 -18553.641 0.011 0.012
Chain 1: 3400 -18028.989 0.013 0.015
Chain 1: 3500 -18640.184 0.015 0.015
Chain 1: 3600 -17947.775 0.017 0.015
Chain 1: 3700 -18333.866 0.019 0.017
Chain 1: 3800 -17294.949 0.023 0.021
Chain 1: 3900 -17291.140 0.021 0.021
Chain 1: 4000 -17408.440 0.022 0.021
Chain 1: 4100 -17322.248 0.022 0.021
Chain 1: 4200 -17138.840 0.022 0.021
Chain 1: 4300 -17277.016 0.021 0.021
Chain 1: 4400 -17234.089 0.019 0.011
Chain 1: 4500 -17136.665 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001522 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.22 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48874.189 1.000 1.000
Chain 1: 200 -16707.660 1.463 1.925
Chain 1: 300 -14851.874 1.017 1.000
Chain 1: 400 -12543.229 0.809 1.000
Chain 1: 500 -12886.247 0.652 0.184
Chain 1: 600 -27486.298 0.632 0.531
Chain 1: 700 -15356.628 0.655 0.531
Chain 1: 800 -13496.002 0.590 0.531
Chain 1: 900 -16185.630 0.543 0.184
Chain 1: 1000 -30596.060 0.536 0.471
Chain 1: 1100 -10839.512 0.618 0.471 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -21733.962 0.476 0.471
Chain 1: 1300 -10190.235 0.576 0.501 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1400 -10639.596 0.562 0.501 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1500 -19826.548 0.606 0.501 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1600 -12645.869 0.610 0.501 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1700 -11297.409 0.542 0.471 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1800 -10034.995 0.541 0.471 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1900 -11045.928 0.534 0.471 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2000 -14620.986 0.511 0.463 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2100 -10165.132 0.373 0.438
Chain 1: 2200 -9935.348 0.325 0.245
Chain 1: 2300 -12241.492 0.230 0.188
Chain 1: 2400 -9310.038 0.258 0.245
Chain 1: 2500 -9888.079 0.217 0.188
Chain 1: 2600 -10814.418 0.169 0.126
Chain 1: 2700 -9102.174 0.176 0.188
Chain 1: 2800 -9676.828 0.169 0.188
Chain 1: 2900 -9288.538 0.164 0.188
Chain 1: 3000 -8986.650 0.143 0.086
Chain 1: 3100 -8623.512 0.104 0.059
Chain 1: 3200 -9219.861 0.108 0.065
Chain 1: 3300 -17333.252 0.136 0.065
Chain 1: 3400 -13628.309 0.131 0.065
Chain 1: 3500 -9970.551 0.162 0.086
Chain 1: 3600 -16007.034 0.191 0.188
Chain 1: 3700 -9392.998 0.243 0.272
Chain 1: 3800 -15479.997 0.276 0.367
Chain 1: 3900 -8807.787 0.348 0.377
Chain 1: 4000 -11004.937 0.365 0.377
Chain 1: 4100 -9014.261 0.382 0.377
Chain 1: 4200 -13419.274 0.409 0.377
Chain 1: 4300 -10188.326 0.394 0.367
Chain 1: 4400 -9570.917 0.373 0.367
Chain 1: 4500 -11891.534 0.356 0.328
Chain 1: 4600 -8701.826 0.355 0.328
Chain 1: 4700 -8652.186 0.285 0.317
Chain 1: 4800 -8777.328 0.247 0.221
Chain 1: 4900 -8752.155 0.171 0.200
Chain 1: 5000 -9897.300 0.163 0.195
Chain 1: 5100 -16162.285 0.180 0.195
Chain 1: 5200 -8807.001 0.230 0.195
Chain 1: 5300 -9622.440 0.207 0.116
Chain 1: 5400 -8935.598 0.208 0.116
Chain 1: 5500 -9729.848 0.197 0.085
Chain 1: 5600 -8819.677 0.171 0.085
Chain 1: 5700 -10816.038 0.189 0.103
Chain 1: 5800 -10762.471 0.188 0.103
Chain 1: 5900 -10727.105 0.188 0.103
Chain 1: 6000 -9399.483 0.190 0.103
Chain 1: 6100 -9209.921 0.154 0.085
Chain 1: 6200 -8696.582 0.076 0.082
Chain 1: 6300 -8383.130 0.071 0.077
Chain 1: 6400 -13298.031 0.101 0.082
Chain 1: 6500 -8595.410 0.147 0.103
Chain 1: 6600 -8847.123 0.140 0.059
Chain 1: 6700 -12488.087 0.150 0.059
Chain 1: 6800 -11982.953 0.154 0.059
Chain 1: 6900 -8856.868 0.189 0.141
Chain 1: 7000 -8827.598 0.175 0.059
Chain 1: 7100 -8481.086 0.177 0.059
Chain 1: 7200 -8562.938 0.172 0.042
Chain 1: 7300 -9372.423 0.177 0.086
Chain 1: 7400 -8356.810 0.152 0.086
Chain 1: 7500 -8881.467 0.104 0.059
Chain 1: 7600 -8562.951 0.104 0.059
Chain 1: 7700 -8179.678 0.080 0.047
Chain 1: 7800 -11292.374 0.103 0.059
Chain 1: 7900 -10674.077 0.074 0.058
Chain 1: 8000 -8357.171 0.101 0.059
Chain 1: 8100 -8703.971 0.101 0.059
Chain 1: 8200 -11028.243 0.121 0.086
Chain 1: 8300 -9915.327 0.124 0.112
Chain 1: 8400 -8576.638 0.127 0.112
Chain 1: 8500 -9096.638 0.127 0.112
Chain 1: 8600 -9613.614 0.129 0.112
Chain 1: 8700 -8142.967 0.142 0.156
Chain 1: 8800 -8344.039 0.117 0.112
Chain 1: 8900 -8384.955 0.112 0.112
Chain 1: 9000 -9908.223 0.099 0.112
Chain 1: 9100 -8246.042 0.115 0.154
Chain 1: 9200 -8740.369 0.100 0.112
Chain 1: 9300 -9728.562 0.099 0.102
Chain 1: 9400 -11721.214 0.100 0.102
Chain 1: 9500 -11634.915 0.095 0.102
Chain 1: 9600 -9015.665 0.119 0.154
Chain 1: 9700 -9820.009 0.109 0.102
Chain 1: 9800 -8959.123 0.116 0.102
Chain 1: 9900 -8462.511 0.122 0.102
Chain 1: 10000 -8319.579 0.108 0.096
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001488 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.88 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -46392.601 1.000 1.000
Chain 1: 200 -15616.097 1.485 1.971
Chain 1: 300 -8748.716 1.252 1.000
Chain 1: 400 -8671.491 0.941 1.000
Chain 1: 500 -8287.772 0.762 0.785
Chain 1: 600 -8536.023 0.640 0.785
Chain 1: 700 -8253.909 0.553 0.046
Chain 1: 800 -8117.112 0.486 0.046
Chain 1: 900 -8093.401 0.433 0.034
Chain 1: 1000 -7844.855 0.393 0.034
Chain 1: 1100 -7822.876 0.293 0.032
Chain 1: 1200 -7940.928 0.097 0.029
Chain 1: 1300 -7790.794 0.021 0.019
Chain 1: 1400 -7726.734 0.021 0.019
Chain 1: 1500 -7689.008 0.016 0.017
Chain 1: 1600 -7762.169 0.015 0.015
Chain 1: 1700 -7616.655 0.013 0.015
Chain 1: 1800 -7682.349 0.012 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002613 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 26.13 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85970.734 1.000 1.000
Chain 1: 200 -13501.836 3.184 5.367
Chain 1: 300 -9902.794 2.244 1.000
Chain 1: 400 -10627.056 1.700 1.000
Chain 1: 500 -8856.946 1.400 0.363
Chain 1: 600 -8388.378 1.176 0.363
Chain 1: 700 -8420.708 1.008 0.200
Chain 1: 800 -9072.812 0.891 0.200
Chain 1: 900 -8680.554 0.797 0.072
Chain 1: 1000 -8530.490 0.719 0.072
Chain 1: 1100 -8772.428 0.622 0.068 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8308.349 0.091 0.056
Chain 1: 1300 -8607.102 0.058 0.056
Chain 1: 1400 -8612.788 0.051 0.045
Chain 1: 1500 -8492.616 0.033 0.035
Chain 1: 1600 -8598.916 0.028 0.028
Chain 1: 1700 -8685.145 0.029 0.028
Chain 1: 1800 -8282.543 0.027 0.028
Chain 1: 1900 -8381.019 0.023 0.018
Chain 1: 2000 -8352.624 0.022 0.014
Chain 1: 2100 -8472.462 0.021 0.014
Chain 1: 2200 -8278.916 0.017 0.014
Chain 1: 2300 -8416.173 0.015 0.014
Chain 1: 2400 -8291.382 0.017 0.014
Chain 1: 2500 -8356.278 0.016 0.014
Chain 1: 2600 -8379.700 0.015 0.014
Chain 1: 2700 -8298.147 0.015 0.014
Chain 1: 2800 -8270.792 0.011 0.012
Chain 1: 2900 -8326.209 0.010 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002731 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 27.31 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8390450.948 1.000 1.000
Chain 1: 200 -1582627.386 2.651 4.302
Chain 1: 300 -891551.068 2.026 1.000
Chain 1: 400 -458040.114 1.756 1.000
Chain 1: 500 -358418.478 1.460 0.946
Chain 1: 600 -233499.353 1.306 0.946
Chain 1: 700 -119509.380 1.256 0.946
Chain 1: 800 -86612.977 1.146 0.946
Chain 1: 900 -66918.873 1.052 0.775
Chain 1: 1000 -51670.328 0.976 0.775
Chain 1: 1100 -39105.716 0.908 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38277.345 0.480 0.380
Chain 1: 1300 -26203.061 0.449 0.380
Chain 1: 1400 -25918.304 0.355 0.321
Chain 1: 1500 -22496.458 0.342 0.321
Chain 1: 1600 -21709.457 0.293 0.295
Chain 1: 1700 -20580.251 0.203 0.294
Chain 1: 1800 -20523.413 0.165 0.152
Chain 1: 1900 -20849.290 0.137 0.055
Chain 1: 2000 -19359.008 0.115 0.055
Chain 1: 2100 -19597.681 0.084 0.036
Chain 1: 2200 -19824.026 0.083 0.036
Chain 1: 2300 -19441.370 0.039 0.020
Chain 1: 2400 -19213.513 0.039 0.020
Chain 1: 2500 -19015.476 0.025 0.016
Chain 1: 2600 -18646.030 0.024 0.016
Chain 1: 2700 -18603.066 0.018 0.012
Chain 1: 2800 -18319.984 0.020 0.015
Chain 1: 2900 -18601.207 0.020 0.015
Chain 1: 3000 -18587.474 0.012 0.012
Chain 1: 3100 -18672.369 0.011 0.012
Chain 1: 3200 -18363.241 0.012 0.015
Chain 1: 3300 -18567.813 0.011 0.012
Chain 1: 3400 -18043.025 0.013 0.015
Chain 1: 3500 -18654.438 0.015 0.015
Chain 1: 3600 -17961.800 0.017 0.015
Chain 1: 3700 -18348.114 0.019 0.017
Chain 1: 3800 -17308.782 0.023 0.021
Chain 1: 3900 -17304.935 0.021 0.021
Chain 1: 4000 -17422.263 0.022 0.021
Chain 1: 4100 -17336.025 0.022 0.021
Chain 1: 4200 -17152.490 0.022 0.021
Chain 1: 4300 -17290.750 0.021 0.021
Chain 1: 4400 -17247.781 0.019 0.011
Chain 1: 4500 -17150.320 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001282 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.82 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13458.418 1.000 1.000
Chain 1: 200 -10110.648 0.666 1.000
Chain 1: 300 -8680.909 0.499 0.331
Chain 1: 400 -8282.380 0.386 0.331
Chain 1: 500 -8174.234 0.311 0.165
Chain 1: 600 -8128.836 0.260 0.165
Chain 1: 700 -8010.319 0.225 0.048
Chain 1: 800 -8053.876 0.198 0.048
Chain 1: 900 -8157.239 0.177 0.015
Chain 1: 1000 -8061.342 0.161 0.015
Chain 1: 1100 -8021.633 0.061 0.013
Chain 1: 1200 -8018.179 0.028 0.013
Chain 1: 1300 -8118.433 0.013 0.012
Chain 1: 1400 -7997.468 0.010 0.012 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003195 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.95 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -47487.294 1.000 1.000
Chain 1: 200 -16224.010 1.463 1.927
Chain 1: 300 -8776.437 1.259 1.000
Chain 1: 400 -8227.694 0.961 1.000
Chain 1: 500 -8877.496 0.783 0.849
Chain 1: 600 -8282.938 0.665 0.849
Chain 1: 700 -8304.237 0.570 0.073
Chain 1: 800 -8361.719 0.500 0.073
Chain 1: 900 -8204.688 0.446 0.072
Chain 1: 1000 -7922.722 0.405 0.072
Chain 1: 1100 -7745.875 0.307 0.067
Chain 1: 1200 -7768.378 0.115 0.036
Chain 1: 1300 -7848.537 0.031 0.023
Chain 1: 1400 -7702.093 0.026 0.019
Chain 1: 1500 -7588.086 0.021 0.019
Chain 1: 1600 -7782.511 0.016 0.019
Chain 1: 1700 -7563.061 0.019 0.019
Chain 1: 1800 -7725.918 0.020 0.021
Chain 1: 1900 -7641.085 0.019 0.021
Chain 1: 2000 -7664.515 0.016 0.019
Chain 1: 2100 -7667.667 0.014 0.015
Chain 1: 2200 -7777.103 0.015 0.015
Chain 1: 2300 -7576.716 0.016 0.019
Chain 1: 2400 -7601.122 0.015 0.015
Chain 1: 2500 -7739.676 0.015 0.018
Chain 1: 2600 -7569.065 0.015 0.018
Chain 1: 2700 -7601.112 0.012 0.014
Chain 1: 2800 -7588.157 0.010 0.011
Chain 1: 2900 -7469.214 0.011 0.014
Chain 1: 3000 -7604.501 0.012 0.016
Chain 1: 3100 -7576.095 0.013 0.016
Chain 1: 3200 -7778.212 0.014 0.018
Chain 1: 3300 -7495.684 0.015 0.018
Chain 1: 3400 -7729.078 0.018 0.018
Chain 1: 3500 -7481.298 0.019 0.023
Chain 1: 3600 -7547.475 0.018 0.018
Chain 1: 3700 -7497.329 0.018 0.018
Chain 1: 3800 -7496.274 0.018 0.018
Chain 1: 3900 -7459.129 0.017 0.018
Chain 1: 4000 -7450.811 0.015 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003039 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.39 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87788.719 1.000 1.000
Chain 1: 200 -13893.251 3.159 5.319
Chain 1: 300 -10125.078 2.230 1.000
Chain 1: 400 -11744.284 1.707 1.000
Chain 1: 500 -8786.595 1.433 0.372
Chain 1: 600 -8552.572 1.199 0.372
Chain 1: 700 -8601.159 1.028 0.337
Chain 1: 800 -9648.220 0.913 0.337
Chain 1: 900 -8898.445 0.821 0.138
Chain 1: 1000 -9046.763 0.741 0.138
Chain 1: 1100 -8825.596 0.643 0.109 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8477.252 0.115 0.084
Chain 1: 1300 -8772.299 0.082 0.041
Chain 1: 1400 -8679.556 0.069 0.034
Chain 1: 1500 -8625.407 0.036 0.027
Chain 1: 1600 -8736.700 0.034 0.025
Chain 1: 1700 -8790.622 0.034 0.025
Chain 1: 1800 -8345.829 0.029 0.025
Chain 1: 1900 -8451.069 0.022 0.016
Chain 1: 2000 -8432.944 0.020 0.013
Chain 1: 2100 -8570.687 0.019 0.013
Chain 1: 2200 -8345.453 0.018 0.013
Chain 1: 2300 -8443.060 0.016 0.012
Chain 1: 2400 -8517.932 0.016 0.012
Chain 1: 2500 -8457.698 0.016 0.012
Chain 1: 2600 -8474.580 0.015 0.012
Chain 1: 2700 -8380.757 0.015 0.012
Chain 1: 2800 -8326.291 0.010 0.011
Chain 1: 2900 -8431.797 0.010 0.011
Chain 1: 3000 -8270.637 0.012 0.012
Chain 1: 3100 -8410.822 0.012 0.012
Chain 1: 3200 -8279.991 0.011 0.012
Chain 1: 3300 -8509.687 0.013 0.013
Chain 1: 3400 -8517.048 0.012 0.013
Chain 1: 3500 -8385.087 0.013 0.016
Chain 1: 3600 -8237.319 0.014 0.016
Chain 1: 3700 -8384.272 0.015 0.017
Chain 1: 3800 -8239.993 0.016 0.018
Chain 1: 3900 -8171.796 0.016 0.018
Chain 1: 4000 -8282.014 0.015 0.017
Chain 1: 4100 -8247.004 0.014 0.016
Chain 1: 4200 -8232.843 0.012 0.016
Chain 1: 4300 -8266.286 0.010 0.013
Chain 1: 4400 -8223.186 0.011 0.013
Chain 1: 4500 -8321.189 0.010 0.012
Chain 1: 4600 -8212.803 0.010 0.012 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.0031 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8444436.133 1.000 1.000
Chain 1: 200 -1591610.979 2.653 4.306
Chain 1: 300 -891713.133 2.030 1.000
Chain 1: 400 -457696.601 1.760 1.000
Chain 1: 500 -357582.675 1.464 0.948
Chain 1: 600 -232447.523 1.310 0.948
Chain 1: 700 -119168.397 1.258 0.948
Chain 1: 800 -86463.325 1.148 0.948
Chain 1: 900 -66915.394 1.053 0.785
Chain 1: 1000 -51801.885 0.977 0.785
Chain 1: 1100 -39352.874 0.909 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38544.802 0.480 0.378
Chain 1: 1300 -26570.522 0.447 0.378
Chain 1: 1400 -26298.147 0.353 0.316
Chain 1: 1500 -22902.431 0.340 0.316
Chain 1: 1600 -22124.419 0.289 0.292
Chain 1: 1700 -21006.297 0.200 0.292
Chain 1: 1800 -20952.593 0.162 0.148
Chain 1: 1900 -21279.405 0.134 0.053
Chain 1: 2000 -19793.329 0.113 0.053
Chain 1: 2100 -20031.748 0.082 0.035
Chain 1: 2200 -20257.843 0.081 0.035
Chain 1: 2300 -19875.180 0.038 0.019
Chain 1: 2400 -19647.151 0.038 0.019
Chain 1: 2500 -19448.643 0.025 0.015
Chain 1: 2600 -19078.751 0.023 0.015
Chain 1: 2700 -19035.699 0.018 0.012
Chain 1: 2800 -18752.069 0.019 0.015
Chain 1: 2900 -19033.493 0.019 0.015
Chain 1: 3000 -19019.760 0.012 0.012
Chain 1: 3100 -19104.805 0.011 0.012
Chain 1: 3200 -18795.225 0.011 0.015
Chain 1: 3300 -19000.142 0.011 0.012
Chain 1: 3400 -18474.366 0.012 0.015
Chain 1: 3500 -19087.153 0.014 0.015
Chain 1: 3600 -18392.629 0.016 0.015
Chain 1: 3700 -18780.267 0.018 0.016
Chain 1: 3800 -17737.957 0.022 0.021
Chain 1: 3900 -17733.965 0.021 0.021
Chain 1: 4000 -17851.358 0.022 0.021
Chain 1: 4100 -17764.996 0.022 0.021
Chain 1: 4200 -17580.783 0.021 0.021
Chain 1: 4300 -17719.563 0.021 0.021
Chain 1: 4400 -17676.052 0.018 0.010
Chain 1: 4500 -17578.434 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001433 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.33 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12749.500 1.000 1.000
Chain 1: 200 -9700.973 0.657 1.000
Chain 1: 300 -8504.946 0.485 0.314
Chain 1: 400 -8698.310 0.369 0.314
Chain 1: 500 -8505.163 0.300 0.141
Chain 1: 600 -8428.156 0.251 0.141
Chain 1: 700 -8338.150 0.217 0.023
Chain 1: 800 -8344.428 0.190 0.023
Chain 1: 900 -8261.105 0.170 0.022
Chain 1: 1000 -8447.422 0.155 0.022
Chain 1: 1100 -8479.206 0.056 0.022
Chain 1: 1200 -8351.353 0.026 0.015
Chain 1: 1300 -8327.109 0.012 0.011
Chain 1: 1400 -8333.167 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001435 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.35 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -62206.801 1.000 1.000
Chain 1: 200 -18368.221 1.693 2.387
Chain 1: 300 -9095.950 1.469 1.019
Chain 1: 400 -8728.526 1.112 1.019
Chain 1: 500 -8773.059 0.891 1.000
Chain 1: 600 -9387.979 0.753 1.000
Chain 1: 700 -8269.620 0.665 0.135
Chain 1: 800 -8625.963 0.587 0.135
Chain 1: 900 -7675.773 0.535 0.124
Chain 1: 1000 -8343.000 0.490 0.124
Chain 1: 1100 -7943.251 0.395 0.080
Chain 1: 1200 -7790.870 0.158 0.066
Chain 1: 1300 -7779.396 0.056 0.050
Chain 1: 1400 -7668.471 0.054 0.050
Chain 1: 1500 -7579.129 0.054 0.050
Chain 1: 1600 -7879.052 0.052 0.041
Chain 1: 1700 -7649.172 0.041 0.038
Chain 1: 1800 -7694.189 0.038 0.030
Chain 1: 1900 -7603.184 0.026 0.020
Chain 1: 2000 -7597.942 0.018 0.014
Chain 1: 2100 -7552.211 0.014 0.012
Chain 1: 2200 -7833.719 0.016 0.012
Chain 1: 2300 -7636.253 0.018 0.014
Chain 1: 2400 -7657.303 0.017 0.012
Chain 1: 2500 -7644.137 0.016 0.012
Chain 1: 2600 -7564.386 0.013 0.011
Chain 1: 2700 -7560.850 0.010 0.006 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002567 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.67 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86137.793 1.000 1.000
Chain 1: 200 -13952.708 3.087 5.174
Chain 1: 300 -10326.696 2.175 1.000
Chain 1: 400 -11175.920 1.650 1.000
Chain 1: 500 -9300.273 1.360 0.351
Chain 1: 600 -8925.560 1.141 0.351
Chain 1: 700 -8883.074 0.978 0.202
Chain 1: 800 -9252.219 0.861 0.202
Chain 1: 900 -9119.797 0.767 0.076
Chain 1: 1000 -9038.704 0.691 0.076
Chain 1: 1100 -9192.810 0.593 0.042 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8809.543 0.080 0.042
Chain 1: 1300 -9019.806 0.047 0.040
Chain 1: 1400 -9033.178 0.040 0.023
Chain 1: 1500 -8885.762 0.021 0.017
Chain 1: 1600 -8998.769 0.018 0.017
Chain 1: 1700 -9082.317 0.019 0.017
Chain 1: 1800 -8670.012 0.019 0.017
Chain 1: 1900 -8765.947 0.019 0.017
Chain 1: 2000 -8739.289 0.018 0.017
Chain 1: 2100 -8861.666 0.018 0.014
Chain 1: 2200 -8681.768 0.016 0.014
Chain 1: 2300 -8761.014 0.014 0.013
Chain 1: 2400 -8830.715 0.015 0.013
Chain 1: 2500 -8776.107 0.014 0.011
Chain 1: 2600 -8775.448 0.013 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003133 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.33 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8382400.872 1.000 1.000
Chain 1: 200 -1581644.141 2.650 4.300
Chain 1: 300 -892097.513 2.024 1.000
Chain 1: 400 -458689.292 1.754 1.000
Chain 1: 500 -359313.658 1.459 0.945
Chain 1: 600 -234209.998 1.305 0.945
Chain 1: 700 -120080.089 1.254 0.945
Chain 1: 800 -87173.267 1.145 0.945
Chain 1: 900 -67446.113 1.050 0.773
Chain 1: 1000 -52186.209 0.974 0.773
Chain 1: 1100 -39606.334 0.906 0.534 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38777.571 0.478 0.377
Chain 1: 1300 -26679.751 0.446 0.377
Chain 1: 1400 -26393.370 0.353 0.318
Chain 1: 1500 -22966.256 0.340 0.318
Chain 1: 1600 -22178.255 0.290 0.292
Chain 1: 1700 -21045.753 0.200 0.292
Chain 1: 1800 -20988.463 0.163 0.149
Chain 1: 1900 -21314.527 0.135 0.054
Chain 1: 2000 -19822.565 0.114 0.054
Chain 1: 2100 -20061.043 0.083 0.036
Chain 1: 2200 -20287.948 0.082 0.036
Chain 1: 2300 -19904.826 0.039 0.019
Chain 1: 2400 -19676.902 0.039 0.019
Chain 1: 2500 -19479.073 0.025 0.015
Chain 1: 2600 -19109.083 0.023 0.015
Chain 1: 2700 -19066.064 0.018 0.012
Chain 1: 2800 -18782.943 0.019 0.015
Chain 1: 2900 -19064.337 0.019 0.015
Chain 1: 3000 -19050.472 0.012 0.012
Chain 1: 3100 -19135.406 0.011 0.012
Chain 1: 3200 -18826.089 0.011 0.015
Chain 1: 3300 -19030.858 0.011 0.012
Chain 1: 3400 -18505.768 0.012 0.015
Chain 1: 3500 -19117.673 0.014 0.015
Chain 1: 3600 -18424.478 0.016 0.015
Chain 1: 3700 -18811.185 0.018 0.016
Chain 1: 3800 -17771.009 0.022 0.021
Chain 1: 3900 -17767.217 0.021 0.021
Chain 1: 4000 -17884.488 0.022 0.021
Chain 1: 4100 -17798.201 0.022 0.021
Chain 1: 4200 -17614.541 0.021 0.021
Chain 1: 4300 -17752.851 0.021 0.021
Chain 1: 4400 -17709.706 0.018 0.010
Chain 1: 4500 -17612.303 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001292 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.92 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49089.025 1.000 1.000
Chain 1: 200 -22030.971 1.114 1.228
Chain 1: 300 -16803.435 0.846 1.000
Chain 1: 400 -33087.165 0.758 1.000
Chain 1: 500 -12141.309 0.951 1.000
Chain 1: 600 -24548.379 0.877 1.000
Chain 1: 700 -10675.861 0.937 1.000
Chain 1: 800 -10787.779 0.821 1.000
Chain 1: 900 -18989.411 0.778 0.505
Chain 1: 1000 -12358.298 0.754 0.537
Chain 1: 1100 -12320.198 0.654 0.505 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -14660.610 0.547 0.492 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1300 -13012.815 0.529 0.492 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1400 -12363.751 0.485 0.432
Chain 1: 1500 -11521.344 0.320 0.160
Chain 1: 1600 -10413.095 0.280 0.127
Chain 1: 1700 -14267.611 0.177 0.127
Chain 1: 1800 -10758.342 0.209 0.160
Chain 1: 1900 -9828.177 0.175 0.127
Chain 1: 2000 -10997.483 0.132 0.106
Chain 1: 2100 -9384.103 0.149 0.127
Chain 1: 2200 -11148.389 0.149 0.127
Chain 1: 2300 -14953.454 0.161 0.158
Chain 1: 2400 -9333.482 0.216 0.172
Chain 1: 2500 -9496.065 0.211 0.172
Chain 1: 2600 -10157.472 0.207 0.172
Chain 1: 2700 -12330.895 0.197 0.172
Chain 1: 2800 -14986.165 0.182 0.172
Chain 1: 2900 -9380.613 0.233 0.176
Chain 1: 3000 -9100.403 0.225 0.176
Chain 1: 3100 -9787.254 0.215 0.176
Chain 1: 3200 -10780.301 0.208 0.176
Chain 1: 3300 -9521.222 0.196 0.132
Chain 1: 3400 -10421.531 0.144 0.092
Chain 1: 3500 -9409.069 0.154 0.108
Chain 1: 3600 -9757.643 0.151 0.108
Chain 1: 3700 -9905.617 0.134 0.092
Chain 1: 3800 -16387.966 0.156 0.092
Chain 1: 3900 -11076.595 0.145 0.092
Chain 1: 4000 -10490.498 0.147 0.092
Chain 1: 4100 -8867.637 0.158 0.108
Chain 1: 4200 -9477.773 0.156 0.108
Chain 1: 4300 -9915.927 0.147 0.086
Chain 1: 4400 -9043.225 0.148 0.097
Chain 1: 4500 -9225.611 0.139 0.064
Chain 1: 4600 -11637.014 0.156 0.097
Chain 1: 4700 -10836.919 0.162 0.097
Chain 1: 4800 -8677.140 0.147 0.097
Chain 1: 4900 -8850.024 0.101 0.074
Chain 1: 5000 -10980.125 0.115 0.097
Chain 1: 5100 -17335.311 0.133 0.097
Chain 1: 5200 -9099.714 0.218 0.194
Chain 1: 5300 -12599.725 0.241 0.207
Chain 1: 5400 -8822.409 0.274 0.249
Chain 1: 5500 -13306.294 0.306 0.278
Chain 1: 5600 -10254.777 0.315 0.298
Chain 1: 5700 -9746.305 0.313 0.298
Chain 1: 5800 -16326.108 0.328 0.337
Chain 1: 5900 -11914.944 0.363 0.367
Chain 1: 6000 -9237.662 0.373 0.367
Chain 1: 6100 -9883.087 0.343 0.337
Chain 1: 6200 -8259.868 0.272 0.298
Chain 1: 6300 -9241.484 0.255 0.298
Chain 1: 6400 -8650.383 0.219 0.290
Chain 1: 6500 -9328.324 0.192 0.197
Chain 1: 6600 -8883.570 0.167 0.106
Chain 1: 6700 -8525.037 0.166 0.106
Chain 1: 6800 -9055.173 0.132 0.073
Chain 1: 6900 -12676.299 0.124 0.073
Chain 1: 7000 -9746.644 0.125 0.073
Chain 1: 7100 -8339.752 0.135 0.106
Chain 1: 7200 -9878.353 0.131 0.106
Chain 1: 7300 -8329.396 0.139 0.156
Chain 1: 7400 -9794.019 0.147 0.156
Chain 1: 7500 -11522.021 0.155 0.156
Chain 1: 7600 -11174.655 0.153 0.156
Chain 1: 7700 -9112.491 0.171 0.169
Chain 1: 7800 -13011.615 0.195 0.186
Chain 1: 7900 -8418.399 0.221 0.186
Chain 1: 8000 -12827.924 0.226 0.186
Chain 1: 8100 -8364.195 0.262 0.226
Chain 1: 8200 -9092.241 0.255 0.226
Chain 1: 8300 -9346.624 0.239 0.226
Chain 1: 8400 -10912.508 0.238 0.226
Chain 1: 8500 -8439.341 0.252 0.293
Chain 1: 8600 -11106.119 0.273 0.293
Chain 1: 8700 -8941.267 0.275 0.293
Chain 1: 8800 -10980.155 0.263 0.242
Chain 1: 8900 -11330.979 0.212 0.240
Chain 1: 9000 -9198.422 0.201 0.232
Chain 1: 9100 -10353.836 0.159 0.186
Chain 1: 9200 -11955.717 0.164 0.186
Chain 1: 9300 -10133.617 0.179 0.186
Chain 1: 9400 -8901.304 0.179 0.186
Chain 1: 9500 -8337.536 0.156 0.180
Chain 1: 9600 -10041.176 0.149 0.170
Chain 1: 9700 -8330.827 0.145 0.170
Chain 1: 9800 -8470.613 0.129 0.138
Chain 1: 9900 -10343.636 0.144 0.170
Chain 1: 10000 -10631.802 0.123 0.138
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001531 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.31 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57149.348 1.000 1.000
Chain 1: 200 -17597.793 1.624 2.248
Chain 1: 300 -8810.933 1.415 1.000
Chain 1: 400 -8349.598 1.075 1.000
Chain 1: 500 -8272.844 0.862 0.997
Chain 1: 600 -9152.744 0.734 0.997
Chain 1: 700 -7860.637 0.653 0.164
Chain 1: 800 -8272.924 0.577 0.164
Chain 1: 900 -7973.086 0.517 0.096
Chain 1: 1000 -7836.550 0.467 0.096
Chain 1: 1100 -7686.619 0.369 0.055
Chain 1: 1200 -7803.197 0.146 0.050
Chain 1: 1300 -7772.160 0.047 0.038
Chain 1: 1400 -7873.093 0.043 0.020
Chain 1: 1500 -7600.726 0.045 0.036
Chain 1: 1600 -7572.091 0.036 0.020
Chain 1: 1700 -7531.656 0.020 0.017
Chain 1: 1800 -7607.446 0.016 0.015
Chain 1: 1900 -7609.182 0.012 0.013
Chain 1: 2000 -7660.868 0.011 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003415 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.15 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86907.751 1.000 1.000
Chain 1: 200 -13618.242 3.191 5.382
Chain 1: 300 -9970.624 2.249 1.000
Chain 1: 400 -10908.511 1.708 1.000
Chain 1: 500 -8948.683 1.411 0.366
Chain 1: 600 -8425.944 1.186 0.366
Chain 1: 700 -8475.848 1.017 0.219
Chain 1: 800 -8688.615 0.893 0.219
Chain 1: 900 -8804.200 0.795 0.086
Chain 1: 1000 -8568.928 0.719 0.086
Chain 1: 1100 -8827.720 0.621 0.062 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8409.265 0.088 0.050
Chain 1: 1300 -8682.096 0.055 0.031
Chain 1: 1400 -8658.749 0.047 0.029
Chain 1: 1500 -8542.159 0.026 0.027
Chain 1: 1600 -8650.735 0.021 0.024
Chain 1: 1700 -8737.063 0.021 0.024
Chain 1: 1800 -8322.751 0.024 0.027
Chain 1: 1900 -8419.355 0.024 0.027
Chain 1: 2000 -8392.706 0.021 0.014
Chain 1: 2100 -8515.511 0.020 0.014
Chain 1: 2200 -8335.484 0.017 0.014
Chain 1: 2300 -8414.145 0.015 0.013
Chain 1: 2400 -8483.884 0.015 0.013
Chain 1: 2500 -8429.394 0.015 0.011
Chain 1: 2600 -8428.949 0.013 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002996 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.96 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8417033.184 1.000 1.000
Chain 1: 200 -1586061.682 2.653 4.307
Chain 1: 300 -891762.960 2.028 1.000
Chain 1: 400 -458032.852 1.758 1.000
Chain 1: 500 -358201.437 1.462 0.947
Chain 1: 600 -233086.689 1.308 0.947
Chain 1: 700 -119330.896 1.257 0.947
Chain 1: 800 -86514.851 1.148 0.947
Chain 1: 900 -66870.215 1.053 0.779
Chain 1: 1000 -51673.415 0.977 0.779
Chain 1: 1100 -39155.366 0.909 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38333.164 0.480 0.379
Chain 1: 1300 -26299.953 0.448 0.379
Chain 1: 1400 -26018.933 0.355 0.320
Chain 1: 1500 -22608.590 0.342 0.320
Chain 1: 1600 -21825.371 0.292 0.294
Chain 1: 1700 -20700.926 0.202 0.294
Chain 1: 1800 -20645.359 0.164 0.151
Chain 1: 1900 -20971.496 0.136 0.054
Chain 1: 2000 -19483.475 0.115 0.054
Chain 1: 2100 -19721.815 0.084 0.036
Chain 1: 2200 -19948.026 0.083 0.036
Chain 1: 2300 -19565.464 0.039 0.020
Chain 1: 2400 -19337.637 0.039 0.020
Chain 1: 2500 -19139.425 0.025 0.016
Chain 1: 2600 -18769.791 0.023 0.016
Chain 1: 2700 -18726.841 0.018 0.012
Chain 1: 2800 -18443.626 0.019 0.015
Chain 1: 2900 -18724.867 0.019 0.015
Chain 1: 3000 -18711.111 0.012 0.012
Chain 1: 3100 -18796.055 0.011 0.012
Chain 1: 3200 -18486.785 0.012 0.015
Chain 1: 3300 -18691.481 0.011 0.012
Chain 1: 3400 -18166.393 0.012 0.015
Chain 1: 3500 -18778.182 0.015 0.015
Chain 1: 3600 -18085.053 0.017 0.015
Chain 1: 3700 -18471.695 0.018 0.017
Chain 1: 3800 -17431.551 0.023 0.021
Chain 1: 3900 -17427.695 0.021 0.021
Chain 1: 4000 -17545.028 0.022 0.021
Chain 1: 4100 -17458.755 0.022 0.021
Chain 1: 4200 -17275.076 0.021 0.021
Chain 1: 4300 -17413.448 0.021 0.021
Chain 1: 4400 -17370.323 0.018 0.011
Chain 1: 4500 -17272.856 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00117 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 11.7 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13129.335 1.000 1.000
Chain 1: 200 -10017.203 0.655 1.000
Chain 1: 300 -8651.114 0.490 0.311
Chain 1: 400 -8852.466 0.373 0.311
Chain 1: 500 -8793.160 0.300 0.158
Chain 1: 600 -8568.188 0.254 0.158
Chain 1: 700 -8536.748 0.218 0.026
Chain 1: 800 -8511.863 0.191 0.026
Chain 1: 900 -8587.588 0.171 0.023
Chain 1: 1000 -8498.291 0.155 0.023
Chain 1: 1100 -8554.522 0.056 0.011
Chain 1: 1200 -8479.337 0.026 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001404 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.04 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -64293.135 1.000 1.000
Chain 1: 200 -19000.493 1.692 2.384
Chain 1: 300 -9210.493 1.482 1.063
Chain 1: 400 -9156.770 1.113 1.063
Chain 1: 500 -8640.939 0.902 1.000
Chain 1: 600 -8995.707 0.759 1.000
Chain 1: 700 -9228.824 0.654 0.060
Chain 1: 800 -8522.747 0.582 0.083
Chain 1: 900 -8481.626 0.518 0.060
Chain 1: 1000 -7644.430 0.477 0.083
Chain 1: 1100 -7816.081 0.380 0.060
Chain 1: 1200 -7616.061 0.144 0.039
Chain 1: 1300 -8026.922 0.043 0.039
Chain 1: 1400 -8149.521 0.044 0.039
Chain 1: 1500 -7598.217 0.045 0.039
Chain 1: 1600 -7703.484 0.042 0.026
Chain 1: 1700 -7688.339 0.040 0.026
Chain 1: 1800 -7618.944 0.033 0.022
Chain 1: 1900 -7608.995 0.032 0.022
Chain 1: 2000 -7716.480 0.023 0.015
Chain 1: 2100 -7538.799 0.023 0.015
Chain 1: 2200 -7809.070 0.024 0.015
Chain 1: 2300 -7693.663 0.020 0.015
Chain 1: 2400 -7673.465 0.019 0.014
Chain 1: 2500 -7630.978 0.012 0.014
Chain 1: 2600 -7594.365 0.011 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003187 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.87 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87262.199 1.000 1.000
Chain 1: 200 -14326.553 3.045 5.091
Chain 1: 300 -10596.386 2.148 1.000
Chain 1: 400 -12095.634 1.642 1.000
Chain 1: 500 -9459.848 1.369 0.352
Chain 1: 600 -9865.645 1.148 0.352
Chain 1: 700 -9180.474 0.994 0.279
Chain 1: 800 -9301.086 0.872 0.279
Chain 1: 900 -9701.353 0.780 0.124
Chain 1: 1000 -9183.752 0.707 0.124
Chain 1: 1100 -9245.204 0.608 0.075 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8911.433 0.103 0.056
Chain 1: 1300 -9229.021 0.071 0.041
Chain 1: 1400 -9011.087 0.061 0.041
Chain 1: 1500 -9082.614 0.034 0.037
Chain 1: 1600 -9187.504 0.031 0.034
Chain 1: 1700 -9244.981 0.024 0.024
Chain 1: 1800 -8797.741 0.028 0.034
Chain 1: 1900 -8906.113 0.025 0.024
Chain 1: 2000 -8888.080 0.019 0.012
Chain 1: 2100 -9019.081 0.020 0.015
Chain 1: 2200 -8802.351 0.019 0.015
Chain 1: 2300 -8905.908 0.017 0.012
Chain 1: 2400 -8966.656 0.015 0.012
Chain 1: 2500 -8914.732 0.015 0.012
Chain 1: 2600 -8929.415 0.014 0.012
Chain 1: 2700 -8836.301 0.014 0.012
Chain 1: 2800 -8782.174 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003208 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.08 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8420015.330 1.000 1.000
Chain 1: 200 -1582169.273 2.661 4.322
Chain 1: 300 -889835.694 2.033 1.000
Chain 1: 400 -458305.199 1.760 1.000
Chain 1: 500 -358765.392 1.464 0.942
Chain 1: 600 -233690.229 1.309 0.942
Chain 1: 700 -119957.080 1.257 0.942
Chain 1: 800 -87256.222 1.147 0.942
Chain 1: 900 -67605.660 1.052 0.778
Chain 1: 1000 -52422.080 0.976 0.778
Chain 1: 1100 -39920.538 0.907 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39100.904 0.477 0.375
Chain 1: 1300 -27058.526 0.444 0.375
Chain 1: 1400 -26780.683 0.351 0.313
Chain 1: 1500 -23369.862 0.337 0.313
Chain 1: 1600 -22588.604 0.287 0.291
Chain 1: 1700 -21461.119 0.198 0.290
Chain 1: 1800 -21405.660 0.161 0.146
Chain 1: 1900 -21732.423 0.133 0.053
Chain 1: 2000 -20242.471 0.111 0.053
Chain 1: 2100 -20480.542 0.081 0.035
Chain 1: 2200 -20707.820 0.080 0.035
Chain 1: 2300 -20324.193 0.038 0.019
Chain 1: 2400 -20096.089 0.038 0.019
Chain 1: 2500 -19898.369 0.024 0.015
Chain 1: 2600 -19527.624 0.023 0.015
Chain 1: 2700 -19484.337 0.018 0.012
Chain 1: 2800 -19201.120 0.019 0.015
Chain 1: 2900 -19482.600 0.019 0.014
Chain 1: 3000 -19468.595 0.011 0.012
Chain 1: 3100 -19553.759 0.011 0.011
Chain 1: 3200 -19243.953 0.011 0.014
Chain 1: 3300 -19449.067 0.010 0.011
Chain 1: 3400 -18923.287 0.012 0.014
Chain 1: 3500 -19536.280 0.014 0.015
Chain 1: 3600 -18841.416 0.016 0.015
Chain 1: 3700 -19229.413 0.018 0.016
Chain 1: 3800 -18186.841 0.022 0.020
Chain 1: 3900 -18182.967 0.021 0.020
Chain 1: 4000 -18300.220 0.021 0.020
Chain 1: 4100 -18213.955 0.021 0.020
Chain 1: 4200 -18029.659 0.021 0.020
Chain 1: 4300 -18168.386 0.020 0.020
Chain 1: 4400 -18124.775 0.018 0.010
Chain 1: 4500 -18027.276 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001348 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.48 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12280.188 1.000 1.000
Chain 1: 200 -9164.613 0.670 1.000
Chain 1: 300 -7893.223 0.500 0.340
Chain 1: 400 -8052.684 0.380 0.340
Chain 1: 500 -7953.304 0.307 0.161
Chain 1: 600 -7872.660 0.257 0.161
Chain 1: 700 -7780.399 0.222 0.020
Chain 1: 800 -7815.950 0.195 0.020
Chain 1: 900 -7943.895 0.175 0.016
Chain 1: 1000 -7841.416 0.159 0.016
Chain 1: 1100 -7920.127 0.060 0.013
Chain 1: 1200 -7787.853 0.028 0.013
Chain 1: 1300 -7795.161 0.012 0.012
Chain 1: 1400 -7773.444 0.010 0.012 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001391 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.91 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57012.902 1.000 1.000
Chain 1: 200 -17368.123 1.641 2.283
Chain 1: 300 -8675.193 1.428 1.002
Chain 1: 400 -8355.469 1.081 1.002
Chain 1: 500 -8131.780 0.870 1.000
Chain 1: 600 -8844.561 0.739 1.000
Chain 1: 700 -8000.982 0.648 0.105
Chain 1: 800 -8112.886 0.569 0.105
Chain 1: 900 -7819.893 0.510 0.081
Chain 1: 1000 -7922.475 0.460 0.081
Chain 1: 1100 -7751.487 0.362 0.038
Chain 1: 1200 -7783.430 0.134 0.037
Chain 1: 1300 -7684.545 0.036 0.028
Chain 1: 1400 -7622.128 0.032 0.022
Chain 1: 1500 -7578.226 0.030 0.014
Chain 1: 1600 -7741.700 0.024 0.014
Chain 1: 1700 -7500.317 0.017 0.014
Chain 1: 1800 -7554.134 0.016 0.013
Chain 1: 1900 -7571.247 0.013 0.013
Chain 1: 2000 -7593.278 0.012 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003412 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.12 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86345.877 1.000 1.000
Chain 1: 200 -13387.782 3.225 5.450
Chain 1: 300 -9756.477 2.274 1.000
Chain 1: 400 -10785.724 1.729 1.000
Chain 1: 500 -8699.604 1.431 0.372
Chain 1: 600 -8248.612 1.202 0.372
Chain 1: 700 -8437.236 1.033 0.240
Chain 1: 800 -8851.778 0.910 0.240
Chain 1: 900 -8589.459 0.812 0.095
Chain 1: 1000 -8432.296 0.733 0.095
Chain 1: 1100 -8608.130 0.635 0.055 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8242.598 0.095 0.047
Chain 1: 1300 -8455.110 0.060 0.044
Chain 1: 1400 -8467.865 0.050 0.031
Chain 1: 1500 -8322.936 0.028 0.025
Chain 1: 1600 -8435.857 0.024 0.022
Chain 1: 1700 -8519.224 0.023 0.020
Chain 1: 1800 -8107.366 0.023 0.020
Chain 1: 1900 -8203.356 0.021 0.019
Chain 1: 2000 -8176.457 0.020 0.017
Chain 1: 2100 -8298.829 0.019 0.015
Chain 1: 2200 -8118.524 0.017 0.015
Chain 1: 2300 -8198.492 0.015 0.013
Chain 1: 2400 -8268.049 0.016 0.013
Chain 1: 2500 -8213.339 0.015 0.012
Chain 1: 2600 -8212.598 0.014 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003279 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.79 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8419906.276 1.000 1.000
Chain 1: 200 -1591744.096 2.645 4.290
Chain 1: 300 -892302.154 2.025 1.000
Chain 1: 400 -457781.840 1.756 1.000
Chain 1: 500 -357527.330 1.461 0.949
Chain 1: 600 -232359.429 1.307 0.949
Chain 1: 700 -118799.745 1.257 0.949
Chain 1: 800 -86057.874 1.147 0.949
Chain 1: 900 -66465.153 1.053 0.784
Chain 1: 1000 -51324.032 0.977 0.784
Chain 1: 1100 -38850.612 0.909 0.539 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38035.709 0.482 0.380
Chain 1: 1300 -26044.847 0.450 0.380
Chain 1: 1400 -25769.458 0.356 0.321
Chain 1: 1500 -22369.713 0.343 0.321
Chain 1: 1600 -21590.002 0.293 0.295
Chain 1: 1700 -20470.187 0.203 0.295
Chain 1: 1800 -20415.868 0.165 0.152
Chain 1: 1900 -20741.968 0.137 0.055
Chain 1: 2000 -19256.307 0.115 0.055
Chain 1: 2100 -19494.699 0.084 0.036
Chain 1: 2200 -19720.484 0.083 0.036
Chain 1: 2300 -19338.262 0.039 0.020
Chain 1: 2400 -19110.425 0.039 0.020
Chain 1: 2500 -18912.142 0.025 0.016
Chain 1: 2600 -18542.752 0.024 0.016
Chain 1: 2700 -18499.810 0.018 0.012
Chain 1: 2800 -18216.523 0.020 0.016
Chain 1: 2900 -18497.709 0.020 0.015
Chain 1: 3000 -18484.008 0.012 0.012
Chain 1: 3100 -18568.968 0.011 0.012
Chain 1: 3200 -18259.753 0.012 0.015
Chain 1: 3300 -18464.397 0.011 0.012
Chain 1: 3400 -17939.380 0.013 0.015
Chain 1: 3500 -18551.051 0.015 0.016
Chain 1: 3600 -17857.997 0.017 0.016
Chain 1: 3700 -18244.545 0.019 0.017
Chain 1: 3800 -17204.599 0.023 0.021
Chain 1: 3900 -17200.692 0.022 0.021
Chain 1: 4000 -17318.060 0.022 0.021
Chain 1: 4100 -17231.795 0.022 0.021
Chain 1: 4200 -17048.121 0.022 0.021
Chain 1: 4300 -17186.501 0.021 0.021
Chain 1: 4400 -17143.398 0.019 0.011
Chain 1: 4500 -17045.892 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001506 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.06 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49093.771 1.000 1.000
Chain 1: 200 -15426.644 1.591 2.182
Chain 1: 300 -19548.322 1.131 1.000
Chain 1: 400 -21851.895 0.875 1.000
Chain 1: 500 -12367.900 0.853 0.767
Chain 1: 600 -22091.037 0.784 0.767
Chain 1: 700 -12213.852 0.788 0.767
Chain 1: 800 -18963.845 0.734 0.767
Chain 1: 900 -13886.872 0.693 0.440
Chain 1: 1000 -12189.603 0.638 0.440
Chain 1: 1100 -10528.651 0.553 0.366 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -10486.590 0.335 0.356
Chain 1: 1300 -13015.828 0.334 0.356
Chain 1: 1400 -10178.021 0.351 0.356
Chain 1: 1500 -10145.673 0.275 0.279
Chain 1: 1600 -11531.894 0.243 0.194
Chain 1: 1700 -10299.961 0.174 0.158
Chain 1: 1800 -18329.879 0.182 0.158
Chain 1: 1900 -9800.141 0.233 0.158
Chain 1: 2000 -11667.272 0.235 0.160
Chain 1: 2100 -9801.524 0.238 0.190
Chain 1: 2200 -10577.780 0.245 0.190
Chain 1: 2300 -15946.903 0.259 0.190
Chain 1: 2400 -9727.901 0.295 0.190
Chain 1: 2500 -10950.350 0.306 0.190
Chain 1: 2600 -9174.033 0.313 0.194
Chain 1: 2700 -10461.081 0.314 0.194
Chain 1: 2800 -14832.057 0.299 0.194
Chain 1: 2900 -10074.078 0.260 0.194
Chain 1: 3000 -9614.821 0.248 0.194
Chain 1: 3100 -8831.194 0.238 0.194
Chain 1: 3200 -9639.944 0.239 0.194
Chain 1: 3300 -9249.174 0.210 0.123
Chain 1: 3400 -9473.362 0.148 0.112
Chain 1: 3500 -9205.161 0.140 0.089
Chain 1: 3600 -9529.909 0.124 0.084
Chain 1: 3700 -10072.536 0.117 0.054
Chain 1: 3800 -8573.199 0.105 0.054
Chain 1: 3900 -12893.483 0.091 0.054
Chain 1: 4000 -9068.236 0.129 0.084
Chain 1: 4100 -9283.965 0.122 0.054
Chain 1: 4200 -9389.999 0.115 0.042
Chain 1: 4300 -10858.118 0.124 0.054
Chain 1: 4400 -15076.377 0.150 0.135
Chain 1: 4500 -12347.815 0.169 0.175
Chain 1: 4600 -9572.913 0.195 0.221
Chain 1: 4700 -10045.713 0.194 0.221
Chain 1: 4800 -8596.282 0.193 0.221
Chain 1: 4900 -8866.984 0.163 0.169
Chain 1: 5000 -13363.079 0.154 0.169
Chain 1: 5100 -8888.594 0.202 0.221
Chain 1: 5200 -8792.466 0.202 0.221
Chain 1: 5300 -13268.264 0.222 0.280
Chain 1: 5400 -11871.521 0.206 0.221
Chain 1: 5500 -12933.863 0.192 0.169
Chain 1: 5600 -8824.220 0.210 0.169
Chain 1: 5700 -11449.653 0.228 0.229
Chain 1: 5800 -8814.239 0.241 0.299
Chain 1: 5900 -8595.645 0.241 0.299
Chain 1: 6000 -8787.926 0.209 0.229
Chain 1: 6100 -10469.865 0.175 0.161
Chain 1: 6200 -9165.917 0.188 0.161
Chain 1: 6300 -8543.744 0.162 0.142
Chain 1: 6400 -11987.835 0.179 0.161
Chain 1: 6500 -8599.052 0.210 0.229
Chain 1: 6600 -8424.167 0.165 0.161
Chain 1: 6700 -9365.936 0.152 0.142
Chain 1: 6800 -10847.263 0.136 0.137
Chain 1: 6900 -9846.402 0.144 0.137
Chain 1: 7000 -8621.997 0.156 0.142
Chain 1: 7100 -12793.283 0.172 0.142
Chain 1: 7200 -9697.687 0.190 0.142
Chain 1: 7300 -11753.679 0.200 0.175
Chain 1: 7400 -11841.805 0.172 0.142
Chain 1: 7500 -10908.847 0.141 0.137
Chain 1: 7600 -9103.471 0.159 0.142
Chain 1: 7700 -8465.034 0.157 0.142
Chain 1: 7800 -9184.725 0.151 0.142
Chain 1: 7900 -8856.409 0.144 0.142
Chain 1: 8000 -9689.886 0.139 0.086
Chain 1: 8100 -9682.459 0.106 0.086
Chain 1: 8200 -11835.929 0.093 0.086
Chain 1: 8300 -8591.584 0.113 0.086
Chain 1: 8400 -10419.386 0.130 0.086
Chain 1: 8500 -8345.582 0.146 0.175
Chain 1: 8600 -9921.997 0.142 0.159
Chain 1: 8700 -8559.621 0.150 0.159
Chain 1: 8800 -8152.322 0.148 0.159
Chain 1: 8900 -9248.174 0.156 0.159
Chain 1: 9000 -10286.348 0.157 0.159
Chain 1: 9100 -8525.605 0.178 0.175
Chain 1: 9200 -9635.593 0.171 0.159
Chain 1: 9300 -8930.226 0.141 0.159
Chain 1: 9400 -8566.057 0.128 0.118
Chain 1: 9500 -8244.448 0.107 0.115
Chain 1: 9600 -9964.965 0.108 0.115
Chain 1: 9700 -8187.278 0.114 0.115
Chain 1: 9800 -12922.854 0.146 0.118
Chain 1: 9900 -8596.693 0.184 0.173
Chain 1: 10000 -8262.355 0.178 0.173
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001561 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.61 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61696.888 1.000 1.000
Chain 1: 200 -17908.151 1.723 2.445
Chain 1: 300 -8868.577 1.488 1.019
Chain 1: 400 -8328.848 1.132 1.019
Chain 1: 500 -8830.786 0.917 1.000
Chain 1: 600 -8826.879 0.764 1.000
Chain 1: 700 -8448.686 0.662 0.065
Chain 1: 800 -8113.182 0.584 0.065
Chain 1: 900 -7976.503 0.521 0.057
Chain 1: 1000 -7772.398 0.472 0.057
Chain 1: 1100 -7665.957 0.373 0.045
Chain 1: 1200 -7660.289 0.129 0.041
Chain 1: 1300 -7492.708 0.029 0.026
Chain 1: 1400 -7808.527 0.026 0.026
Chain 1: 1500 -7517.049 0.025 0.026
Chain 1: 1600 -7689.230 0.027 0.026
Chain 1: 1700 -7478.417 0.025 0.026
Chain 1: 1800 -7518.638 0.022 0.022
Chain 1: 1900 -7528.682 0.020 0.022
Chain 1: 2000 -7585.519 0.018 0.022
Chain 1: 2100 -7459.170 0.018 0.022
Chain 1: 2200 -7684.892 0.021 0.022
Chain 1: 2300 -7552.635 0.021 0.022
Chain 1: 2400 -7605.416 0.017 0.018
Chain 1: 2500 -7697.676 0.015 0.017
Chain 1: 2600 -7472.241 0.016 0.017
Chain 1: 2700 -7484.150 0.013 0.012
Chain 1: 2800 -7529.289 0.013 0.012
Chain 1: 2900 -7355.481 0.015 0.017
Chain 1: 3000 -7489.372 0.016 0.018
Chain 1: 3100 -7481.414 0.015 0.018
Chain 1: 3200 -7667.891 0.014 0.018
Chain 1: 3300 -7418.192 0.016 0.018
Chain 1: 3400 -7616.101 0.018 0.024
Chain 1: 3500 -7394.719 0.019 0.024
Chain 1: 3600 -7456.571 0.017 0.024
Chain 1: 3700 -7407.280 0.018 0.024
Chain 1: 3800 -7414.871 0.017 0.024
Chain 1: 3900 -7385.876 0.015 0.018
Chain 1: 4000 -7373.825 0.014 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003172 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.72 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86677.564 1.000 1.000
Chain 1: 200 -13542.069 3.200 5.401
Chain 1: 300 -9906.510 2.256 1.000
Chain 1: 400 -10878.204 1.714 1.000
Chain 1: 500 -8875.621 1.417 0.367
Chain 1: 600 -8611.401 1.186 0.367
Chain 1: 700 -8423.221 1.019 0.226
Chain 1: 800 -9245.791 0.903 0.226
Chain 1: 900 -8722.275 0.809 0.089
Chain 1: 1000 -8392.625 0.732 0.089
Chain 1: 1100 -8749.447 0.636 0.089 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8367.465 0.101 0.060
Chain 1: 1300 -8496.237 0.066 0.046
Chain 1: 1400 -8573.689 0.058 0.041
Chain 1: 1500 -8467.329 0.036 0.039
Chain 1: 1600 -8571.707 0.035 0.039
Chain 1: 1700 -8661.130 0.033 0.039
Chain 1: 1800 -8245.906 0.030 0.039
Chain 1: 1900 -8343.017 0.025 0.015
Chain 1: 2000 -8316.416 0.021 0.013
Chain 1: 2100 -8439.430 0.018 0.013
Chain 1: 2200 -8258.823 0.016 0.013
Chain 1: 2300 -8337.634 0.016 0.012
Chain 1: 2400 -8407.362 0.015 0.012
Chain 1: 2500 -8352.974 0.015 0.012
Chain 1: 2600 -8352.681 0.014 0.010
Chain 1: 2700 -8269.981 0.014 0.010
Chain 1: 2800 -8232.929 0.009 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003388 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.88 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8435311.541 1.000 1.000
Chain 1: 200 -1586493.565 2.658 4.317
Chain 1: 300 -890658.250 2.033 1.000
Chain 1: 400 -457669.436 1.761 1.000
Chain 1: 500 -357632.424 1.465 0.946
Chain 1: 600 -232652.546 1.310 0.946
Chain 1: 700 -119056.903 1.259 0.946
Chain 1: 800 -86333.460 1.149 0.946
Chain 1: 900 -66710.628 1.054 0.781
Chain 1: 1000 -51538.166 0.978 0.781
Chain 1: 1100 -39049.381 0.910 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38228.172 0.481 0.379
Chain 1: 1300 -26215.404 0.448 0.379
Chain 1: 1400 -25938.010 0.355 0.320
Chain 1: 1500 -22533.615 0.342 0.320
Chain 1: 1600 -21752.895 0.292 0.294
Chain 1: 1700 -20630.110 0.202 0.294
Chain 1: 1800 -20575.103 0.164 0.151
Chain 1: 1900 -20901.279 0.136 0.054
Chain 1: 2000 -19414.393 0.115 0.054
Chain 1: 2100 -19652.581 0.084 0.036
Chain 1: 2200 -19878.818 0.083 0.036
Chain 1: 2300 -19496.236 0.039 0.020
Chain 1: 2400 -19268.341 0.039 0.020
Chain 1: 2500 -19070.313 0.025 0.016
Chain 1: 2600 -18700.534 0.023 0.016
Chain 1: 2700 -18657.552 0.018 0.012
Chain 1: 2800 -18374.396 0.020 0.015
Chain 1: 2900 -18655.593 0.019 0.015
Chain 1: 3000 -18641.810 0.012 0.012
Chain 1: 3100 -18726.782 0.011 0.012
Chain 1: 3200 -18417.484 0.012 0.015
Chain 1: 3300 -18622.202 0.011 0.012
Chain 1: 3400 -18097.130 0.013 0.015
Chain 1: 3500 -18708.939 0.015 0.015
Chain 1: 3600 -18015.693 0.017 0.015
Chain 1: 3700 -18402.407 0.018 0.017
Chain 1: 3800 -17362.207 0.023 0.021
Chain 1: 3900 -17358.335 0.021 0.021
Chain 1: 4000 -17475.662 0.022 0.021
Chain 1: 4100 -17389.412 0.022 0.021
Chain 1: 4200 -17205.678 0.021 0.021
Chain 1: 4300 -17344.066 0.021 0.021
Chain 1: 4400 -17300.899 0.019 0.011
Chain 1: 4500 -17203.416 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001297 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.97 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13155.472 1.000 1.000
Chain 1: 200 -9624.350 0.683 1.000
Chain 1: 300 -8592.051 0.496 0.367
Chain 1: 400 -8406.154 0.377 0.367
Chain 1: 500 -8078.055 0.310 0.120
Chain 1: 600 -7897.887 0.262 0.120
Chain 1: 700 -7835.671 0.226 0.041
Chain 1: 800 -7850.524 0.198 0.041
Chain 1: 900 -7810.348 0.176 0.023
Chain 1: 1000 -7887.970 0.160 0.023
Chain 1: 1100 -7933.960 0.060 0.022
Chain 1: 1200 -7848.619 0.025 0.011
Chain 1: 1300 -7784.630 0.014 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001519 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.19 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -63150.859 1.000 1.000
Chain 1: 200 -18259.529 1.729 2.459
Chain 1: 300 -8780.718 1.513 1.080
Chain 1: 400 -8667.150 1.138 1.080
Chain 1: 500 -8599.725 0.912 1.000
Chain 1: 600 -8845.357 0.764 1.000
Chain 1: 700 -7694.528 0.677 0.150
Chain 1: 800 -8069.828 0.598 0.150
Chain 1: 900 -8027.659 0.532 0.047
Chain 1: 1000 -7958.473 0.480 0.047
Chain 1: 1100 -7715.447 0.383 0.031
Chain 1: 1200 -7559.635 0.139 0.028
Chain 1: 1300 -7678.485 0.033 0.021
Chain 1: 1400 -7889.794 0.034 0.027
Chain 1: 1500 -7534.222 0.038 0.028
Chain 1: 1600 -7621.879 0.036 0.027
Chain 1: 1700 -7496.589 0.023 0.021
Chain 1: 1800 -7506.594 0.019 0.017
Chain 1: 1900 -7532.183 0.018 0.017
Chain 1: 2000 -7562.707 0.018 0.017
Chain 1: 2100 -7515.945 0.015 0.015
Chain 1: 2200 -7645.722 0.015 0.015
Chain 1: 2300 -7498.986 0.015 0.017
Chain 1: 2400 -7533.074 0.013 0.012
Chain 1: 2500 -7524.377 0.009 0.006 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003023 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.23 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86933.238 1.000 1.000
Chain 1: 200 -13568.404 3.204 5.407
Chain 1: 300 -9878.573 2.260 1.000
Chain 1: 400 -10890.300 1.718 1.000
Chain 1: 500 -8869.998 1.420 0.374
Chain 1: 600 -8316.157 1.195 0.374
Chain 1: 700 -8313.820 1.024 0.228
Chain 1: 800 -8506.181 0.899 0.228
Chain 1: 900 -8682.330 0.801 0.093
Chain 1: 1000 -8463.776 0.724 0.093
Chain 1: 1100 -8627.525 0.626 0.067 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8286.835 0.089 0.041
Chain 1: 1300 -8543.010 0.055 0.030
Chain 1: 1400 -8552.499 0.045 0.026
Chain 1: 1500 -8401.920 0.024 0.023
Chain 1: 1600 -8517.381 0.019 0.020
Chain 1: 1700 -8585.776 0.020 0.020
Chain 1: 1800 -8154.101 0.023 0.020
Chain 1: 1900 -8258.139 0.022 0.019
Chain 1: 2000 -8233.486 0.020 0.018
Chain 1: 2100 -8369.027 0.020 0.016
Chain 1: 2200 -8163.424 0.018 0.016
Chain 1: 2300 -8304.192 0.017 0.016
Chain 1: 2400 -8160.823 0.018 0.017
Chain 1: 2500 -8232.272 0.017 0.016
Chain 1: 2600 -8146.105 0.017 0.016
Chain 1: 2700 -8177.989 0.017 0.016
Chain 1: 2800 -8140.996 0.012 0.013
Chain 1: 2900 -8231.266 0.012 0.011
Chain 1: 3000 -8056.156 0.014 0.016
Chain 1: 3100 -8222.249 0.014 0.017
Chain 1: 3200 -8095.553 0.013 0.016
Chain 1: 3300 -8114.568 0.012 0.011
Chain 1: 3400 -8242.827 0.011 0.011
Chain 1: 3500 -8238.280 0.011 0.011
Chain 1: 3600 -8053.874 0.012 0.016
Chain 1: 3700 -8194.167 0.013 0.016
Chain 1: 3800 -8060.910 0.014 0.017
Chain 1: 3900 -7996.502 0.014 0.017
Chain 1: 4000 -8071.591 0.013 0.016
Chain 1: 4100 -8061.025 0.011 0.016
Chain 1: 4200 -8049.515 0.010 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002928 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.28 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8443592.390 1.000 1.000
Chain 1: 200 -1589019.810 2.657 4.314
Chain 1: 300 -890752.170 2.033 1.000
Chain 1: 400 -457507.091 1.761 1.000
Chain 1: 500 -357172.443 1.465 0.947
Chain 1: 600 -232238.285 1.311 0.947
Chain 1: 700 -118845.337 1.260 0.947
Chain 1: 800 -86170.445 1.150 0.947
Chain 1: 900 -66599.906 1.055 0.784
Chain 1: 1000 -51473.774 0.978 0.784
Chain 1: 1100 -39021.806 0.910 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38210.033 0.481 0.379
Chain 1: 1300 -26232.054 0.448 0.379
Chain 1: 1400 -25959.539 0.355 0.319
Chain 1: 1500 -22563.666 0.342 0.319
Chain 1: 1600 -21786.234 0.291 0.294
Chain 1: 1700 -20667.320 0.201 0.294
Chain 1: 1800 -20613.550 0.164 0.151
Chain 1: 1900 -20939.962 0.136 0.054
Chain 1: 2000 -19454.729 0.114 0.054
Chain 1: 2100 -19692.888 0.084 0.036
Chain 1: 2200 -19918.864 0.083 0.036
Chain 1: 2300 -19536.447 0.039 0.020
Chain 1: 2400 -19308.512 0.039 0.020
Chain 1: 2500 -19110.275 0.025 0.016
Chain 1: 2600 -18740.369 0.023 0.016
Chain 1: 2700 -18697.422 0.018 0.012
Chain 1: 2800 -18414.002 0.019 0.015
Chain 1: 2900 -18695.331 0.019 0.015
Chain 1: 3000 -18681.538 0.012 0.012
Chain 1: 3100 -18766.525 0.011 0.012
Chain 1: 3200 -18457.078 0.012 0.015
Chain 1: 3300 -18661.956 0.011 0.012
Chain 1: 3400 -18136.526 0.012 0.015
Chain 1: 3500 -18748.762 0.015 0.015
Chain 1: 3600 -18054.978 0.017 0.015
Chain 1: 3700 -18442.030 0.018 0.017
Chain 1: 3800 -17400.914 0.023 0.021
Chain 1: 3900 -17397.003 0.021 0.021
Chain 1: 4000 -17514.358 0.022 0.021
Chain 1: 4100 -17428.007 0.022 0.021
Chain 1: 4200 -17244.129 0.021 0.021
Chain 1: 4300 -17382.645 0.021 0.021
Chain 1: 4400 -17339.315 0.018 0.011
Chain 1: 4500 -17241.797 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001482 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.82 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12151.191 1.000 1.000
Chain 1: 200 -9081.091 0.669 1.000
Chain 1: 300 -7939.485 0.494 0.338
Chain 1: 400 -8148.686 0.377 0.338
Chain 1: 500 -8001.779 0.305 0.144
Chain 1: 600 -7871.086 0.257 0.144
Chain 1: 700 -7805.103 0.222 0.026
Chain 1: 800 -7813.124 0.194 0.026
Chain 1: 900 -7815.050 0.172 0.018
Chain 1: 1000 -7866.569 0.156 0.018
Chain 1: 1100 -7929.437 0.057 0.017
Chain 1: 1200 -7816.795 0.024 0.014
Chain 1: 1300 -7817.816 0.010 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001384 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.84 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61807.664 1.000 1.000
Chain 1: 200 -17521.654 1.764 2.528
Chain 1: 300 -8654.920 1.517 1.024
Chain 1: 400 -8087.121 1.156 1.024
Chain 1: 500 -8112.105 0.925 1.000
Chain 1: 600 -8751.580 0.783 1.000
Chain 1: 700 -7725.563 0.690 0.133
Chain 1: 800 -7927.012 0.607 0.133
Chain 1: 900 -7778.586 0.542 0.073
Chain 1: 1000 -7830.301 0.488 0.073
Chain 1: 1100 -7515.955 0.392 0.070
Chain 1: 1200 -7500.487 0.140 0.042
Chain 1: 1300 -7569.162 0.038 0.025
Chain 1: 1400 -7794.587 0.034 0.025
Chain 1: 1500 -7487.930 0.038 0.029
Chain 1: 1600 -7412.352 0.032 0.025
Chain 1: 1700 -7408.449 0.018 0.019
Chain 1: 1800 -7472.969 0.017 0.010
Chain 1: 1900 -7464.961 0.015 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002945 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.45 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85711.177 1.000 1.000
Chain 1: 200 -13230.405 3.239 5.478
Chain 1: 300 -9687.941 2.281 1.000
Chain 1: 400 -10550.102 1.731 1.000
Chain 1: 500 -8595.491 1.431 0.366
Chain 1: 600 -8225.241 1.200 0.366
Chain 1: 700 -8329.549 1.030 0.227
Chain 1: 800 -8686.849 0.906 0.227
Chain 1: 900 -8504.467 0.808 0.082
Chain 1: 1000 -8267.092 0.730 0.082
Chain 1: 1100 -8544.248 0.633 0.045 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8190.244 0.090 0.043
Chain 1: 1300 -8280.565 0.054 0.041
Chain 1: 1400 -8343.248 0.047 0.032
Chain 1: 1500 -8284.943 0.025 0.029
Chain 1: 1600 -8284.449 0.020 0.021
Chain 1: 1700 -8214.369 0.020 0.021
Chain 1: 1800 -8099.403 0.017 0.014
Chain 1: 1900 -8217.041 0.017 0.014
Chain 1: 2000 -8177.153 0.014 0.011
Chain 1: 2100 -8307.766 0.013 0.011
Chain 1: 2200 -8098.416 0.011 0.011
Chain 1: 2300 -8239.737 0.012 0.014
Chain 1: 2400 -8253.609 0.011 0.014
Chain 1: 2500 -8221.155 0.011 0.014
Chain 1: 2600 -8219.343 0.011 0.014
Chain 1: 2700 -8127.962 0.011 0.014
Chain 1: 2800 -8104.112 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003015 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.15 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8368602.781 1.000 1.000
Chain 1: 200 -1578867.352 2.650 4.300
Chain 1: 300 -889583.674 2.025 1.000
Chain 1: 400 -457106.656 1.755 1.000
Chain 1: 500 -358114.298 1.460 0.946
Chain 1: 600 -233099.121 1.306 0.946
Chain 1: 700 -119142.188 1.256 0.946
Chain 1: 800 -86334.503 1.146 0.946
Chain 1: 900 -66626.029 1.052 0.775
Chain 1: 1000 -51384.017 0.976 0.775
Chain 1: 1100 -38826.874 0.909 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37992.477 0.481 0.380
Chain 1: 1300 -25917.920 0.450 0.380
Chain 1: 1400 -25631.001 0.356 0.323
Chain 1: 1500 -22211.849 0.344 0.323
Chain 1: 1600 -21425.474 0.294 0.297
Chain 1: 1700 -20296.013 0.204 0.296
Chain 1: 1800 -20239.080 0.166 0.154
Chain 1: 1900 -20564.727 0.138 0.056
Chain 1: 2000 -19075.252 0.117 0.056
Chain 1: 2100 -19313.448 0.085 0.037
Chain 1: 2200 -19540.036 0.084 0.037
Chain 1: 2300 -19157.262 0.040 0.020
Chain 1: 2400 -18929.525 0.040 0.020
Chain 1: 2500 -18731.797 0.026 0.016
Chain 1: 2600 -18362.325 0.024 0.016
Chain 1: 2700 -18319.317 0.019 0.012
Chain 1: 2800 -18036.621 0.020 0.016
Chain 1: 2900 -18317.597 0.020 0.015
Chain 1: 3000 -18303.733 0.012 0.012
Chain 1: 3100 -18388.712 0.011 0.012
Chain 1: 3200 -18079.667 0.012 0.015
Chain 1: 3300 -18284.130 0.011 0.012
Chain 1: 3400 -17759.743 0.013 0.015
Chain 1: 3500 -18370.728 0.015 0.016
Chain 1: 3600 -17678.512 0.017 0.016
Chain 1: 3700 -18064.591 0.019 0.017
Chain 1: 3800 -17026.125 0.023 0.021
Chain 1: 3900 -17022.355 0.022 0.021
Chain 1: 4000 -17139.594 0.022 0.021
Chain 1: 4100 -17053.547 0.022 0.021
Chain 1: 4200 -16870.139 0.022 0.021
Chain 1: 4300 -17008.251 0.022 0.021
Chain 1: 4400 -16965.384 0.019 0.011
Chain 1: 4500 -16868.025 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001244 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.44 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -50066.971 1.000 1.000
Chain 1: 200 -17434.888 1.436 1.872
Chain 1: 300 -20693.425 1.010 1.000
Chain 1: 400 -17265.278 0.807 1.000
Chain 1: 500 -16386.866 0.656 0.199
Chain 1: 600 -12217.948 0.604 0.341
Chain 1: 700 -15413.132 0.547 0.207
Chain 1: 800 -27471.724 0.534 0.341
Chain 1: 900 -11903.809 0.620 0.341
Chain 1: 1000 -25204.929 0.610 0.439
Chain 1: 1100 -13521.148 0.597 0.439 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -11910.639 0.423 0.341
Chain 1: 1300 -14028.948 0.423 0.341
Chain 1: 1400 -13428.131 0.407 0.341
Chain 1: 1500 -15736.279 0.416 0.341
Chain 1: 1600 -13104.283 0.402 0.207
Chain 1: 1700 -10855.176 0.402 0.207
Chain 1: 1800 -10905.395 0.359 0.201
Chain 1: 1900 -12397.982 0.240 0.151
Chain 1: 2000 -10686.059 0.203 0.151
Chain 1: 2100 -11397.728 0.123 0.147
Chain 1: 2200 -10391.599 0.119 0.147
Chain 1: 2300 -10359.753 0.105 0.120
Chain 1: 2400 -10338.202 0.100 0.120
Chain 1: 2500 -12108.057 0.100 0.120
Chain 1: 2600 -10792.282 0.092 0.120
Chain 1: 2700 -10155.370 0.078 0.097
Chain 1: 2800 -10902.429 0.084 0.097
Chain 1: 2900 -10497.415 0.076 0.069
Chain 1: 3000 -9841.679 0.067 0.067
Chain 1: 3100 -10452.883 0.066 0.067
Chain 1: 3200 -9867.554 0.063 0.063
Chain 1: 3300 -12939.575 0.086 0.067
Chain 1: 3400 -9902.351 0.117 0.069
Chain 1: 3500 -10630.262 0.109 0.068
Chain 1: 3600 -11449.734 0.104 0.068
Chain 1: 3700 -9594.157 0.117 0.069
Chain 1: 3800 -16732.135 0.153 0.072
Chain 1: 3900 -12128.329 0.187 0.193
Chain 1: 4000 -9956.821 0.202 0.218
Chain 1: 4100 -11936.920 0.213 0.218
Chain 1: 4200 -16417.485 0.234 0.237
Chain 1: 4300 -11898.728 0.248 0.273
Chain 1: 4400 -10351.985 0.233 0.218
Chain 1: 4500 -12109.698 0.240 0.218
Chain 1: 4600 -9689.403 0.258 0.250
Chain 1: 4700 -11768.645 0.256 0.250
Chain 1: 4800 -9670.022 0.235 0.218
Chain 1: 4900 -9317.539 0.201 0.217
Chain 1: 5000 -10276.078 0.189 0.177
Chain 1: 5100 -9794.024 0.177 0.177
Chain 1: 5200 -15094.473 0.185 0.177
Chain 1: 5300 -12864.007 0.164 0.173
Chain 1: 5400 -9296.639 0.188 0.177
Chain 1: 5500 -9300.082 0.173 0.177
Chain 1: 5600 -15098.195 0.187 0.177
Chain 1: 5700 -9551.038 0.227 0.217
Chain 1: 5800 -9727.104 0.207 0.173
Chain 1: 5900 -14047.081 0.234 0.308
Chain 1: 6000 -10270.531 0.262 0.351
Chain 1: 6100 -9255.434 0.268 0.351
Chain 1: 6200 -9101.571 0.234 0.308
Chain 1: 6300 -12016.330 0.241 0.308
Chain 1: 6400 -11514.960 0.207 0.243
Chain 1: 6500 -13506.279 0.222 0.243
Chain 1: 6600 -9241.283 0.230 0.243
Chain 1: 6700 -13422.246 0.203 0.243
Chain 1: 6800 -10701.205 0.226 0.254
Chain 1: 6900 -9355.638 0.210 0.243
Chain 1: 7000 -12316.752 0.197 0.240
Chain 1: 7100 -13243.076 0.193 0.240
Chain 1: 7200 -9321.031 0.234 0.243
Chain 1: 7300 -10767.353 0.223 0.240
Chain 1: 7400 -9027.178 0.238 0.240
Chain 1: 7500 -9332.221 0.226 0.240
Chain 1: 7600 -9427.737 0.181 0.193
Chain 1: 7700 -9395.302 0.150 0.144
Chain 1: 7800 -9070.540 0.128 0.134
Chain 1: 7900 -8961.541 0.115 0.070
Chain 1: 8000 -9363.486 0.095 0.043
Chain 1: 8100 -12774.149 0.115 0.043
Chain 1: 8200 -9303.668 0.110 0.043
Chain 1: 8300 -12016.184 0.120 0.043
Chain 1: 8400 -13368.698 0.110 0.043
Chain 1: 8500 -12058.587 0.118 0.101
Chain 1: 8600 -9419.616 0.145 0.109
Chain 1: 8700 -9574.107 0.146 0.109
Chain 1: 8800 -9107.176 0.148 0.109
Chain 1: 8900 -9603.409 0.152 0.109
Chain 1: 9000 -13236.167 0.175 0.226
Chain 1: 9100 -8900.089 0.197 0.226
Chain 1: 9200 -9196.120 0.163 0.109
Chain 1: 9300 -10231.187 0.150 0.101
Chain 1: 9400 -9276.204 0.151 0.103
Chain 1: 9500 -8920.072 0.144 0.101
Chain 1: 9600 -9105.075 0.118 0.052
Chain 1: 9700 -8983.113 0.117 0.052
Chain 1: 9800 -8784.912 0.115 0.052
Chain 1: 9900 -10634.641 0.127 0.101
Chain 1: 10000 -9671.492 0.109 0.100
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001428 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.28 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58160.794 1.000 1.000
Chain 1: 200 -18389.470 1.581 2.163
Chain 1: 300 -9224.216 1.385 1.000
Chain 1: 400 -8353.881 1.065 1.000
Chain 1: 500 -8803.367 0.862 0.994
Chain 1: 600 -8785.652 0.719 0.994
Chain 1: 700 -8020.374 0.630 0.104
Chain 1: 800 -7965.117 0.552 0.104
Chain 1: 900 -8057.402 0.492 0.095
Chain 1: 1000 -8333.700 0.446 0.095
Chain 1: 1100 -7790.575 0.353 0.070
Chain 1: 1200 -8180.011 0.142 0.051
Chain 1: 1300 -8258.862 0.043 0.048
Chain 1: 1400 -8023.021 0.036 0.033
Chain 1: 1500 -8284.392 0.034 0.032
Chain 1: 1600 -7942.706 0.038 0.033
Chain 1: 1700 -7856.526 0.029 0.032
Chain 1: 1800 -7712.278 0.031 0.032
Chain 1: 1900 -7708.729 0.029 0.032
Chain 1: 2000 -7811.854 0.027 0.029
Chain 1: 2100 -7563.515 0.024 0.029
Chain 1: 2200 -8138.909 0.026 0.029
Chain 1: 2300 -7726.619 0.030 0.032
Chain 1: 2400 -7858.362 0.029 0.032
Chain 1: 2500 -7654.008 0.029 0.027
Chain 1: 2600 -7623.272 0.025 0.019
Chain 1: 2700 -7508.502 0.025 0.019
Chain 1: 2800 -7728.410 0.026 0.027
Chain 1: 2900 -7455.538 0.030 0.028
Chain 1: 3000 -7611.748 0.031 0.028
Chain 1: 3100 -7600.712 0.027 0.027
Chain 1: 3200 -7877.234 0.024 0.027
Chain 1: 3300 -7541.573 0.023 0.027
Chain 1: 3400 -7800.385 0.025 0.028
Chain 1: 3500 -7531.605 0.025 0.033
Chain 1: 3600 -7593.216 0.026 0.033
Chain 1: 3700 -7532.001 0.025 0.033
Chain 1: 3800 -7507.150 0.023 0.033
Chain 1: 3900 -7492.167 0.019 0.021
Chain 1: 4000 -7470.567 0.017 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003019 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.19 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86185.211 1.000 1.000
Chain 1: 200 -14439.187 2.984 4.969
Chain 1: 300 -10705.586 2.106 1.000
Chain 1: 400 -12092.271 1.608 1.000
Chain 1: 500 -9482.504 1.341 0.349
Chain 1: 600 -9370.625 1.120 0.349
Chain 1: 700 -9064.578 0.965 0.275
Chain 1: 800 -9619.995 0.851 0.275
Chain 1: 900 -9352.379 0.760 0.115
Chain 1: 1000 -9450.543 0.685 0.115
Chain 1: 1100 -9455.162 0.585 0.058 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8971.519 0.094 0.054
Chain 1: 1300 -9324.436 0.062 0.038
Chain 1: 1400 -9079.492 0.054 0.034
Chain 1: 1500 -9159.968 0.027 0.029
Chain 1: 1600 -9265.506 0.027 0.029
Chain 1: 1700 -9323.377 0.024 0.027
Chain 1: 1800 -8877.378 0.023 0.027
Chain 1: 1900 -8979.981 0.022 0.011
Chain 1: 2000 -8981.671 0.021 0.011
Chain 1: 2100 -9083.415 0.022 0.011
Chain 1: 2200 -8874.292 0.019 0.011
Chain 1: 2300 -9058.814 0.017 0.011
Chain 1: 2400 -8878.670 0.016 0.011
Chain 1: 2500 -8953.786 0.016 0.011
Chain 1: 2600 -8864.936 0.016 0.011
Chain 1: 2700 -8897.477 0.016 0.011
Chain 1: 2800 -8849.339 0.011 0.011
Chain 1: 2900 -8960.982 0.012 0.011
Chain 1: 3000 -8896.926 0.012 0.011
Chain 1: 3100 -8841.327 0.012 0.010
Chain 1: 3200 -8814.541 0.010 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003942 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 39.42 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8400261.072 1.000 1.000
Chain 1: 200 -1581031.233 2.657 4.313
Chain 1: 300 -890499.801 2.030 1.000
Chain 1: 400 -458609.468 1.758 1.000
Chain 1: 500 -359275.613 1.461 0.942
Chain 1: 600 -234304.738 1.307 0.942
Chain 1: 700 -120355.222 1.255 0.942
Chain 1: 800 -87608.145 1.145 0.942
Chain 1: 900 -67905.740 1.050 0.775
Chain 1: 1000 -52681.833 0.974 0.775
Chain 1: 1100 -40137.149 0.905 0.533 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39315.119 0.476 0.374
Chain 1: 1300 -27222.849 0.443 0.374
Chain 1: 1400 -26941.974 0.350 0.313
Chain 1: 1500 -23517.816 0.337 0.313
Chain 1: 1600 -22732.940 0.287 0.290
Chain 1: 1700 -21599.182 0.197 0.289
Chain 1: 1800 -21542.454 0.160 0.146
Chain 1: 1900 -21869.311 0.133 0.052
Chain 1: 2000 -20375.963 0.111 0.052
Chain 1: 2100 -20614.237 0.081 0.035
Chain 1: 2200 -20842.074 0.080 0.035
Chain 1: 2300 -20457.949 0.038 0.019
Chain 1: 2400 -20229.705 0.038 0.019
Chain 1: 2500 -20032.263 0.024 0.015
Chain 1: 2600 -19661.132 0.022 0.015
Chain 1: 2700 -19617.795 0.017 0.012
Chain 1: 2800 -19334.565 0.019 0.015
Chain 1: 2900 -19616.229 0.019 0.014
Chain 1: 3000 -19602.176 0.011 0.012
Chain 1: 3100 -19687.318 0.011 0.011
Chain 1: 3200 -19377.407 0.011 0.014
Chain 1: 3300 -19582.639 0.010 0.011
Chain 1: 3400 -19056.696 0.012 0.014
Chain 1: 3500 -19669.998 0.014 0.015
Chain 1: 3600 -18974.833 0.016 0.015
Chain 1: 3700 -19363.037 0.018 0.016
Chain 1: 3800 -18320.011 0.022 0.020
Chain 1: 3900 -18316.176 0.020 0.020
Chain 1: 4000 -18433.399 0.021 0.020
Chain 1: 4100 -18347.056 0.021 0.020
Chain 1: 4200 -18162.724 0.020 0.020
Chain 1: 4300 -18301.451 0.020 0.020
Chain 1: 4400 -18257.740 0.018 0.010
Chain 1: 4500 -18160.280 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001276 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.76 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48430.794 1.000 1.000
Chain 1: 200 -15902.125 1.523 2.046
Chain 1: 300 -13061.335 1.088 1.000
Chain 1: 400 -12147.078 0.835 1.000
Chain 1: 500 -11619.136 0.677 0.217
Chain 1: 600 -12767.367 0.579 0.217
Chain 1: 700 -14092.498 0.510 0.094
Chain 1: 800 -16244.465 0.463 0.132
Chain 1: 900 -10799.841 0.467 0.132
Chain 1: 1000 -12245.518 0.432 0.132
Chain 1: 1100 -12232.745 0.332 0.118
Chain 1: 1200 -10158.720 0.148 0.118
Chain 1: 1300 -15607.957 0.161 0.118
Chain 1: 1400 -19149.060 0.172 0.132
Chain 1: 1500 -11404.989 0.236 0.185
Chain 1: 1600 -10535.670 0.235 0.185
Chain 1: 1700 -22038.302 0.278 0.204
Chain 1: 1800 -9673.540 0.392 0.349
Chain 1: 1900 -10081.821 0.346 0.204
Chain 1: 2000 -20441.872 0.385 0.349
Chain 1: 2100 -19152.289 0.391 0.349
Chain 1: 2200 -10865.673 0.447 0.507 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2300 -9167.750 0.431 0.507 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2400 -10100.550 0.422 0.507 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2500 -11102.892 0.363 0.185
Chain 1: 2600 -9133.367 0.376 0.216
Chain 1: 2700 -9001.259 0.325 0.185
Chain 1: 2800 -16287.325 0.242 0.185
Chain 1: 2900 -11719.511 0.277 0.216
Chain 1: 3000 -8897.733 0.258 0.216
Chain 1: 3100 -15367.812 0.294 0.317
Chain 1: 3200 -8981.762 0.288 0.317
Chain 1: 3300 -9161.588 0.272 0.317
Chain 1: 3400 -15464.784 0.303 0.390
Chain 1: 3500 -9198.984 0.362 0.408
Chain 1: 3600 -8887.116 0.344 0.408
Chain 1: 3700 -8567.681 0.347 0.408
Chain 1: 3800 -15364.056 0.346 0.408
Chain 1: 3900 -8700.620 0.384 0.421
Chain 1: 4000 -9731.757 0.363 0.421
Chain 1: 4100 -8861.110 0.330 0.408
Chain 1: 4200 -13732.466 0.295 0.355
Chain 1: 4300 -12408.752 0.303 0.355
Chain 1: 4400 -9596.987 0.292 0.293
Chain 1: 4500 -8825.811 0.233 0.107
Chain 1: 4600 -11929.644 0.255 0.260
Chain 1: 4700 -9039.179 0.283 0.293
Chain 1: 4800 -8300.909 0.248 0.260
Chain 1: 4900 -8536.858 0.174 0.107
Chain 1: 5000 -8417.149 0.165 0.107
Chain 1: 5100 -10768.483 0.177 0.218
Chain 1: 5200 -14537.286 0.168 0.218
Chain 1: 5300 -9141.074 0.216 0.259
Chain 1: 5400 -13019.835 0.216 0.259
Chain 1: 5500 -12076.276 0.215 0.259
Chain 1: 5600 -8407.223 0.233 0.259
Chain 1: 5700 -13992.146 0.241 0.259
Chain 1: 5800 -8410.320 0.299 0.298
Chain 1: 5900 -15267.186 0.341 0.399
Chain 1: 6000 -8339.226 0.422 0.436
Chain 1: 6100 -12973.362 0.436 0.436
Chain 1: 6200 -8620.072 0.461 0.449
Chain 1: 6300 -8278.444 0.406 0.436
Chain 1: 6400 -12313.593 0.409 0.436
Chain 1: 6500 -13037.754 0.407 0.436
Chain 1: 6600 -8154.718 0.423 0.449
Chain 1: 6700 -11414.055 0.411 0.449
Chain 1: 6800 -9110.669 0.370 0.357
Chain 1: 6900 -8607.846 0.331 0.328
Chain 1: 7000 -9767.858 0.260 0.286
Chain 1: 7100 -8223.975 0.243 0.253
Chain 1: 7200 -8290.187 0.193 0.188
Chain 1: 7300 -9781.123 0.205 0.188
Chain 1: 7400 -8064.595 0.193 0.188
Chain 1: 7500 -8253.425 0.190 0.188
Chain 1: 7600 -9166.283 0.140 0.152
Chain 1: 7700 -9733.541 0.117 0.119
Chain 1: 7800 -8169.221 0.111 0.119
Chain 1: 7900 -9561.166 0.120 0.146
Chain 1: 8000 -8024.351 0.127 0.152
Chain 1: 8100 -8286.726 0.111 0.146
Chain 1: 8200 -8171.035 0.112 0.146
Chain 1: 8300 -8058.871 0.098 0.100
Chain 1: 8400 -11977.535 0.110 0.100
Chain 1: 8500 -7946.598 0.158 0.146
Chain 1: 8600 -8162.293 0.151 0.146
Chain 1: 8700 -8591.619 0.150 0.146
Chain 1: 8800 -9045.598 0.136 0.050
Chain 1: 8900 -8999.373 0.122 0.050
Chain 1: 9000 -8581.272 0.107 0.049
Chain 1: 9100 -10451.709 0.122 0.050
Chain 1: 9200 -9410.435 0.132 0.050
Chain 1: 9300 -9760.338 0.134 0.050
Chain 1: 9400 -8174.778 0.121 0.050
Chain 1: 9500 -9627.307 0.085 0.050
Chain 1: 9600 -8182.481 0.100 0.111
Chain 1: 9700 -9454.492 0.109 0.135
Chain 1: 9800 -8214.634 0.119 0.151
Chain 1: 9900 -8260.682 0.119 0.151
Chain 1: 10000 -7957.663 0.118 0.151
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001412 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.12 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56946.045 1.000 1.000
Chain 1: 200 -17248.724 1.651 2.301
Chain 1: 300 -8538.071 1.441 1.020
Chain 1: 400 -8185.946 1.091 1.020
Chain 1: 500 -8443.033 0.879 1.000
Chain 1: 600 -8552.782 0.735 1.000
Chain 1: 700 -7614.063 0.647 0.123
Chain 1: 800 -7954.011 0.572 0.123
Chain 1: 900 -7764.544 0.511 0.043
Chain 1: 1000 -7425.992 0.464 0.046
Chain 1: 1100 -7536.971 0.366 0.043
Chain 1: 1200 -7469.600 0.137 0.043
Chain 1: 1300 -7425.274 0.035 0.030
Chain 1: 1400 -7488.977 0.032 0.024
Chain 1: 1500 -7435.748 0.029 0.015
Chain 1: 1600 -7469.619 0.029 0.015
Chain 1: 1700 -7398.658 0.017 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003447 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.47 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86219.630 1.000 1.000
Chain 1: 200 -13222.120 3.260 5.521
Chain 1: 300 -9614.760 2.299 1.000
Chain 1: 400 -10479.064 1.745 1.000
Chain 1: 500 -8556.545 1.441 0.375
Chain 1: 600 -8480.696 1.202 0.375
Chain 1: 700 -8357.693 1.032 0.225
Chain 1: 800 -8403.137 0.904 0.225
Chain 1: 900 -8424.816 0.804 0.082
Chain 1: 1000 -8287.952 0.725 0.082
Chain 1: 1100 -8494.892 0.628 0.024 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8124.893 0.080 0.024
Chain 1: 1300 -8323.368 0.045 0.024
Chain 1: 1400 -8329.922 0.037 0.017
Chain 1: 1500 -8188.244 0.016 0.017
Chain 1: 1600 -8300.913 0.016 0.017
Chain 1: 1700 -8386.542 0.016 0.017
Chain 1: 1800 -7978.654 0.021 0.017
Chain 1: 1900 -8074.611 0.022 0.017
Chain 1: 2000 -8047.022 0.020 0.017
Chain 1: 2100 -8168.178 0.019 0.015
Chain 1: 2200 -8281.838 0.016 0.014
Chain 1: 2300 -8072.084 0.016 0.014
Chain 1: 2400 -8139.932 0.017 0.014
Chain 1: 2500 -8086.511 0.016 0.014
Chain 1: 2600 -8083.561 0.015 0.012
Chain 1: 2700 -8001.100 0.015 0.012
Chain 1: 2800 -7965.689 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003133 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.33 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8407030.676 1.000 1.000
Chain 1: 200 -1584223.814 2.653 4.307
Chain 1: 300 -890446.543 2.029 1.000
Chain 1: 400 -457302.705 1.758 1.000
Chain 1: 500 -357579.160 1.462 0.947
Chain 1: 600 -232548.699 1.308 0.947
Chain 1: 700 -118867.607 1.258 0.947
Chain 1: 800 -86098.685 1.148 0.947
Chain 1: 900 -66459.332 1.054 0.779
Chain 1: 1000 -51267.661 0.978 0.779
Chain 1: 1100 -38758.477 0.910 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37933.502 0.482 0.381
Chain 1: 1300 -25908.011 0.450 0.381
Chain 1: 1400 -25627.658 0.356 0.323
Chain 1: 1500 -22219.985 0.344 0.323
Chain 1: 1600 -21437.760 0.294 0.296
Chain 1: 1700 -20313.895 0.204 0.296
Chain 1: 1800 -20258.426 0.166 0.153
Chain 1: 1900 -20584.449 0.138 0.055
Chain 1: 2000 -19097.082 0.116 0.055
Chain 1: 2100 -19335.360 0.085 0.036
Chain 1: 2200 -19561.571 0.084 0.036
Chain 1: 2300 -19179.028 0.040 0.020
Chain 1: 2400 -18951.174 0.040 0.020
Chain 1: 2500 -18753.112 0.025 0.016
Chain 1: 2600 -18383.541 0.024 0.016
Chain 1: 2700 -18340.559 0.019 0.012
Chain 1: 2800 -18057.474 0.020 0.016
Chain 1: 2900 -18338.602 0.020 0.015
Chain 1: 3000 -18324.830 0.012 0.012
Chain 1: 3100 -18409.798 0.011 0.012
Chain 1: 3200 -18100.597 0.012 0.015
Chain 1: 3300 -18305.220 0.011 0.012
Chain 1: 3400 -17780.324 0.013 0.015
Chain 1: 3500 -18391.919 0.015 0.016
Chain 1: 3600 -17698.947 0.017 0.016
Chain 1: 3700 -18085.474 0.019 0.017
Chain 1: 3800 -17045.727 0.023 0.021
Chain 1: 3900 -17041.873 0.022 0.021
Chain 1: 4000 -17159.184 0.022 0.021
Chain 1: 4100 -17072.985 0.022 0.021
Chain 1: 4200 -16889.328 0.022 0.021
Chain 1: 4300 -17027.667 0.022 0.021
Chain 1: 4400 -16984.585 0.019 0.011
Chain 1: 4500 -16887.114 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001294 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.94 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12336.552 1.000 1.000
Chain 1: 200 -9322.528 0.662 1.000
Chain 1: 300 -8074.356 0.493 0.323
Chain 1: 400 -8235.431 0.374 0.323
Chain 1: 500 -8163.191 0.301 0.155
Chain 1: 600 -8018.646 0.254 0.155
Chain 1: 700 -7944.875 0.219 0.020
Chain 1: 800 -7954.315 0.192 0.020
Chain 1: 900 -7855.496 0.172 0.018
Chain 1: 1000 -8053.345 0.157 0.020
Chain 1: 1100 -7986.219 0.058 0.018
Chain 1: 1200 -7984.499 0.026 0.013
Chain 1: 1300 -7912.954 0.011 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001402 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.02 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -63158.183 1.000 1.000
Chain 1: 200 -18107.197 1.744 2.488
Chain 1: 300 -8755.118 1.519 1.068
Chain 1: 400 -8455.281 1.148 1.068
Chain 1: 500 -8401.306 0.920 1.000
Chain 1: 600 -8038.876 0.774 1.000
Chain 1: 700 -7760.250 0.668 0.045
Chain 1: 800 -8171.105 0.591 0.050
Chain 1: 900 -7982.445 0.528 0.045
Chain 1: 1000 -7899.278 0.476 0.045
Chain 1: 1100 -7681.810 0.379 0.036
Chain 1: 1200 -7580.845 0.132 0.035
Chain 1: 1300 -7803.207 0.028 0.028
Chain 1: 1400 -7686.287 0.026 0.028
Chain 1: 1500 -7651.602 0.026 0.028
Chain 1: 1600 -7549.328 0.022 0.024
Chain 1: 1700 -7652.964 0.020 0.015
Chain 1: 1800 -7649.706 0.015 0.014
Chain 1: 1900 -7621.539 0.013 0.014
Chain 1: 2000 -7582.900 0.013 0.014
Chain 1: 2100 -7620.228 0.010 0.013
Chain 1: 2200 -7720.263 0.010 0.013
Chain 1: 2300 -7623.912 0.009 0.013 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003037 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.37 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86387.630 1.000 1.000
Chain 1: 200 -13427.260 3.217 5.434
Chain 1: 300 -9855.059 2.265 1.000
Chain 1: 400 -10565.837 1.716 1.000
Chain 1: 500 -8759.327 1.414 0.362
Chain 1: 600 -8435.428 1.185 0.362
Chain 1: 700 -8482.742 1.016 0.206
Chain 1: 800 -8952.363 0.896 0.206
Chain 1: 900 -8699.552 0.799 0.067
Chain 1: 1000 -8430.098 0.723 0.067
Chain 1: 1100 -8672.700 0.626 0.052 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8338.859 0.086 0.040
Chain 1: 1300 -8407.380 0.051 0.038
Chain 1: 1400 -8501.686 0.045 0.032
Chain 1: 1500 -8436.731 0.025 0.029
Chain 1: 1600 -8433.976 0.021 0.028
Chain 1: 1700 -8357.959 0.022 0.028
Chain 1: 1800 -8245.834 0.018 0.014
Chain 1: 1900 -8365.192 0.016 0.014
Chain 1: 2000 -8325.571 0.014 0.011
Chain 1: 2100 -8450.902 0.012 0.011
Chain 1: 2200 -8237.924 0.011 0.011
Chain 1: 2300 -8387.070 0.012 0.014
Chain 1: 2400 -8400.958 0.011 0.014
Chain 1: 2500 -8370.395 0.011 0.014
Chain 1: 2600 -8372.154 0.011 0.014
Chain 1: 2700 -8278.579 0.011 0.014
Chain 1: 2800 -8250.592 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00329 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.9 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8408287.034 1.000 1.000
Chain 1: 200 -1582851.889 2.656 4.312
Chain 1: 300 -891185.897 2.029 1.000
Chain 1: 400 -458477.714 1.758 1.000
Chain 1: 500 -358960.544 1.462 0.944
Chain 1: 600 -233651.415 1.308 0.944
Chain 1: 700 -119483.899 1.257 0.944
Chain 1: 800 -86625.149 1.148 0.944
Chain 1: 900 -66881.918 1.053 0.776
Chain 1: 1000 -51611.909 0.977 0.776
Chain 1: 1100 -39042.372 0.909 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38204.154 0.480 0.379
Chain 1: 1300 -26121.144 0.449 0.379
Chain 1: 1400 -25833.540 0.356 0.322
Chain 1: 1500 -22412.415 0.343 0.322
Chain 1: 1600 -21625.964 0.293 0.296
Chain 1: 1700 -20495.535 0.203 0.295
Chain 1: 1800 -20438.351 0.166 0.153
Chain 1: 1900 -20764.061 0.138 0.055
Chain 1: 2000 -19273.969 0.116 0.055
Chain 1: 2100 -19512.225 0.085 0.036
Chain 1: 2200 -19738.944 0.084 0.036
Chain 1: 2300 -19356.004 0.039 0.020
Chain 1: 2400 -19128.199 0.040 0.020
Chain 1: 2500 -18930.525 0.025 0.016
Chain 1: 2600 -18560.874 0.024 0.016
Chain 1: 2700 -18517.764 0.018 0.012
Chain 1: 2800 -18235.042 0.020 0.016
Chain 1: 2900 -18516.092 0.020 0.015
Chain 1: 3000 -18502.158 0.012 0.012
Chain 1: 3100 -18587.202 0.011 0.012
Chain 1: 3200 -18278.037 0.012 0.015
Chain 1: 3300 -18482.582 0.011 0.012
Chain 1: 3400 -17958.025 0.013 0.015
Chain 1: 3500 -18569.247 0.015 0.016
Chain 1: 3600 -17876.659 0.017 0.016
Chain 1: 3700 -18263.052 0.019 0.017
Chain 1: 3800 -17224.035 0.023 0.021
Chain 1: 3900 -17220.230 0.022 0.021
Chain 1: 4000 -17337.482 0.022 0.021
Chain 1: 4100 -17251.418 0.022 0.021
Chain 1: 4200 -17067.843 0.022 0.021
Chain 1: 4300 -17206.059 0.021 0.021
Chain 1: 4400 -17163.113 0.019 0.011
Chain 1: 4500 -17065.693 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001456 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.56 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13173.946 1.000 1.000
Chain 1: 200 -9780.841 0.673 1.000
Chain 1: 300 -8594.919 0.495 0.347
Chain 1: 400 -8759.718 0.376 0.347
Chain 1: 500 -8697.094 0.302 0.138
Chain 1: 600 -8482.976 0.256 0.138
Chain 1: 700 -8400.915 0.221 0.025
Chain 1: 800 -8405.188 0.193 0.025
Chain 1: 900 -8539.634 0.174 0.019
Chain 1: 1000 -8456.510 0.157 0.019
Chain 1: 1100 -8476.348 0.057 0.016
Chain 1: 1200 -8403.094 0.024 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001392 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.92 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -62649.995 1.000 1.000
Chain 1: 200 -18780.040 1.668 2.336
Chain 1: 300 -9344.845 1.449 1.010
Chain 1: 400 -10150.294 1.106 1.010
Chain 1: 500 -8211.500 0.932 1.000
Chain 1: 600 -8569.336 0.784 1.000
Chain 1: 700 -8499.949 0.673 0.236
Chain 1: 800 -8346.993 0.591 0.236
Chain 1: 900 -7770.840 0.534 0.079
Chain 1: 1000 -7667.514 0.482 0.079
Chain 1: 1100 -7688.610 0.382 0.074
Chain 1: 1200 -7787.207 0.150 0.042
Chain 1: 1300 -7620.733 0.051 0.022
Chain 1: 1400 -7692.064 0.044 0.018
Chain 1: 1500 -7629.462 0.021 0.013
Chain 1: 1600 -7881.342 0.020 0.013
Chain 1: 1700 -7728.806 0.021 0.018
Chain 1: 1800 -7565.170 0.022 0.020
Chain 1: 1900 -7658.460 0.015 0.013
Chain 1: 2000 -7698.836 0.015 0.013
Chain 1: 2100 -7581.755 0.016 0.015
Chain 1: 2200 -7832.147 0.018 0.020
Chain 1: 2300 -7694.978 0.017 0.018
Chain 1: 2400 -7603.429 0.018 0.018
Chain 1: 2500 -7471.248 0.019 0.018
Chain 1: 2600 -7672.840 0.018 0.018
Chain 1: 2700 -7671.913 0.016 0.018
Chain 1: 2800 -7555.160 0.015 0.015
Chain 1: 2900 -7433.392 0.016 0.016
Chain 1: 3000 -7605.463 0.018 0.018
Chain 1: 3100 -7569.564 0.017 0.018
Chain 1: 3200 -7830.892 0.017 0.018
Chain 1: 3300 -7519.258 0.019 0.018
Chain 1: 3400 -7753.624 0.021 0.023
Chain 1: 3500 -7472.935 0.023 0.026
Chain 1: 3600 -7536.915 0.021 0.023
Chain 1: 3700 -7518.012 0.021 0.023
Chain 1: 3800 -7442.676 0.021 0.023
Chain 1: 3900 -7458.740 0.019 0.023
Chain 1: 4000 -7448.188 0.017 0.010
Chain 1: 4100 -7457.772 0.017 0.010
Chain 1: 4200 -7548.582 0.015 0.010
Chain 1: 4300 -7437.643 0.012 0.010
Chain 1: 4400 -7484.846 0.010 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003085 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.85 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87288.337 1.000 1.000
Chain 1: 200 -14340.587 3.043 5.087
Chain 1: 300 -10568.912 2.148 1.000
Chain 1: 400 -12448.576 1.649 1.000
Chain 1: 500 -8983.001 1.396 0.386
Chain 1: 600 -8835.619 1.166 0.386
Chain 1: 700 -9270.522 1.006 0.357
Chain 1: 800 -9195.652 0.882 0.357
Chain 1: 900 -9206.977 0.784 0.151
Chain 1: 1000 -9021.382 0.707 0.151
Chain 1: 1100 -9392.534 0.611 0.047 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8840.476 0.109 0.047
Chain 1: 1300 -9174.995 0.077 0.040
Chain 1: 1400 -9043.803 0.063 0.036
Chain 1: 1500 -9039.291 0.025 0.021
Chain 1: 1600 -9130.388 0.024 0.021
Chain 1: 1700 -9182.759 0.020 0.015
Chain 1: 1800 -8727.328 0.024 0.021
Chain 1: 1900 -8837.944 0.025 0.021
Chain 1: 2000 -8847.637 0.023 0.015
Chain 1: 2100 -8770.809 0.020 0.013
Chain 1: 2200 -8760.386 0.014 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003067 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.67 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8426588.485 1.000 1.000
Chain 1: 200 -1588474.557 2.652 4.305
Chain 1: 300 -891574.725 2.029 1.000
Chain 1: 400 -458636.658 1.758 1.000
Chain 1: 500 -358573.301 1.462 0.944
Chain 1: 600 -233632.773 1.307 0.944
Chain 1: 700 -119933.991 1.256 0.944
Chain 1: 800 -87214.013 1.146 0.944
Chain 1: 900 -67584.249 1.051 0.782
Chain 1: 1000 -52415.446 0.975 0.782
Chain 1: 1100 -39920.324 0.906 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39106.893 0.478 0.375
Chain 1: 1300 -27072.529 0.444 0.375
Chain 1: 1400 -26797.834 0.351 0.313
Chain 1: 1500 -23387.051 0.337 0.313
Chain 1: 1600 -22605.753 0.287 0.290
Chain 1: 1700 -21479.156 0.198 0.289
Chain 1: 1800 -21424.054 0.160 0.146
Chain 1: 1900 -21750.913 0.133 0.052
Chain 1: 2000 -20260.703 0.111 0.052
Chain 1: 2100 -20499.264 0.081 0.035
Chain 1: 2200 -20726.249 0.080 0.035
Chain 1: 2300 -20342.784 0.038 0.019
Chain 1: 2400 -20114.528 0.038 0.019
Chain 1: 2500 -19916.629 0.024 0.015
Chain 1: 2600 -19545.979 0.023 0.015
Chain 1: 2700 -19502.708 0.018 0.012
Chain 1: 2800 -19219.201 0.019 0.015
Chain 1: 2900 -19500.811 0.019 0.014
Chain 1: 3000 -19486.968 0.011 0.012
Chain 1: 3100 -19572.079 0.011 0.011
Chain 1: 3200 -19262.225 0.011 0.014
Chain 1: 3300 -19467.377 0.010 0.011
Chain 1: 3400 -18941.368 0.012 0.014
Chain 1: 3500 -19554.644 0.014 0.015
Chain 1: 3600 -18859.441 0.016 0.015
Chain 1: 3700 -19247.585 0.018 0.016
Chain 1: 3800 -18204.457 0.022 0.020
Chain 1: 3900 -18200.503 0.021 0.020
Chain 1: 4000 -18317.824 0.021 0.020
Chain 1: 4100 -18231.418 0.021 0.020
Chain 1: 4200 -18047.043 0.021 0.020
Chain 1: 4300 -18185.882 0.020 0.020
Chain 1: 4400 -18142.156 0.018 0.010
Chain 1: 4500 -18044.582 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001192 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 11.92 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12112.838 1.000 1.000
Chain 1: 200 -9070.306 0.668 1.000
Chain 1: 300 -7885.137 0.495 0.335
Chain 1: 400 -8099.480 0.378 0.335
Chain 1: 500 -7968.919 0.306 0.150
Chain 1: 600 -7830.216 0.258 0.150
Chain 1: 700 -7752.905 0.222 0.026
Chain 1: 800 -7762.991 0.195 0.026
Chain 1: 900 -7661.138 0.175 0.018
Chain 1: 1000 -7806.117 0.159 0.019
Chain 1: 1100 -7801.390 0.059 0.018
Chain 1: 1200 -7790.460 0.026 0.016
Chain 1: 1300 -7731.243 0.011 0.013
Chain 1: 1400 -7751.673 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001422 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.22 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61581.698 1.000 1.000
Chain 1: 200 -17573.128 1.752 2.504
Chain 1: 300 -8703.867 1.508 1.019
Chain 1: 400 -8977.978 1.138 1.019
Chain 1: 500 -7939.700 0.937 1.000
Chain 1: 600 -8898.533 0.799 1.000
Chain 1: 700 -7813.792 0.704 0.139
Chain 1: 800 -8698.708 0.629 0.139
Chain 1: 900 -7955.639 0.570 0.131
Chain 1: 1000 -7728.970 0.516 0.131
Chain 1: 1100 -7697.985 0.416 0.108
Chain 1: 1200 -7761.860 0.166 0.102
Chain 1: 1300 -7691.625 0.065 0.093
Chain 1: 1400 -7626.941 0.063 0.093
Chain 1: 1500 -7566.130 0.051 0.029
Chain 1: 1600 -7498.176 0.041 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003793 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 37.93 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85247.181 1.000 1.000
Chain 1: 200 -13233.543 3.221 5.442
Chain 1: 300 -9662.290 2.270 1.000
Chain 1: 400 -10542.125 1.724 1.000
Chain 1: 500 -8581.915 1.425 0.370
Chain 1: 600 -8365.854 1.192 0.370
Chain 1: 700 -8522.521 1.024 0.228
Chain 1: 800 -8704.468 0.899 0.228
Chain 1: 900 -8510.156 0.801 0.083
Chain 1: 1000 -8341.829 0.723 0.083
Chain 1: 1100 -8525.687 0.625 0.026 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8070.413 0.087 0.026
Chain 1: 1300 -8360.860 0.053 0.026
Chain 1: 1400 -8377.612 0.045 0.023
Chain 1: 1500 -8261.712 0.024 0.022
Chain 1: 1600 -8367.091 0.022 0.021
Chain 1: 1700 -8452.042 0.022 0.021
Chain 1: 1800 -8055.760 0.024 0.022
Chain 1: 1900 -8156.717 0.023 0.020
Chain 1: 2000 -8127.483 0.022 0.014
Chain 1: 2100 -8248.472 0.021 0.014
Chain 1: 2200 -8026.051 0.018 0.014
Chain 1: 2300 -8185.621 0.017 0.014
Chain 1: 2400 -8196.010 0.016 0.014
Chain 1: 2500 -8169.751 0.015 0.013
Chain 1: 2600 -8172.183 0.014 0.012
Chain 1: 2700 -8078.240 0.014 0.012
Chain 1: 2800 -8048.426 0.010 0.012 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003329 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.29 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8369181.869 1.000 1.000
Chain 1: 200 -1579470.292 2.649 4.299
Chain 1: 300 -889366.991 2.025 1.000
Chain 1: 400 -456593.893 1.756 1.000
Chain 1: 500 -357294.399 1.460 0.948
Chain 1: 600 -232661.176 1.306 0.948
Chain 1: 700 -118982.392 1.256 0.948
Chain 1: 800 -86196.149 1.146 0.948
Chain 1: 900 -66546.593 1.052 0.776
Chain 1: 1000 -51334.418 0.976 0.776
Chain 1: 1100 -38800.525 0.909 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37975.520 0.481 0.380
Chain 1: 1300 -25923.096 0.450 0.380
Chain 1: 1400 -25640.990 0.356 0.323
Chain 1: 1500 -22225.320 0.344 0.323
Chain 1: 1600 -21440.740 0.294 0.296
Chain 1: 1700 -20313.568 0.204 0.295
Chain 1: 1800 -20257.447 0.166 0.154
Chain 1: 1900 -20583.272 0.138 0.055
Chain 1: 2000 -19094.528 0.116 0.055
Chain 1: 2100 -19332.978 0.085 0.037
Chain 1: 2200 -19559.198 0.084 0.037
Chain 1: 2300 -19176.697 0.040 0.020
Chain 1: 2400 -18948.870 0.040 0.020
Chain 1: 2500 -18750.953 0.026 0.016
Chain 1: 2600 -18381.494 0.024 0.016
Chain 1: 2700 -18338.612 0.019 0.012
Chain 1: 2800 -18055.589 0.020 0.016
Chain 1: 2900 -18336.732 0.020 0.015
Chain 1: 3000 -18322.977 0.012 0.012
Chain 1: 3100 -18407.863 0.011 0.012
Chain 1: 3200 -18098.823 0.012 0.015
Chain 1: 3300 -18303.369 0.011 0.012
Chain 1: 3400 -17778.728 0.013 0.015
Chain 1: 3500 -18389.957 0.015 0.016
Chain 1: 3600 -17697.538 0.017 0.016
Chain 1: 3700 -18083.642 0.019 0.017
Chain 1: 3800 -17044.751 0.023 0.021
Chain 1: 3900 -17040.949 0.022 0.021
Chain 1: 4000 -17158.238 0.022 0.021
Chain 1: 4100 -17072.019 0.022 0.021
Chain 1: 4200 -16888.627 0.022 0.021
Chain 1: 4300 -17026.780 0.022 0.021
Chain 1: 4400 -16983.853 0.019 0.011
Chain 1: 4500 -16886.436 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001342 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.42 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48845.976 1.000 1.000
Chain 1: 200 -16173.786 1.510 2.020
Chain 1: 300 -18494.247 1.049 1.000
Chain 1: 400 -13878.222 0.870 1.000
Chain 1: 500 -20906.567 0.763 0.336
Chain 1: 600 -15987.898 0.687 0.336
Chain 1: 700 -11408.819 0.646 0.336
Chain 1: 800 -14632.709 0.593 0.336
Chain 1: 900 -19997.909 0.557 0.333
Chain 1: 1000 -10868.418 0.585 0.336
Chain 1: 1100 -22680.079 0.537 0.336 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -10141.239 0.459 0.336
Chain 1: 1300 -13203.554 0.470 0.336
Chain 1: 1400 -18049.735 0.463 0.336
Chain 1: 1500 -11775.867 0.483 0.401
Chain 1: 1600 -10415.939 0.465 0.401
Chain 1: 1700 -9452.357 0.435 0.268
Chain 1: 1800 -15795.493 0.453 0.402
Chain 1: 1900 -10037.791 0.484 0.521 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2000 -9386.964 0.407 0.402
Chain 1: 2100 -10050.631 0.361 0.268
Chain 1: 2200 -10688.437 0.244 0.232
Chain 1: 2300 -18032.659 0.261 0.268
Chain 1: 2400 -9518.003 0.324 0.402
Chain 1: 2500 -11586.558 0.288 0.179
Chain 1: 2600 -10543.935 0.285 0.179
Chain 1: 2700 -9435.286 0.287 0.179
Chain 1: 2800 -11516.184 0.265 0.179
Chain 1: 2900 -9696.873 0.226 0.179
Chain 1: 3000 -9896.411 0.221 0.179
Chain 1: 3100 -12032.093 0.232 0.179
Chain 1: 3200 -9598.363 0.252 0.181
Chain 1: 3300 -9397.165 0.213 0.179
Chain 1: 3400 -9015.009 0.128 0.177
Chain 1: 3500 -9368.299 0.114 0.118
Chain 1: 3600 -17045.954 0.149 0.177
Chain 1: 3700 -9256.404 0.221 0.181
Chain 1: 3800 -11002.443 0.219 0.177
Chain 1: 3900 -9803.577 0.213 0.159
Chain 1: 4000 -11317.424 0.224 0.159
Chain 1: 4100 -10562.377 0.213 0.134
Chain 1: 4200 -10438.461 0.189 0.122
Chain 1: 4300 -14916.009 0.217 0.134
Chain 1: 4400 -9135.972 0.276 0.159
Chain 1: 4500 -9627.066 0.277 0.159
Chain 1: 4600 -10300.202 0.239 0.134
Chain 1: 4700 -15323.256 0.188 0.134
Chain 1: 4800 -9668.636 0.230 0.134
Chain 1: 4900 -9132.783 0.224 0.134
Chain 1: 5000 -13684.408 0.244 0.300
Chain 1: 5100 -12244.403 0.248 0.300
Chain 1: 5200 -14643.752 0.263 0.300
Chain 1: 5300 -10860.110 0.268 0.328
Chain 1: 5400 -9093.793 0.224 0.194
Chain 1: 5500 -8419.314 0.227 0.194
Chain 1: 5600 -9163.754 0.229 0.194
Chain 1: 5700 -9083.504 0.197 0.164
Chain 1: 5800 -8488.242 0.146 0.118
Chain 1: 5900 -14059.097 0.179 0.164
Chain 1: 6000 -9055.751 0.201 0.164
Chain 1: 6100 -13865.923 0.224 0.194
Chain 1: 6200 -8579.717 0.269 0.347
Chain 1: 6300 -9365.569 0.243 0.194
Chain 1: 6400 -8222.751 0.237 0.139
Chain 1: 6500 -13995.506 0.271 0.347
Chain 1: 6600 -10901.321 0.291 0.347
Chain 1: 6700 -8874.158 0.313 0.347
Chain 1: 6800 -9059.010 0.308 0.347
Chain 1: 6900 -8469.645 0.275 0.284
Chain 1: 7000 -11642.704 0.247 0.273
Chain 1: 7100 -9122.688 0.240 0.273
Chain 1: 7200 -8260.211 0.189 0.228
Chain 1: 7300 -11837.040 0.211 0.273
Chain 1: 7400 -13422.582 0.209 0.273
Chain 1: 7500 -9516.765 0.209 0.273
Chain 1: 7600 -9691.968 0.182 0.228
Chain 1: 7700 -8375.215 0.175 0.157
Chain 1: 7800 -10590.599 0.194 0.209
Chain 1: 7900 -8245.019 0.215 0.273
Chain 1: 8000 -8118.282 0.190 0.209
Chain 1: 8100 -8859.334 0.170 0.157
Chain 1: 8200 -8166.457 0.168 0.157
Chain 1: 8300 -8257.907 0.139 0.118
Chain 1: 8400 -8229.710 0.128 0.085
Chain 1: 8500 -8081.389 0.089 0.084
Chain 1: 8600 -11054.446 0.114 0.085
Chain 1: 8700 -8745.386 0.124 0.085
Chain 1: 8800 -9085.935 0.107 0.084
Chain 1: 8900 -8722.589 0.083 0.042
Chain 1: 9000 -8518.055 0.084 0.042
Chain 1: 9100 -8200.521 0.079 0.039
Chain 1: 9200 -8412.162 0.073 0.037
Chain 1: 9300 -8145.057 0.075 0.037
Chain 1: 9400 -10439.689 0.097 0.039
Chain 1: 9500 -8337.024 0.120 0.042
Chain 1: 9600 -10121.594 0.111 0.042
Chain 1: 9700 -10253.329 0.086 0.039
Chain 1: 9800 -11119.878 0.090 0.042
Chain 1: 9900 -8649.018 0.115 0.078
Chain 1: 10000 -9862.015 0.124 0.123
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001397 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.97 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57326.863 1.000 1.000
Chain 1: 200 -17460.492 1.642 2.283
Chain 1: 300 -8694.259 1.431 1.008
Chain 1: 400 -8317.151 1.084 1.008
Chain 1: 500 -8153.684 0.871 1.000
Chain 1: 600 -8713.289 0.737 1.000
Chain 1: 700 -7767.520 0.649 0.122
Chain 1: 800 -8217.965 0.575 0.122
Chain 1: 900 -7919.353 0.515 0.064
Chain 1: 1000 -7677.851 0.467 0.064
Chain 1: 1100 -7724.173 0.367 0.055
Chain 1: 1200 -7610.294 0.140 0.045
Chain 1: 1300 -7661.077 0.040 0.038
Chain 1: 1400 -7761.660 0.037 0.031
Chain 1: 1500 -7565.355 0.038 0.031
Chain 1: 1600 -7719.824 0.033 0.026
Chain 1: 1700 -7472.581 0.024 0.026
Chain 1: 1800 -7550.191 0.020 0.020
Chain 1: 1900 -7499.946 0.017 0.015
Chain 1: 2000 -7518.351 0.014 0.013
Chain 1: 2100 -7473.998 0.014 0.013
Chain 1: 2200 -7687.136 0.015 0.013
Chain 1: 2300 -7545.491 0.016 0.019
Chain 1: 2400 -7596.506 0.016 0.019
Chain 1: 2500 -7546.931 0.014 0.010
Chain 1: 2600 -7499.452 0.012 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002742 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 27.42 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87030.570 1.000 1.000
Chain 1: 200 -13443.878 3.237 5.474
Chain 1: 300 -9822.875 2.281 1.000
Chain 1: 400 -10781.653 1.733 1.000
Chain 1: 500 -8769.064 1.432 0.369
Chain 1: 600 -8290.080 1.203 0.369
Chain 1: 700 -8662.955 1.037 0.230
Chain 1: 800 -9366.786 0.917 0.230
Chain 1: 900 -8681.738 0.824 0.089
Chain 1: 1000 -8406.793 0.745 0.089
Chain 1: 1100 -8620.554 0.647 0.079 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8199.053 0.105 0.075
Chain 1: 1300 -8465.344 0.071 0.058
Chain 1: 1400 -8502.180 0.063 0.051
Chain 1: 1500 -8416.282 0.041 0.043
Chain 1: 1600 -8516.459 0.036 0.033
Chain 1: 1700 -8595.179 0.033 0.031
Chain 1: 1800 -8189.673 0.030 0.031
Chain 1: 1900 -8285.892 0.024 0.025
Chain 1: 2000 -8258.369 0.021 0.012
Chain 1: 2100 -8379.361 0.020 0.012
Chain 1: 2200 -8276.542 0.016 0.012
Chain 1: 2300 -8326.048 0.013 0.012
Chain 1: 2400 -8210.492 0.014 0.012
Chain 1: 2500 -8262.224 0.014 0.012
Chain 1: 2600 -8290.387 0.013 0.012
Chain 1: 2700 -8206.442 0.013 0.012
Chain 1: 2800 -8174.783 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003792 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 37.92 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8438937.065 1.000 1.000
Chain 1: 200 -1590400.281 2.653 4.306
Chain 1: 300 -891782.537 2.030 1.000
Chain 1: 400 -457802.814 1.759 1.000
Chain 1: 500 -357688.018 1.463 0.948
Chain 1: 600 -232396.788 1.309 0.948
Chain 1: 700 -118851.129 1.259 0.948
Chain 1: 800 -86144.048 1.149 0.948
Chain 1: 900 -66541.581 1.054 0.783
Chain 1: 1000 -51386.147 0.978 0.783
Chain 1: 1100 -38910.616 0.910 0.539 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38089.834 0.482 0.380
Chain 1: 1300 -26095.702 0.449 0.380
Chain 1: 1400 -25818.435 0.356 0.321
Chain 1: 1500 -22419.626 0.343 0.321
Chain 1: 1600 -21640.048 0.292 0.295
Chain 1: 1700 -20519.821 0.202 0.295
Chain 1: 1800 -20465.267 0.165 0.152
Chain 1: 1900 -20791.324 0.137 0.055
Chain 1: 2000 -19305.791 0.115 0.055
Chain 1: 2100 -19543.935 0.084 0.036
Chain 1: 2200 -19769.979 0.083 0.036
Chain 1: 2300 -19387.570 0.039 0.020
Chain 1: 2400 -19159.760 0.039 0.020
Chain 1: 2500 -18961.625 0.025 0.016
Chain 1: 2600 -18592.165 0.024 0.016
Chain 1: 2700 -18549.124 0.018 0.012
Chain 1: 2800 -18266.061 0.020 0.015
Chain 1: 2900 -18547.135 0.020 0.015
Chain 1: 3000 -18533.313 0.012 0.012
Chain 1: 3100 -18618.356 0.011 0.012
Chain 1: 3200 -18309.133 0.012 0.015
Chain 1: 3300 -18513.733 0.011 0.012
Chain 1: 3400 -17988.854 0.013 0.015
Chain 1: 3500 -18600.428 0.015 0.015
Chain 1: 3600 -17907.389 0.017 0.015
Chain 1: 3700 -18294.012 0.019 0.017
Chain 1: 3800 -17254.204 0.023 0.021
Chain 1: 3900 -17250.312 0.022 0.021
Chain 1: 4000 -17367.641 0.022 0.021
Chain 1: 4100 -17281.507 0.022 0.021
Chain 1: 4200 -17097.758 0.022 0.021
Chain 1: 4300 -17236.146 0.021 0.021
Chain 1: 4400 -17193.057 0.019 0.011
Chain 1: 4500 -17095.563 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001175 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 11.75 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48875.117 1.000 1.000
Chain 1: 200 -23248.666 1.051 1.102
Chain 1: 300 -14792.095 0.891 1.000
Chain 1: 400 -13626.089 0.690 1.000
Chain 1: 500 -21617.253 0.626 0.572
Chain 1: 600 -15652.430 0.585 0.572
Chain 1: 700 -11888.416 0.547 0.381
Chain 1: 800 -13216.267 0.491 0.381
Chain 1: 900 -12067.931 0.447 0.370
Chain 1: 1000 -21921.918 0.447 0.381
Chain 1: 1100 -13249.667 0.413 0.381
Chain 1: 1200 -11048.924 0.322 0.370
Chain 1: 1300 -12053.964 0.274 0.317
Chain 1: 1400 -9768.073 0.288 0.317
Chain 1: 1500 -21388.743 0.306 0.317
Chain 1: 1600 -11238.014 0.358 0.317
Chain 1: 1700 -12909.798 0.339 0.234
Chain 1: 1800 -11693.192 0.340 0.234
Chain 1: 1900 -10062.014 0.346 0.234
Chain 1: 2000 -11086.204 0.311 0.199
Chain 1: 2100 -13856.330 0.265 0.199
Chain 1: 2200 -9760.885 0.287 0.200
Chain 1: 2300 -17438.470 0.323 0.234
Chain 1: 2400 -9271.861 0.388 0.420
Chain 1: 2500 -9570.034 0.336 0.200
Chain 1: 2600 -9521.368 0.246 0.162
Chain 1: 2700 -9162.166 0.237 0.162
Chain 1: 2800 -9021.696 0.229 0.162
Chain 1: 2900 -9299.360 0.215 0.092
Chain 1: 3000 -15337.956 0.246 0.200
Chain 1: 3100 -9796.118 0.282 0.394
Chain 1: 3200 -9854.901 0.241 0.039
Chain 1: 3300 -9200.518 0.204 0.039
Chain 1: 3400 -9499.302 0.119 0.031
Chain 1: 3500 -9313.111 0.118 0.031
Chain 1: 3600 -9886.279 0.123 0.039
Chain 1: 3700 -9714.540 0.121 0.031
Chain 1: 3800 -8733.874 0.131 0.058
Chain 1: 3900 -10513.099 0.145 0.071
Chain 1: 4000 -10456.460 0.106 0.058
Chain 1: 4100 -9017.435 0.065 0.058
Chain 1: 4200 -9521.186 0.070 0.058
Chain 1: 4300 -8975.073 0.069 0.058
Chain 1: 4400 -8571.936 0.070 0.058
Chain 1: 4500 -8913.695 0.072 0.058
Chain 1: 4600 -10542.841 0.082 0.061
Chain 1: 4700 -8452.767 0.105 0.112
Chain 1: 4800 -9005.596 0.100 0.061
Chain 1: 4900 -8737.529 0.086 0.061
Chain 1: 5000 -9134.425 0.090 0.061
Chain 1: 5100 -11362.868 0.093 0.061
Chain 1: 5200 -9334.693 0.110 0.061
Chain 1: 5300 -10035.829 0.111 0.070
Chain 1: 5400 -13931.332 0.134 0.155
Chain 1: 5500 -10297.578 0.165 0.196
Chain 1: 5600 -9703.658 0.156 0.196
Chain 1: 5700 -13568.692 0.160 0.196
Chain 1: 5800 -8501.635 0.213 0.217
Chain 1: 5900 -12411.912 0.242 0.280
Chain 1: 6000 -10330.430 0.257 0.280
Chain 1: 6100 -9338.356 0.248 0.280
Chain 1: 6200 -8518.516 0.236 0.280
Chain 1: 6300 -14032.227 0.269 0.285
Chain 1: 6400 -9028.080 0.296 0.315
Chain 1: 6500 -8457.874 0.268 0.285
Chain 1: 6600 -8468.841 0.262 0.285
Chain 1: 6700 -9482.978 0.244 0.201
Chain 1: 6800 -11899.852 0.204 0.201
Chain 1: 6900 -11401.873 0.177 0.107
Chain 1: 7000 -8492.038 0.191 0.107
Chain 1: 7100 -12975.172 0.215 0.203
Chain 1: 7200 -8364.137 0.261 0.343
Chain 1: 7300 -9463.024 0.233 0.203
Chain 1: 7400 -9091.954 0.182 0.116
Chain 1: 7500 -8956.430 0.177 0.116
Chain 1: 7600 -9833.547 0.185 0.116
Chain 1: 7700 -9626.853 0.177 0.116
Chain 1: 7800 -9110.550 0.162 0.089
Chain 1: 7900 -8544.704 0.165 0.089
Chain 1: 8000 -10620.376 0.150 0.089
Chain 1: 8100 -10528.084 0.116 0.066
Chain 1: 8200 -9672.087 0.070 0.066
Chain 1: 8300 -11560.248 0.075 0.066
Chain 1: 8400 -8299.041 0.110 0.089
Chain 1: 8500 -8331.806 0.109 0.089
Chain 1: 8600 -8733.125 0.104 0.066
Chain 1: 8700 -8264.169 0.108 0.066
Chain 1: 8800 -8532.738 0.105 0.066
Chain 1: 8900 -9142.159 0.105 0.067
Chain 1: 9000 -8377.455 0.095 0.067
Chain 1: 9100 -9405.273 0.105 0.089
Chain 1: 9200 -10649.307 0.108 0.091
Chain 1: 9300 -8401.099 0.118 0.091
Chain 1: 9400 -8308.410 0.080 0.067
Chain 1: 9500 -8273.675 0.080 0.067
Chain 1: 9600 -9023.412 0.084 0.083
Chain 1: 9700 -11060.829 0.097 0.091
Chain 1: 9800 -8380.174 0.125 0.109
Chain 1: 9900 -9689.079 0.132 0.117
Chain 1: 10000 -8653.805 0.135 0.120
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00141 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.1 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61837.571 1.000 1.000
Chain 1: 200 -17771.067 1.740 2.480
Chain 1: 300 -8825.665 1.498 1.014
Chain 1: 400 -9178.010 1.133 1.014
Chain 1: 500 -8014.384 0.935 1.000
Chain 1: 600 -8897.897 0.796 1.000
Chain 1: 700 -7714.240 0.704 0.153
Chain 1: 800 -8154.385 0.623 0.153
Chain 1: 900 -7838.485 0.558 0.145
Chain 1: 1000 -7759.512 0.503 0.145
Chain 1: 1100 -7791.106 0.404 0.099
Chain 1: 1200 -7812.260 0.156 0.054
Chain 1: 1300 -7703.664 0.056 0.040
Chain 1: 1400 -7670.637 0.053 0.040
Chain 1: 1500 -7612.099 0.039 0.014
Chain 1: 1600 -7860.969 0.032 0.014
Chain 1: 1700 -7511.363 0.022 0.014
Chain 1: 1800 -7621.222 0.018 0.014
Chain 1: 1900 -7558.658 0.014 0.010
Chain 1: 2000 -7593.060 0.014 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002604 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 26.04 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85982.936 1.000 1.000
Chain 1: 200 -13442.277 3.198 5.396
Chain 1: 300 -9871.148 2.253 1.000
Chain 1: 400 -10696.312 1.709 1.000
Chain 1: 500 -8813.831 1.410 0.362
Chain 1: 600 -8494.645 1.181 0.362
Chain 1: 700 -8412.144 1.014 0.214
Chain 1: 800 -9093.675 0.896 0.214
Chain 1: 900 -8672.515 0.802 0.077
Chain 1: 1000 -8433.759 0.725 0.077
Chain 1: 1100 -8590.153 0.627 0.075 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8363.517 0.090 0.049
Chain 1: 1300 -8567.727 0.056 0.038
Chain 1: 1400 -8564.477 0.048 0.028
Chain 1: 1500 -8461.834 0.028 0.027
Chain 1: 1600 -8564.421 0.026 0.024
Chain 1: 1700 -8652.059 0.026 0.024
Chain 1: 1800 -8252.113 0.023 0.024
Chain 1: 1900 -8352.612 0.019 0.018
Chain 1: 2000 -8323.558 0.017 0.012
Chain 1: 2100 -8443.906 0.016 0.012
Chain 1: 2200 -8220.744 0.016 0.012
Chain 1: 2300 -8381.970 0.016 0.012
Chain 1: 2400 -8263.972 0.017 0.014
Chain 1: 2500 -8327.784 0.017 0.014
Chain 1: 2600 -8348.967 0.016 0.014
Chain 1: 2700 -8268.416 0.016 0.014
Chain 1: 2800 -8243.076 0.011 0.012
Chain 1: 2900 -8297.826 0.011 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003807 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 38.07 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8399556.120 1.000 1.000
Chain 1: 200 -1580699.512 2.657 4.314
Chain 1: 300 -889768.962 2.030 1.000
Chain 1: 400 -457572.114 1.759 1.000
Chain 1: 500 -358167.740 1.462 0.945
Chain 1: 600 -233164.161 1.308 0.945
Chain 1: 700 -119271.452 1.258 0.945
Chain 1: 800 -86489.338 1.148 0.945
Chain 1: 900 -66798.517 1.053 0.777
Chain 1: 1000 -51568.386 0.977 0.777
Chain 1: 1100 -39028.621 0.909 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38196.748 0.480 0.379
Chain 1: 1300 -26134.731 0.449 0.379
Chain 1: 1400 -25850.752 0.355 0.321
Chain 1: 1500 -22434.536 0.343 0.321
Chain 1: 1600 -21650.122 0.293 0.295
Chain 1: 1700 -20521.587 0.203 0.295
Chain 1: 1800 -20465.107 0.165 0.152
Chain 1: 1900 -20790.923 0.137 0.055
Chain 1: 2000 -19301.907 0.115 0.055
Chain 1: 2100 -19540.015 0.085 0.036
Chain 1: 2200 -19766.627 0.084 0.036
Chain 1: 2300 -19383.818 0.039 0.020
Chain 1: 2400 -19156.008 0.039 0.020
Chain 1: 2500 -18958.329 0.025 0.016
Chain 1: 2600 -18588.559 0.024 0.016
Chain 1: 2700 -18545.565 0.018 0.012
Chain 1: 2800 -18262.733 0.020 0.015
Chain 1: 2900 -18543.806 0.020 0.015
Chain 1: 3000 -18529.938 0.012 0.012
Chain 1: 3100 -18614.916 0.011 0.012
Chain 1: 3200 -18305.770 0.012 0.015
Chain 1: 3300 -18510.364 0.011 0.012
Chain 1: 3400 -17985.718 0.013 0.015
Chain 1: 3500 -18597.052 0.015 0.015
Chain 1: 3600 -17904.407 0.017 0.015
Chain 1: 3700 -18290.725 0.019 0.017
Chain 1: 3800 -17251.627 0.023 0.021
Chain 1: 3900 -17247.854 0.022 0.021
Chain 1: 4000 -17365.085 0.022 0.021
Chain 1: 4100 -17278.959 0.022 0.021
Chain 1: 4200 -17095.459 0.022 0.021
Chain 1: 4300 -17233.637 0.021 0.021
Chain 1: 4400 -17190.645 0.019 0.011
Chain 1: 4500 -17093.268 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001408 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.08 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48233.421 1.000 1.000
Chain 1: 200 -17885.800 1.348 1.697
Chain 1: 300 -13004.736 1.024 1.000
Chain 1: 400 -13507.058 0.777 1.000
Chain 1: 500 -13474.266 0.622 0.375
Chain 1: 600 -14876.244 0.534 0.375
Chain 1: 700 -12168.151 0.490 0.223
Chain 1: 800 -13300.175 0.439 0.223
Chain 1: 900 -10115.011 0.425 0.223
Chain 1: 1000 -25339.296 0.443 0.315
Chain 1: 1100 -10227.272 0.491 0.315
Chain 1: 1200 -10458.604 0.323 0.223
Chain 1: 1300 -11986.561 0.298 0.127
Chain 1: 1400 -9995.463 0.315 0.199
Chain 1: 1500 -25019.934 0.374 0.223
Chain 1: 1600 -10857.783 0.495 0.315
Chain 1: 1700 -11177.039 0.476 0.315
Chain 1: 1800 -9367.594 0.487 0.315
Chain 1: 1900 -10520.931 0.466 0.199
Chain 1: 2000 -17340.601 0.446 0.199
Chain 1: 2100 -11163.545 0.353 0.199
Chain 1: 2200 -12040.919 0.358 0.199
Chain 1: 2300 -11454.340 0.351 0.199
Chain 1: 2400 -9363.169 0.353 0.223
Chain 1: 2500 -9424.305 0.294 0.193
Chain 1: 2600 -8879.051 0.169 0.110
Chain 1: 2700 -9562.798 0.174 0.110
Chain 1: 2800 -10142.056 0.160 0.073
Chain 1: 2900 -14730.748 0.180 0.073
Chain 1: 3000 -9407.695 0.197 0.073
Chain 1: 3100 -14165.249 0.176 0.073
Chain 1: 3200 -14165.631 0.168 0.072
Chain 1: 3300 -12787.178 0.174 0.108
Chain 1: 3400 -9067.042 0.193 0.108
Chain 1: 3500 -9773.633 0.199 0.108
Chain 1: 3600 -9204.292 0.199 0.108
Chain 1: 3700 -9545.893 0.196 0.108
Chain 1: 3800 -14766.550 0.225 0.312
Chain 1: 3900 -8666.638 0.265 0.336
Chain 1: 4000 -10828.076 0.228 0.200
Chain 1: 4100 -8852.394 0.217 0.200
Chain 1: 4200 -10356.609 0.231 0.200
Chain 1: 4300 -9338.730 0.231 0.200
Chain 1: 4400 -10646.257 0.203 0.145
Chain 1: 4500 -9185.170 0.211 0.159
Chain 1: 4600 -8657.650 0.211 0.159
Chain 1: 4700 -8941.504 0.211 0.159
Chain 1: 4800 -8421.967 0.182 0.145
Chain 1: 4900 -8145.642 0.115 0.123
Chain 1: 5000 -8841.157 0.103 0.109
Chain 1: 5100 -8365.900 0.086 0.079
Chain 1: 5200 -14548.236 0.114 0.079
Chain 1: 5300 -10729.595 0.139 0.079
Chain 1: 5400 -13653.418 0.148 0.079
Chain 1: 5500 -8931.949 0.185 0.079
Chain 1: 5600 -8846.548 0.180 0.079
Chain 1: 5700 -8700.077 0.178 0.079
Chain 1: 5800 -8344.866 0.176 0.079
Chain 1: 5900 -11617.785 0.201 0.214
Chain 1: 6000 -8976.868 0.223 0.282
Chain 1: 6100 -11520.085 0.239 0.282
Chain 1: 6200 -11038.199 0.201 0.221
Chain 1: 6300 -9061.657 0.187 0.218
Chain 1: 6400 -8795.226 0.169 0.218
Chain 1: 6500 -9230.873 0.120 0.047
Chain 1: 6600 -8751.606 0.125 0.055
Chain 1: 6700 -8715.710 0.124 0.055
Chain 1: 6800 -8810.143 0.121 0.055
Chain 1: 6900 -11949.454 0.119 0.055
Chain 1: 7000 -8262.655 0.134 0.055
Chain 1: 7100 -8123.865 0.113 0.047
Chain 1: 7200 -10231.759 0.130 0.055
Chain 1: 7300 -8304.282 0.131 0.055
Chain 1: 7400 -8850.225 0.134 0.062
Chain 1: 7500 -8423.709 0.135 0.062
Chain 1: 7600 -8232.881 0.131 0.062
Chain 1: 7700 -8932.118 0.139 0.078
Chain 1: 7800 -8912.814 0.138 0.078
Chain 1: 7900 -8048.532 0.122 0.078
Chain 1: 8000 -11295.841 0.107 0.078
Chain 1: 8100 -7900.064 0.148 0.107
Chain 1: 8200 -9872.842 0.147 0.107
Chain 1: 8300 -9022.552 0.133 0.094
Chain 1: 8400 -8630.245 0.132 0.094
Chain 1: 8500 -10608.648 0.145 0.107
Chain 1: 8600 -7989.136 0.176 0.186
Chain 1: 8700 -8267.070 0.171 0.186
Chain 1: 8800 -9775.368 0.187 0.186
Chain 1: 8900 -9583.173 0.178 0.186
Chain 1: 9000 -10311.908 0.156 0.154
Chain 1: 9100 -8336.846 0.137 0.154
Chain 1: 9200 -8203.810 0.119 0.094
Chain 1: 9300 -7862.696 0.113 0.071
Chain 1: 9400 -8160.043 0.113 0.071
Chain 1: 9500 -10404.255 0.116 0.071
Chain 1: 9600 -8046.622 0.112 0.071
Chain 1: 9700 -8213.471 0.111 0.071
Chain 1: 9800 -8096.644 0.097 0.043
Chain 1: 9900 -9856.081 0.113 0.071
Chain 1: 10000 -7890.118 0.130 0.179
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001681 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 16.81 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61465.587 1.000 1.000
Chain 1: 200 -17302.500 1.776 2.552
Chain 1: 300 -8641.692 1.518 1.002
Chain 1: 400 -8159.612 1.153 1.002
Chain 1: 500 -8260.426 0.925 1.000
Chain 1: 600 -9025.719 0.785 1.000
Chain 1: 700 -7966.014 0.692 0.133
Chain 1: 800 -8046.197 0.607 0.133
Chain 1: 900 -7851.649 0.542 0.085
Chain 1: 1000 -7654.301 0.490 0.085
Chain 1: 1100 -7666.163 0.391 0.059
Chain 1: 1200 -7612.831 0.136 0.026
Chain 1: 1300 -7709.523 0.037 0.025
Chain 1: 1400 -7837.884 0.033 0.016
Chain 1: 1500 -7646.388 0.034 0.025
Chain 1: 1600 -7558.701 0.027 0.016
Chain 1: 1700 -7520.600 0.014 0.013
Chain 1: 1800 -7556.400 0.013 0.013
Chain 1: 1900 -7618.468 0.012 0.012
Chain 1: 2000 -7614.463 0.009 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002559 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.59 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85962.535 1.000 1.000
Chain 1: 200 -12961.434 3.316 5.632
Chain 1: 300 -9498.262 2.332 1.000
Chain 1: 400 -9919.317 1.760 1.000
Chain 1: 500 -8840.689 1.432 0.365
Chain 1: 600 -8184.882 1.207 0.365
Chain 1: 700 -8354.576 1.037 0.122
Chain 1: 800 -8527.652 0.910 0.122
Chain 1: 900 -8416.809 0.811 0.080
Chain 1: 1000 -8169.604 0.733 0.080
Chain 1: 1100 -8441.158 0.636 0.042 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8214.068 0.075 0.032
Chain 1: 1300 -8159.769 0.040 0.030
Chain 1: 1400 -8170.443 0.035 0.028
Chain 1: 1500 -8188.361 0.023 0.020
Chain 1: 1600 -8184.578 0.015 0.020
Chain 1: 1700 -8136.623 0.014 0.013
Chain 1: 1800 -8015.638 0.013 0.013
Chain 1: 1900 -8124.970 0.014 0.013
Chain 1: 2000 -8091.676 0.011 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003282 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.82 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8415584.762 1.000 1.000
Chain 1: 200 -1587890.428 2.650 4.300
Chain 1: 300 -891108.556 2.027 1.000
Chain 1: 400 -457181.733 1.758 1.000
Chain 1: 500 -357195.217 1.462 0.949
Chain 1: 600 -232086.638 1.308 0.949
Chain 1: 700 -118447.731 1.258 0.949
Chain 1: 800 -85701.590 1.149 0.949
Chain 1: 900 -66074.072 1.054 0.782
Chain 1: 1000 -50886.052 0.979 0.782
Chain 1: 1100 -38391.282 0.911 0.539 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37560.527 0.483 0.382
Chain 1: 1300 -25565.314 0.452 0.382
Chain 1: 1400 -25283.860 0.358 0.325
Chain 1: 1500 -21884.740 0.346 0.325
Chain 1: 1600 -21103.768 0.296 0.298
Chain 1: 1700 -19984.383 0.205 0.297
Chain 1: 1800 -19929.479 0.167 0.155
Chain 1: 1900 -20254.498 0.139 0.056
Chain 1: 2000 -18771.291 0.117 0.056
Chain 1: 2100 -19009.328 0.086 0.037
Chain 1: 2200 -19234.497 0.085 0.037
Chain 1: 2300 -18853.118 0.040 0.020
Chain 1: 2400 -18625.631 0.040 0.020
Chain 1: 2500 -18427.542 0.026 0.016
Chain 1: 2600 -18059.057 0.024 0.016
Chain 1: 2700 -18016.368 0.019 0.013
Chain 1: 2800 -17733.669 0.020 0.016
Chain 1: 2900 -18014.335 0.020 0.016
Chain 1: 3000 -18000.627 0.012 0.013
Chain 1: 3100 -18085.461 0.011 0.012
Chain 1: 3200 -17776.933 0.012 0.016
Chain 1: 3300 -17981.027 0.011 0.012
Chain 1: 3400 -17457.325 0.013 0.016
Chain 1: 3500 -18067.117 0.015 0.016
Chain 1: 3600 -17376.458 0.017 0.016
Chain 1: 3700 -17761.244 0.019 0.017
Chain 1: 3800 -16725.142 0.024 0.022
Chain 1: 3900 -16721.362 0.022 0.022
Chain 1: 4000 -16838.670 0.023 0.022
Chain 1: 4100 -16752.644 0.023 0.022
Chain 1: 4200 -16569.779 0.022 0.022
Chain 1: 4300 -16707.560 0.022 0.022
Chain 1: 4400 -16665.120 0.019 0.011
Chain 1: 4500 -16567.758 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001275 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.75 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12205.630 1.000 1.000
Chain 1: 200 -9110.796 0.670 1.000
Chain 1: 300 -7798.285 0.503 0.340
Chain 1: 400 -7942.338 0.382 0.340
Chain 1: 500 -7932.246 0.305 0.168
Chain 1: 600 -7726.615 0.259 0.168
Chain 1: 700 -7663.712 0.223 0.027
Chain 1: 800 -7696.739 0.196 0.027
Chain 1: 900 -7863.770 0.176 0.021
Chain 1: 1000 -7699.175 0.161 0.021
Chain 1: 1100 -7697.023 0.061 0.021
Chain 1: 1200 -7679.434 0.027 0.018
Chain 1: 1300 -7648.022 0.011 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001432 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.32 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -46784.756 1.000 1.000
Chain 1: 200 -15343.959 1.525 2.049
Chain 1: 300 -8373.220 1.294 1.000
Chain 1: 400 -8387.666 0.971 1.000
Chain 1: 500 -8308.412 0.779 0.833
Chain 1: 600 -8506.513 0.653 0.833
Chain 1: 700 -7814.697 0.572 0.089
Chain 1: 800 -7976.891 0.503 0.089
Chain 1: 900 -7825.029 0.449 0.023
Chain 1: 1000 -7810.714 0.405 0.023
Chain 1: 1100 -7599.331 0.307 0.023
Chain 1: 1200 -7599.914 0.103 0.020
Chain 1: 1300 -7633.759 0.020 0.019
Chain 1: 1400 -7790.781 0.022 0.020
Chain 1: 1500 -7580.458 0.023 0.020
Chain 1: 1600 -7470.318 0.023 0.020
Chain 1: 1700 -7450.908 0.014 0.019
Chain 1: 1800 -7504.673 0.013 0.015
Chain 1: 1900 -7531.021 0.011 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00325 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.5 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86058.931 1.000 1.000
Chain 1: 200 -13077.870 3.290 5.581
Chain 1: 300 -9527.304 2.318 1.000
Chain 1: 400 -10262.466 1.756 1.000
Chain 1: 500 -8461.121 1.448 0.373
Chain 1: 600 -8042.172 1.215 0.373
Chain 1: 700 -8494.466 1.049 0.213
Chain 1: 800 -8865.275 0.923 0.213
Chain 1: 900 -8346.615 0.827 0.072
Chain 1: 1000 -8125.152 0.747 0.072
Chain 1: 1100 -8288.186 0.649 0.062 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8079.110 0.094 0.053
Chain 1: 1300 -8285.767 0.059 0.052
Chain 1: 1400 -8277.212 0.052 0.042
Chain 1: 1500 -8174.718 0.032 0.027
Chain 1: 1600 -8266.081 0.028 0.026
Chain 1: 1700 -8356.572 0.024 0.025
Chain 1: 1800 -7972.221 0.024 0.025
Chain 1: 1900 -8074.472 0.019 0.020
Chain 1: 2000 -8044.134 0.017 0.013
Chain 1: 2100 -8178.955 0.017 0.013
Chain 1: 2200 -7963.191 0.017 0.013
Chain 1: 2300 -8104.458 0.016 0.013
Chain 1: 2400 -8115.240 0.016 0.013
Chain 1: 2500 -8083.581 0.015 0.013
Chain 1: 2600 -8081.543 0.014 0.013
Chain 1: 2700 -7990.897 0.014 0.013
Chain 1: 2800 -7969.486 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003214 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.14 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8403462.749 1.000 1.000
Chain 1: 200 -1586009.491 2.649 4.298
Chain 1: 300 -890729.232 2.026 1.000
Chain 1: 400 -457222.654 1.757 1.000
Chain 1: 500 -357279.768 1.461 0.948
Chain 1: 600 -232283.324 1.308 0.948
Chain 1: 700 -118663.961 1.258 0.948
Chain 1: 800 -85859.427 1.148 0.948
Chain 1: 900 -66239.362 1.053 0.781
Chain 1: 1000 -51054.445 0.978 0.781
Chain 1: 1100 -38549.846 0.910 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37726.875 0.483 0.382
Chain 1: 1300 -25718.173 0.451 0.382
Chain 1: 1400 -25437.164 0.358 0.324
Chain 1: 1500 -22033.291 0.345 0.324
Chain 1: 1600 -21251.180 0.295 0.297
Chain 1: 1700 -20130.284 0.205 0.296
Chain 1: 1800 -20075.221 0.167 0.154
Chain 1: 1900 -20400.850 0.139 0.056
Chain 1: 2000 -18915.615 0.117 0.056
Chain 1: 2100 -19153.811 0.086 0.037
Chain 1: 2200 -19379.339 0.085 0.037
Chain 1: 2300 -18997.532 0.040 0.020
Chain 1: 2400 -18769.914 0.040 0.020
Chain 1: 2500 -18571.621 0.026 0.016
Chain 1: 2600 -18202.629 0.024 0.016
Chain 1: 2700 -18159.919 0.019 0.012
Chain 1: 2800 -17876.855 0.020 0.016
Chain 1: 2900 -18157.869 0.020 0.015
Chain 1: 3000 -18144.159 0.012 0.012
Chain 1: 3100 -18228.984 0.011 0.012
Chain 1: 3200 -17920.137 0.012 0.015
Chain 1: 3300 -18124.537 0.011 0.012
Chain 1: 3400 -17600.103 0.013 0.015
Chain 1: 3500 -18210.845 0.015 0.016
Chain 1: 3600 -17519.132 0.017 0.016
Chain 1: 3700 -17904.692 0.019 0.017
Chain 1: 3800 -16866.664 0.024 0.022
Chain 1: 3900 -16862.862 0.022 0.022
Chain 1: 4000 -16980.213 0.023 0.022
Chain 1: 4100 -16893.985 0.023 0.022
Chain 1: 4200 -16710.813 0.022 0.022
Chain 1: 4300 -16848.842 0.022 0.022
Chain 1: 4400 -16806.103 0.019 0.011
Chain 1: 4500 -16708.700 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001276 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.76 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49881.490 1.000 1.000
Chain 1: 200 -17156.358 1.454 1.907
Chain 1: 300 -20056.902 1.017 1.000
Chain 1: 400 -14619.894 0.856 1.000
Chain 1: 500 -12472.526 0.719 0.372
Chain 1: 600 -16064.521 0.637 0.372
Chain 1: 700 -14269.887 0.564 0.224
Chain 1: 800 -14334.657 0.494 0.224
Chain 1: 900 -12625.788 0.454 0.172
Chain 1: 1000 -13999.533 0.418 0.172
Chain 1: 1100 -12497.291 0.330 0.145
Chain 1: 1200 -12228.765 0.142 0.135
Chain 1: 1300 -13195.869 0.135 0.126
Chain 1: 1400 -13120.449 0.098 0.120
Chain 1: 1500 -11347.098 0.096 0.120
Chain 1: 1600 -12751.504 0.085 0.110
Chain 1: 1700 -13242.513 0.076 0.098
Chain 1: 1800 -16613.315 0.096 0.110
Chain 1: 1900 -11002.468 0.134 0.110
Chain 1: 2000 -11532.378 0.128 0.110
Chain 1: 2100 -13225.014 0.129 0.110
Chain 1: 2200 -9607.557 0.165 0.128
Chain 1: 2300 -12435.041 0.180 0.156
Chain 1: 2400 -10409.170 0.199 0.195
Chain 1: 2500 -9432.561 0.194 0.195
Chain 1: 2600 -9683.946 0.185 0.195
Chain 1: 2700 -9650.373 0.182 0.195
Chain 1: 2800 -10954.128 0.173 0.128
Chain 1: 2900 -9934.606 0.133 0.119
Chain 1: 3000 -11796.175 0.144 0.128
Chain 1: 3100 -8987.899 0.162 0.158
Chain 1: 3200 -15181.393 0.165 0.158
Chain 1: 3300 -11784.977 0.172 0.158
Chain 1: 3400 -11139.608 0.158 0.119
Chain 1: 3500 -11422.917 0.150 0.119
Chain 1: 3600 -9539.610 0.167 0.158
Chain 1: 3700 -8969.724 0.173 0.158
Chain 1: 3800 -8973.406 0.161 0.158
Chain 1: 3900 -9647.269 0.158 0.158
Chain 1: 4000 -10326.441 0.149 0.070
Chain 1: 4100 -14749.557 0.148 0.070
Chain 1: 4200 -12126.296 0.128 0.070
Chain 1: 4300 -10413.249 0.116 0.070
Chain 1: 4400 -9268.493 0.123 0.124
Chain 1: 4500 -9633.903 0.124 0.124
Chain 1: 4600 -9211.308 0.109 0.070
Chain 1: 4700 -11233.969 0.120 0.124
Chain 1: 4800 -14073.162 0.141 0.165
Chain 1: 4900 -9723.036 0.178 0.180
Chain 1: 5000 -14885.140 0.206 0.202
Chain 1: 5100 -9432.709 0.234 0.202
Chain 1: 5200 -9462.529 0.213 0.180
Chain 1: 5300 -13232.944 0.225 0.202
Chain 1: 5400 -8968.503 0.260 0.285
Chain 1: 5500 -9395.244 0.261 0.285
Chain 1: 5600 -11371.288 0.274 0.285
Chain 1: 5700 -9405.580 0.277 0.285
Chain 1: 5800 -16320.974 0.299 0.347
Chain 1: 5900 -9343.624 0.329 0.347
Chain 1: 6000 -10168.587 0.302 0.285
Chain 1: 6100 -16324.612 0.282 0.285
Chain 1: 6200 -10445.348 0.338 0.377
Chain 1: 6300 -14097.016 0.335 0.377
Chain 1: 6400 -17036.411 0.305 0.259
Chain 1: 6500 -9436.580 0.381 0.377
Chain 1: 6600 -10129.416 0.371 0.377
Chain 1: 6700 -8446.174 0.370 0.377
Chain 1: 6800 -9134.740 0.335 0.259
Chain 1: 6900 -8719.361 0.265 0.199
Chain 1: 7000 -12372.215 0.286 0.259
Chain 1: 7100 -10070.371 0.271 0.229
Chain 1: 7200 -8755.026 0.230 0.199
Chain 1: 7300 -8597.610 0.206 0.173
Chain 1: 7400 -9028.677 0.194 0.150
Chain 1: 7500 -11705.201 0.136 0.150
Chain 1: 7600 -8625.778 0.165 0.199
Chain 1: 7700 -9227.811 0.151 0.150
Chain 1: 7800 -9718.051 0.149 0.150
Chain 1: 7900 -9045.731 0.152 0.150
Chain 1: 8000 -8747.029 0.125 0.074
Chain 1: 8100 -8437.690 0.106 0.065
Chain 1: 8200 -12585.707 0.124 0.065
Chain 1: 8300 -10930.427 0.138 0.074
Chain 1: 8400 -9016.894 0.154 0.151
Chain 1: 8500 -9936.871 0.140 0.093
Chain 1: 8600 -8788.932 0.118 0.093
Chain 1: 8700 -9282.692 0.117 0.093
Chain 1: 8800 -8544.821 0.120 0.093
Chain 1: 8900 -9384.002 0.122 0.093
Chain 1: 9000 -8742.758 0.126 0.093
Chain 1: 9100 -8431.690 0.126 0.093
Chain 1: 9200 -8481.249 0.093 0.089
Chain 1: 9300 -9710.320 0.091 0.089
Chain 1: 9400 -12652.211 0.093 0.089
Chain 1: 9500 -8472.509 0.133 0.089
Chain 1: 9600 -8594.719 0.121 0.086
Chain 1: 9700 -9126.312 0.122 0.086
Chain 1: 9800 -9702.017 0.119 0.073
Chain 1: 9900 -10993.092 0.122 0.073
Chain 1: 10000 -8278.425 0.147 0.117
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004398 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 43.98 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58642.290 1.000 1.000
Chain 1: 200 -18310.562 1.601 2.203
Chain 1: 300 -8994.624 1.413 1.036
Chain 1: 400 -8137.651 1.086 1.036
Chain 1: 500 -8852.864 0.885 1.000
Chain 1: 600 -9462.537 0.748 1.000
Chain 1: 700 -8313.350 0.661 0.138
Chain 1: 800 -8220.322 0.580 0.138
Chain 1: 900 -8067.755 0.517 0.105
Chain 1: 1000 -8112.122 0.466 0.105
Chain 1: 1100 -7635.950 0.373 0.081
Chain 1: 1200 -7760.597 0.154 0.064
Chain 1: 1300 -7966.886 0.053 0.062
Chain 1: 1400 -7812.077 0.044 0.026
Chain 1: 1500 -7623.883 0.039 0.025
Chain 1: 1600 -7844.066 0.035 0.025
Chain 1: 1700 -7679.758 0.023 0.021
Chain 1: 1800 -7795.008 0.024 0.021
Chain 1: 1900 -7681.077 0.023 0.021
Chain 1: 2000 -7795.359 0.024 0.021
Chain 1: 2100 -7677.968 0.020 0.020
Chain 1: 2200 -7889.029 0.021 0.021
Chain 1: 2300 -7623.114 0.022 0.021
Chain 1: 2400 -7658.045 0.020 0.021
Chain 1: 2500 -7681.688 0.018 0.015
Chain 1: 2600 -7614.056 0.016 0.015
Chain 1: 2700 -7572.054 0.014 0.015
Chain 1: 2800 -7742.494 0.015 0.015
Chain 1: 2900 -7488.943 0.017 0.015
Chain 1: 3000 -7617.664 0.017 0.017
Chain 1: 3100 -7618.783 0.016 0.017
Chain 1: 3200 -7823.335 0.016 0.017
Chain 1: 3300 -7529.333 0.016 0.017
Chain 1: 3400 -7782.279 0.019 0.022
Chain 1: 3500 -7521.351 0.022 0.026
Chain 1: 3600 -7584.998 0.022 0.026
Chain 1: 3700 -7538.319 0.022 0.026
Chain 1: 3800 -7525.279 0.020 0.026
Chain 1: 3900 -7493.640 0.017 0.017
Chain 1: 4000 -7487.146 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003245 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.45 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87603.087 1.000 1.000
Chain 1: 200 -14031.047 3.122 5.244
Chain 1: 300 -10215.505 2.206 1.000
Chain 1: 400 -12084.050 1.693 1.000
Chain 1: 500 -8632.636 1.434 0.400
Chain 1: 600 -8564.582 1.197 0.400
Chain 1: 700 -8598.218 1.026 0.374
Chain 1: 800 -8852.834 0.902 0.374
Chain 1: 900 -8867.194 0.802 0.155
Chain 1: 1000 -9216.718 0.725 0.155
Chain 1: 1100 -8766.171 0.630 0.051 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8490.693 0.109 0.038
Chain 1: 1300 -8885.633 0.076 0.038
Chain 1: 1400 -8762.515 0.062 0.032
Chain 1: 1500 -8678.584 0.023 0.029
Chain 1: 1600 -8758.979 0.023 0.029
Chain 1: 1700 -8829.125 0.024 0.029
Chain 1: 1800 -8364.465 0.026 0.032
Chain 1: 1900 -8487.017 0.028 0.032
Chain 1: 2000 -8504.267 0.024 0.014
Chain 1: 2100 -8589.680 0.020 0.014
Chain 1: 2200 -8374.134 0.019 0.014
Chain 1: 2300 -8536.036 0.017 0.014
Chain 1: 2400 -8382.630 0.017 0.014
Chain 1: 2500 -8456.218 0.017 0.014
Chain 1: 2600 -8366.865 0.017 0.014
Chain 1: 2700 -8400.845 0.017 0.014
Chain 1: 2800 -8352.010 0.012 0.011
Chain 1: 2900 -8466.798 0.012 0.011
Chain 1: 3000 -8379.423 0.013 0.011
Chain 1: 3100 -8344.178 0.012 0.011
Chain 1: 3200 -8315.951 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002754 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 27.54 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8450569.201 1.000 1.000
Chain 1: 200 -1595225.985 2.649 4.297
Chain 1: 300 -892554.733 2.028 1.000
Chain 1: 400 -458245.997 1.758 1.000
Chain 1: 500 -357522.651 1.463 0.948
Chain 1: 600 -232433.661 1.309 0.948
Chain 1: 700 -119184.493 1.258 0.948
Chain 1: 800 -86509.463 1.148 0.948
Chain 1: 900 -66978.104 1.052 0.787
Chain 1: 1000 -51894.440 0.976 0.787
Chain 1: 1100 -39468.428 0.908 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38669.065 0.480 0.378
Chain 1: 1300 -26710.181 0.446 0.378
Chain 1: 1400 -26442.291 0.352 0.315
Chain 1: 1500 -23050.031 0.339 0.315
Chain 1: 1600 -22274.143 0.289 0.292
Chain 1: 1700 -21156.909 0.199 0.291
Chain 1: 1800 -21103.996 0.161 0.147
Chain 1: 1900 -21430.989 0.134 0.053
Chain 1: 2000 -19945.399 0.112 0.053
Chain 1: 2100 -20183.751 0.082 0.035
Chain 1: 2200 -20409.839 0.081 0.035
Chain 1: 2300 -20027.177 0.038 0.019
Chain 1: 2400 -19799.050 0.038 0.019
Chain 1: 2500 -19600.594 0.024 0.015
Chain 1: 2600 -19230.307 0.023 0.015
Chain 1: 2700 -19187.284 0.018 0.012
Chain 1: 2800 -18903.410 0.019 0.015
Chain 1: 2900 -19185.064 0.019 0.015
Chain 1: 3000 -19171.323 0.012 0.012
Chain 1: 3100 -19256.337 0.011 0.012
Chain 1: 3200 -18946.584 0.011 0.015
Chain 1: 3300 -19151.737 0.011 0.012
Chain 1: 3400 -18625.567 0.012 0.015
Chain 1: 3500 -19238.808 0.014 0.015
Chain 1: 3600 -18543.806 0.016 0.015
Chain 1: 3700 -18931.684 0.018 0.016
Chain 1: 3800 -17888.567 0.022 0.020
Chain 1: 3900 -17884.584 0.021 0.020
Chain 1: 4000 -18001.996 0.021 0.020
Chain 1: 4100 -17915.450 0.022 0.020
Chain 1: 4200 -17731.198 0.021 0.020
Chain 1: 4300 -17870.021 0.021 0.020
Chain 1: 4400 -17826.336 0.018 0.010
Chain 1: 4500 -17728.738 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001329 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.29 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49241.880 1.000 1.000
Chain 1: 200 -21065.458 1.169 1.338
Chain 1: 300 -20039.994 0.796 1.000
Chain 1: 400 -14574.916 0.691 1.000
Chain 1: 500 -12874.906 0.579 0.375
Chain 1: 600 -21677.101 0.550 0.406
Chain 1: 700 -12330.247 0.580 0.406
Chain 1: 800 -14825.030 0.529 0.406
Chain 1: 900 -14043.273 0.476 0.375
Chain 1: 1000 -19544.067 0.457 0.375
Chain 1: 1100 -14447.747 0.392 0.353
Chain 1: 1200 -14763.612 0.260 0.281
Chain 1: 1300 -11687.511 0.281 0.281
Chain 1: 1400 -10372.039 0.257 0.263
Chain 1: 1500 -11025.315 0.249 0.263
Chain 1: 1600 -9998.628 0.219 0.168
Chain 1: 1700 -9767.004 0.146 0.127
Chain 1: 1800 -10003.328 0.131 0.103
Chain 1: 1900 -13838.275 0.153 0.127
Chain 1: 2000 -18417.201 0.150 0.127
Chain 1: 2100 -9536.393 0.208 0.127
Chain 1: 2200 -14903.000 0.242 0.249
Chain 1: 2300 -14078.334 0.221 0.127
Chain 1: 2400 -9607.855 0.255 0.249
Chain 1: 2500 -9419.897 0.251 0.249
Chain 1: 2600 -11570.366 0.259 0.249
Chain 1: 2700 -9930.707 0.274 0.249
Chain 1: 2800 -11528.143 0.285 0.249
Chain 1: 2900 -9707.593 0.276 0.188
Chain 1: 3000 -9392.652 0.255 0.186
Chain 1: 3100 -10715.816 0.174 0.165
Chain 1: 3200 -12044.264 0.149 0.139
Chain 1: 3300 -16521.346 0.170 0.165
Chain 1: 3400 -9714.975 0.194 0.165
Chain 1: 3500 -9314.487 0.196 0.165
Chain 1: 3600 -10551.266 0.189 0.139
Chain 1: 3700 -8755.559 0.193 0.139
Chain 1: 3800 -8887.955 0.181 0.123
Chain 1: 3900 -10777.989 0.179 0.123
Chain 1: 4000 -10998.247 0.178 0.123
Chain 1: 4100 -8930.914 0.189 0.175
Chain 1: 4200 -12132.273 0.204 0.205
Chain 1: 4300 -8947.069 0.213 0.205
Chain 1: 4400 -9835.884 0.152 0.175
Chain 1: 4500 -8945.099 0.157 0.175
Chain 1: 4600 -10849.912 0.163 0.176
Chain 1: 4700 -8550.411 0.170 0.176
Chain 1: 4800 -8658.163 0.169 0.176
Chain 1: 4900 -8552.089 0.153 0.176
Chain 1: 5000 -17633.395 0.203 0.231
Chain 1: 5100 -9109.499 0.273 0.264
Chain 1: 5200 -9157.524 0.247 0.176
Chain 1: 5300 -10302.015 0.223 0.111
Chain 1: 5400 -8924.120 0.229 0.154
Chain 1: 5500 -9355.841 0.224 0.154
Chain 1: 5600 -9052.413 0.209 0.111
Chain 1: 5700 -9254.551 0.185 0.046
Chain 1: 5800 -9112.961 0.185 0.046
Chain 1: 5900 -14889.064 0.223 0.111
Chain 1: 6000 -10779.955 0.209 0.111
Chain 1: 6100 -10240.958 0.121 0.053
Chain 1: 6200 -9512.503 0.128 0.077
Chain 1: 6300 -13532.971 0.147 0.077
Chain 1: 6400 -13925.950 0.134 0.053
Chain 1: 6500 -10913.084 0.157 0.077
Chain 1: 6600 -9460.022 0.169 0.154
Chain 1: 6700 -9029.546 0.172 0.154
Chain 1: 6800 -10751.517 0.186 0.160
Chain 1: 6900 -11220.175 0.151 0.154
Chain 1: 7000 -8708.675 0.142 0.154
Chain 1: 7100 -8496.475 0.139 0.154
Chain 1: 7200 -8234.405 0.135 0.154
Chain 1: 7300 -11251.806 0.132 0.154
Chain 1: 7400 -8395.430 0.163 0.160
Chain 1: 7500 -11151.836 0.160 0.160
Chain 1: 7600 -10803.323 0.148 0.160
Chain 1: 7700 -8595.206 0.169 0.247
Chain 1: 7800 -8420.694 0.155 0.247
Chain 1: 7900 -10525.708 0.171 0.247
Chain 1: 8000 -8568.757 0.165 0.228
Chain 1: 8100 -10183.098 0.178 0.228
Chain 1: 8200 -12517.503 0.194 0.228
Chain 1: 8300 -8228.480 0.219 0.228
Chain 1: 8400 -8474.620 0.188 0.200
Chain 1: 8500 -8154.195 0.167 0.186
Chain 1: 8600 -8288.475 0.166 0.186
Chain 1: 8700 -9162.135 0.150 0.159
Chain 1: 8800 -8809.094 0.151 0.159
Chain 1: 8900 -9130.467 0.135 0.095
Chain 1: 9000 -10148.925 0.122 0.095
Chain 1: 9100 -8059.795 0.132 0.095
Chain 1: 9200 -11617.918 0.144 0.095
Chain 1: 9300 -9017.979 0.121 0.095
Chain 1: 9400 -11202.776 0.138 0.100
Chain 1: 9500 -8868.741 0.160 0.195
Chain 1: 9600 -8425.535 0.164 0.195
Chain 1: 9700 -8139.770 0.158 0.195
Chain 1: 9800 -9061.302 0.164 0.195
Chain 1: 9900 -8601.399 0.166 0.195
Chain 1: 10000 -9150.137 0.161 0.195
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001607 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 16.07 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56029.606 1.000 1.000
Chain 1: 200 -17425.812 1.608 2.215
Chain 1: 300 -8805.372 1.398 1.000
Chain 1: 400 -8619.625 1.054 1.000
Chain 1: 500 -8642.128 0.844 0.979
Chain 1: 600 -8758.825 0.705 0.979
Chain 1: 700 -8325.363 0.612 0.052
Chain 1: 800 -8352.210 0.536 0.052
Chain 1: 900 -7984.924 0.481 0.046
Chain 1: 1000 -8076.935 0.434 0.046
Chain 1: 1100 -7769.643 0.338 0.040
Chain 1: 1200 -7709.014 0.118 0.022
Chain 1: 1300 -7954.873 0.023 0.022
Chain 1: 1400 -7894.554 0.021 0.013
Chain 1: 1500 -7661.495 0.024 0.030
Chain 1: 1600 -7591.182 0.024 0.030
Chain 1: 1700 -7590.201 0.019 0.011
Chain 1: 1800 -7651.884 0.019 0.011
Chain 1: 1900 -7666.137 0.015 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002495 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 24.95 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85584.769 1.000 1.000
Chain 1: 200 -13701.113 3.123 5.247
Chain 1: 300 -10024.960 2.204 1.000
Chain 1: 400 -11108.685 1.678 1.000
Chain 1: 500 -8819.996 1.394 0.367
Chain 1: 600 -9148.852 1.168 0.367
Chain 1: 700 -8933.513 1.004 0.259
Chain 1: 800 -8718.681 0.882 0.259
Chain 1: 900 -8765.747 0.784 0.098
Chain 1: 1000 -8754.007 0.706 0.098
Chain 1: 1100 -8615.144 0.608 0.036 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8411.969 0.086 0.025
Chain 1: 1300 -8781.441 0.053 0.025
Chain 1: 1400 -8690.650 0.044 0.024
Chain 1: 1500 -8543.930 0.020 0.024
Chain 1: 1600 -8658.784 0.018 0.017
Chain 1: 1700 -8729.435 0.016 0.016
Chain 1: 1800 -8295.488 0.019 0.016
Chain 1: 1900 -8400.296 0.020 0.016
Chain 1: 2000 -8376.204 0.020 0.016
Chain 1: 2100 -8336.355 0.019 0.013
Chain 1: 2200 -8317.938 0.017 0.012
Chain 1: 2300 -8447.208 0.014 0.012
Chain 1: 2400 -8303.524 0.015 0.013
Chain 1: 2500 -8371.715 0.014 0.012
Chain 1: 2600 -8291.641 0.013 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003153 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.53 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8372243.666 1.000 1.000
Chain 1: 200 -1578985.717 2.651 4.302
Chain 1: 300 -890794.107 2.025 1.000
Chain 1: 400 -458249.046 1.755 1.000
Chain 1: 500 -359258.222 1.459 0.944
Chain 1: 600 -233975.732 1.305 0.944
Chain 1: 700 -119842.285 1.255 0.944
Chain 1: 800 -86997.506 1.145 0.944
Chain 1: 900 -67251.780 1.050 0.773
Chain 1: 1000 -51991.850 0.975 0.773
Chain 1: 1100 -39413.921 0.907 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38584.434 0.479 0.378
Chain 1: 1300 -26472.640 0.447 0.378
Chain 1: 1400 -26186.414 0.354 0.319
Chain 1: 1500 -22756.777 0.341 0.319
Chain 1: 1600 -21969.136 0.291 0.294
Chain 1: 1700 -20833.954 0.201 0.294
Chain 1: 1800 -20776.429 0.164 0.151
Chain 1: 1900 -21102.885 0.136 0.054
Chain 1: 2000 -19609.191 0.114 0.054
Chain 1: 2100 -19847.588 0.084 0.036
Chain 1: 2200 -20075.182 0.083 0.036
Chain 1: 2300 -19691.334 0.039 0.019
Chain 1: 2400 -19463.235 0.039 0.019
Chain 1: 2500 -19265.706 0.025 0.015
Chain 1: 2600 -18895.094 0.023 0.015
Chain 1: 2700 -18851.860 0.018 0.012
Chain 1: 2800 -18568.781 0.019 0.015
Chain 1: 2900 -18850.265 0.019 0.015
Chain 1: 3000 -18836.279 0.012 0.012
Chain 1: 3100 -18921.361 0.011 0.012
Chain 1: 3200 -18611.736 0.012 0.015
Chain 1: 3300 -18816.726 0.011 0.012
Chain 1: 3400 -18291.262 0.012 0.015
Chain 1: 3500 -18903.884 0.015 0.015
Chain 1: 3600 -18209.628 0.016 0.015
Chain 1: 3700 -18597.162 0.018 0.017
Chain 1: 3800 -17555.562 0.023 0.021
Chain 1: 3900 -17551.766 0.021 0.021
Chain 1: 4000 -17668.976 0.022 0.021
Chain 1: 4100 -17582.718 0.022 0.021
Chain 1: 4200 -17398.693 0.021 0.021
Chain 1: 4300 -17537.223 0.021 0.021
Chain 1: 4400 -17493.790 0.018 0.011
Chain 1: 4500 -17396.367 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001293 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.93 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49198.130 1.000 1.000
Chain 1: 200 -16485.357 1.492 1.984
Chain 1: 300 -21905.293 1.077 1.000
Chain 1: 400 -26543.889 0.852 1.000
Chain 1: 500 -16709.516 0.799 0.589
Chain 1: 600 -13142.609 0.711 0.589
Chain 1: 700 -12709.684 0.614 0.271
Chain 1: 800 -14205.654 0.551 0.271
Chain 1: 900 -13960.766 0.491 0.247
Chain 1: 1000 -10995.836 0.469 0.270
Chain 1: 1100 -26971.738 0.429 0.270
Chain 1: 1200 -11768.602 0.359 0.270
Chain 1: 1300 -13448.336 0.347 0.270
Chain 1: 1400 -10915.639 0.353 0.270
Chain 1: 1500 -10205.694 0.301 0.232
Chain 1: 1600 -10656.800 0.278 0.125
Chain 1: 1700 -11241.837 0.280 0.125
Chain 1: 1800 -11339.281 0.270 0.125
Chain 1: 1900 -21629.847 0.316 0.232
Chain 1: 2000 -19525.641 0.300 0.125
Chain 1: 2100 -11290.158 0.313 0.125
Chain 1: 2200 -11598.673 0.187 0.108
Chain 1: 2300 -10103.904 0.189 0.108
Chain 1: 2400 -12246.312 0.183 0.108
Chain 1: 2500 -11218.958 0.186 0.108
Chain 1: 2600 -10631.241 0.187 0.108
Chain 1: 2700 -11816.020 0.192 0.108
Chain 1: 2800 -11127.042 0.197 0.108
Chain 1: 2900 -12372.315 0.160 0.101
Chain 1: 3000 -9345.106 0.181 0.101
Chain 1: 3100 -10680.271 0.121 0.101
Chain 1: 3200 -9667.231 0.129 0.105
Chain 1: 3300 -19674.933 0.165 0.105
Chain 1: 3400 -9276.715 0.259 0.105
Chain 1: 3500 -13348.929 0.281 0.125
Chain 1: 3600 -10254.597 0.305 0.302
Chain 1: 3700 -11077.827 0.303 0.302
Chain 1: 3800 -9129.427 0.318 0.302
Chain 1: 3900 -11826.214 0.331 0.302
Chain 1: 4000 -9787.783 0.319 0.228
Chain 1: 4100 -12484.974 0.328 0.228
Chain 1: 4200 -10806.467 0.333 0.228
Chain 1: 4300 -11659.397 0.290 0.216
Chain 1: 4400 -9874.349 0.196 0.213
Chain 1: 4500 -9269.327 0.172 0.208
Chain 1: 4600 -11607.790 0.162 0.201
Chain 1: 4700 -10815.062 0.162 0.201
Chain 1: 4800 -9284.268 0.157 0.181
Chain 1: 4900 -8958.178 0.137 0.165
Chain 1: 5000 -13119.682 0.148 0.165
Chain 1: 5100 -9008.961 0.172 0.165
Chain 1: 5200 -9125.508 0.158 0.165
Chain 1: 5300 -9445.448 0.154 0.165
Chain 1: 5400 -9857.231 0.140 0.073
Chain 1: 5500 -9715.836 0.135 0.073
Chain 1: 5600 -13434.583 0.143 0.073
Chain 1: 5700 -14052.673 0.140 0.044
Chain 1: 5800 -14211.197 0.124 0.042
Chain 1: 5900 -13921.003 0.123 0.042
Chain 1: 6000 -8790.095 0.150 0.042
Chain 1: 6100 -8779.452 0.104 0.034
Chain 1: 6200 -8606.369 0.105 0.034
Chain 1: 6300 -11175.971 0.124 0.042
Chain 1: 6400 -13677.910 0.139 0.044
Chain 1: 6500 -8803.909 0.192 0.183
Chain 1: 6600 -9076.182 0.168 0.044
Chain 1: 6700 -14144.162 0.199 0.183
Chain 1: 6800 -8933.359 0.256 0.230
Chain 1: 6900 -9482.363 0.260 0.230
Chain 1: 7000 -8897.582 0.208 0.183
Chain 1: 7100 -8663.120 0.211 0.183
Chain 1: 7200 -8878.668 0.211 0.183
Chain 1: 7300 -9709.258 0.197 0.086
Chain 1: 7400 -9045.677 0.186 0.073
Chain 1: 7500 -13380.812 0.163 0.073
Chain 1: 7600 -9110.874 0.207 0.086
Chain 1: 7700 -13691.961 0.204 0.086
Chain 1: 7800 -12648.818 0.154 0.082
Chain 1: 7900 -8807.685 0.192 0.086
Chain 1: 8000 -12506.691 0.215 0.296
Chain 1: 8100 -8784.538 0.255 0.324
Chain 1: 8200 -11836.819 0.278 0.324
Chain 1: 8300 -10346.774 0.284 0.324
Chain 1: 8400 -8460.834 0.299 0.324
Chain 1: 8500 -8484.852 0.267 0.296
Chain 1: 8600 -9936.618 0.235 0.258
Chain 1: 8700 -8752.372 0.215 0.223
Chain 1: 8800 -8717.264 0.207 0.223
Chain 1: 8900 -12276.620 0.192 0.223
Chain 1: 9000 -10719.584 0.177 0.146
Chain 1: 9100 -9237.597 0.151 0.146
Chain 1: 9200 -9779.735 0.131 0.145
Chain 1: 9300 -8384.125 0.133 0.146
Chain 1: 9400 -8841.029 0.116 0.145
Chain 1: 9500 -9333.450 0.121 0.145
Chain 1: 9600 -10659.502 0.119 0.135
Chain 1: 9700 -8635.311 0.128 0.145
Chain 1: 9800 -9087.805 0.133 0.145
Chain 1: 9900 -11060.374 0.122 0.145
Chain 1: 10000 -9090.237 0.129 0.160
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00155 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.5 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58569.181 1.000 1.000
Chain 1: 200 -18104.868 1.617 2.235
Chain 1: 300 -8869.817 1.425 1.041
Chain 1: 400 -8153.231 1.091 1.041
Chain 1: 500 -8247.585 0.875 1.000
Chain 1: 600 -9421.590 0.750 1.000
Chain 1: 700 -7853.514 0.671 0.200
Chain 1: 800 -8185.986 0.593 0.200
Chain 1: 900 -8012.524 0.529 0.125
Chain 1: 1000 -7920.790 0.477 0.125
Chain 1: 1100 -7615.663 0.381 0.088
Chain 1: 1200 -7728.307 0.159 0.041
Chain 1: 1300 -7700.591 0.056 0.040
Chain 1: 1400 -7699.417 0.047 0.022
Chain 1: 1500 -7566.742 0.047 0.022
Chain 1: 1600 -7773.202 0.038 0.022
Chain 1: 1700 -7592.702 0.020 0.022
Chain 1: 1800 -7713.780 0.018 0.018
Chain 1: 1900 -7593.471 0.017 0.016
Chain 1: 2000 -7668.712 0.017 0.016
Chain 1: 2100 -7585.025 0.014 0.016
Chain 1: 2200 -7752.437 0.015 0.016
Chain 1: 2300 -7575.328 0.017 0.018
Chain 1: 2400 -7699.973 0.018 0.018
Chain 1: 2500 -7637.247 0.017 0.016
Chain 1: 2600 -7540.685 0.016 0.016
Chain 1: 2700 -7522.537 0.014 0.016
Chain 1: 2800 -7517.376 0.012 0.013
Chain 1: 2900 -7395.076 0.012 0.013
Chain 1: 3000 -7541.794 0.013 0.016
Chain 1: 3100 -7543.235 0.012 0.016
Chain 1: 3200 -7759.411 0.013 0.016
Chain 1: 3300 -7482.391 0.014 0.016
Chain 1: 3400 -7710.218 0.015 0.017
Chain 1: 3500 -7453.140 0.018 0.019
Chain 1: 3600 -7521.563 0.018 0.019
Chain 1: 3700 -7470.332 0.018 0.019
Chain 1: 3800 -7469.986 0.018 0.019
Chain 1: 3900 -7431.843 0.017 0.019
Chain 1: 4000 -7423.513 0.015 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00288 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 28.8 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86555.217 1.000 1.000
Chain 1: 200 -13903.081 3.113 5.226
Chain 1: 300 -10218.633 2.195 1.000
Chain 1: 400 -11330.358 1.671 1.000
Chain 1: 500 -9214.191 1.383 0.361
Chain 1: 600 -8644.082 1.163 0.361
Chain 1: 700 -8704.430 0.998 0.230
Chain 1: 800 -9618.401 0.885 0.230
Chain 1: 900 -8991.593 0.795 0.098
Chain 1: 1000 -9014.384 0.715 0.098
Chain 1: 1100 -9009.660 0.615 0.095 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8622.238 0.097 0.070
Chain 1: 1300 -8893.042 0.064 0.066
Chain 1: 1400 -8905.745 0.055 0.045
Chain 1: 1500 -8749.749 0.034 0.030
Chain 1: 1600 -8864.635 0.028 0.018
Chain 1: 1700 -8934.633 0.028 0.018
Chain 1: 1800 -8504.108 0.024 0.018
Chain 1: 1900 -8607.924 0.018 0.013
Chain 1: 2000 -8583.220 0.018 0.013
Chain 1: 2100 -8718.946 0.020 0.016
Chain 1: 2200 -8512.617 0.018 0.016
Chain 1: 2300 -8608.933 0.016 0.013
Chain 1: 2400 -8672.248 0.016 0.013
Chain 1: 2500 -8616.473 0.015 0.012
Chain 1: 2600 -8620.652 0.014 0.011
Chain 1: 2700 -8535.930 0.014 0.011
Chain 1: 2800 -8492.589 0.010 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003363 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.63 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8369931.984 1.000 1.000
Chain 1: 200 -1577347.638 2.653 4.306
Chain 1: 300 -890720.808 2.026 1.000
Chain 1: 400 -458563.202 1.755 1.000
Chain 1: 500 -359628.556 1.459 0.942
Chain 1: 600 -234367.041 1.305 0.942
Chain 1: 700 -120144.755 1.254 0.942
Chain 1: 800 -87259.067 1.145 0.942
Chain 1: 900 -67499.154 1.050 0.771
Chain 1: 1000 -52218.012 0.974 0.771
Chain 1: 1100 -39625.075 0.906 0.534 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38792.610 0.478 0.377
Chain 1: 1300 -26667.959 0.446 0.377
Chain 1: 1400 -26380.435 0.353 0.318
Chain 1: 1500 -22947.023 0.340 0.318
Chain 1: 1600 -22158.210 0.290 0.293
Chain 1: 1700 -21021.520 0.201 0.293
Chain 1: 1800 -20963.454 0.163 0.150
Chain 1: 1900 -21289.917 0.135 0.054
Chain 1: 2000 -19795.273 0.114 0.054
Chain 1: 2100 -20033.834 0.083 0.036
Chain 1: 2200 -20261.530 0.082 0.036
Chain 1: 2300 -19877.528 0.039 0.019
Chain 1: 2400 -19649.407 0.039 0.019
Chain 1: 2500 -19451.913 0.025 0.015
Chain 1: 2600 -19081.407 0.023 0.015
Chain 1: 2700 -19038.104 0.018 0.012
Chain 1: 2800 -18755.118 0.019 0.015
Chain 1: 2900 -19036.551 0.019 0.015
Chain 1: 3000 -19022.559 0.012 0.012
Chain 1: 3100 -19107.678 0.011 0.012
Chain 1: 3200 -18798.082 0.011 0.015
Chain 1: 3300 -19002.997 0.011 0.012
Chain 1: 3400 -18477.652 0.012 0.015
Chain 1: 3500 -19090.190 0.014 0.015
Chain 1: 3600 -18395.952 0.016 0.015
Chain 1: 3700 -18783.549 0.018 0.016
Chain 1: 3800 -17742.049 0.022 0.021
Chain 1: 3900 -17738.223 0.021 0.021
Chain 1: 4000 -17855.429 0.022 0.021
Chain 1: 4100 -17769.255 0.022 0.021
Chain 1: 4200 -17585.151 0.021 0.021
Chain 1: 4300 -17723.729 0.021 0.021
Chain 1: 4400 -17680.342 0.018 0.010
Chain 1: 4500 -17582.869 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001619 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 16.19 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12275.802 1.000 1.000
Chain 1: 200 -9193.101 0.668 1.000
Chain 1: 300 -7859.795 0.502 0.335
Chain 1: 400 -8076.243 0.383 0.335
Chain 1: 500 -7808.243 0.313 0.170
Chain 1: 600 -7804.458 0.261 0.170
Chain 1: 700 -7745.513 0.225 0.034
Chain 1: 800 -7771.976 0.197 0.034
Chain 1: 900 -7760.387 0.175 0.027
Chain 1: 1000 -7805.581 0.158 0.027
Chain 1: 1100 -7842.567 0.059 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001766 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 17.66 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58081.995 1.000 1.000
Chain 1: 200 -17673.211 1.643 2.286
Chain 1: 300 -8627.945 1.445 1.048
Chain 1: 400 -8144.361 1.099 1.048
Chain 1: 500 -8346.703 0.884 1.000
Chain 1: 600 -8523.420 0.740 1.000
Chain 1: 700 -7926.039 0.645 0.075
Chain 1: 800 -8082.980 0.567 0.075
Chain 1: 900 -7866.443 0.507 0.059
Chain 1: 1000 -7797.738 0.457 0.059
Chain 1: 1100 -7667.728 0.359 0.028
Chain 1: 1200 -7665.117 0.130 0.024
Chain 1: 1300 -7606.719 0.026 0.021
Chain 1: 1400 -7843.209 0.023 0.021
Chain 1: 1500 -7559.790 0.024 0.021
Chain 1: 1600 -7615.803 0.023 0.019
Chain 1: 1700 -7478.632 0.017 0.018
Chain 1: 1800 -7530.394 0.016 0.017
Chain 1: 1900 -7559.947 0.014 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003876 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 38.76 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86045.228 1.000 1.000
Chain 1: 200 -13434.120 3.202 5.405
Chain 1: 300 -9767.964 2.260 1.000
Chain 1: 400 -10659.210 1.716 1.000
Chain 1: 500 -8752.711 1.416 0.375
Chain 1: 600 -8644.700 1.182 0.375
Chain 1: 700 -8241.211 1.020 0.218
Chain 1: 800 -8940.510 0.903 0.218
Chain 1: 900 -8594.121 0.807 0.084
Chain 1: 1000 -8345.462 0.729 0.084
Chain 1: 1100 -8547.575 0.632 0.078 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8079.007 0.097 0.058
Chain 1: 1300 -8397.767 0.063 0.049
Chain 1: 1400 -8449.026 0.055 0.040
Chain 1: 1500 -8313.037 0.035 0.038
Chain 1: 1600 -8420.597 0.035 0.038
Chain 1: 1700 -8486.526 0.031 0.030
Chain 1: 1800 -8061.995 0.029 0.030
Chain 1: 1900 -8163.992 0.026 0.024
Chain 1: 2000 -8138.773 0.023 0.016
Chain 1: 2100 -8265.973 0.022 0.015
Chain 1: 2200 -8065.038 0.019 0.015
Chain 1: 2300 -8159.405 0.016 0.013
Chain 1: 2400 -8227.328 0.017 0.013
Chain 1: 2500 -8173.506 0.016 0.012
Chain 1: 2600 -8175.894 0.014 0.012
Chain 1: 2700 -8092.161 0.015 0.012
Chain 1: 2800 -8050.735 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003664 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 36.64 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8366601.959 1.000 1.000
Chain 1: 200 -1577925.426 2.651 4.302
Chain 1: 300 -890647.129 2.025 1.000
Chain 1: 400 -458116.678 1.755 1.000
Chain 1: 500 -359149.061 1.459 0.944
Chain 1: 600 -234037.455 1.305 0.944
Chain 1: 700 -119747.564 1.255 0.944
Chain 1: 800 -86811.335 1.145 0.944
Chain 1: 900 -67043.320 1.051 0.772
Chain 1: 1000 -51755.594 0.975 0.772
Chain 1: 1100 -39150.754 0.907 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38319.069 0.479 0.379
Chain 1: 1300 -26189.942 0.449 0.379
Chain 1: 1400 -25901.679 0.355 0.322
Chain 1: 1500 -22466.293 0.343 0.322
Chain 1: 1600 -21676.392 0.293 0.295
Chain 1: 1700 -20539.578 0.203 0.295
Chain 1: 1800 -20481.466 0.166 0.153
Chain 1: 1900 -20807.781 0.138 0.055
Chain 1: 2000 -19313.123 0.116 0.055
Chain 1: 2100 -19551.812 0.085 0.036
Chain 1: 2200 -19779.245 0.084 0.036
Chain 1: 2300 -19395.537 0.040 0.020
Chain 1: 2400 -19167.455 0.040 0.020
Chain 1: 2500 -18969.840 0.025 0.016
Chain 1: 2600 -18599.487 0.024 0.016
Chain 1: 2700 -18556.306 0.018 0.012
Chain 1: 2800 -18273.212 0.020 0.015
Chain 1: 2900 -18554.716 0.020 0.015
Chain 1: 3000 -18540.770 0.012 0.012
Chain 1: 3100 -18625.800 0.011 0.012
Chain 1: 3200 -18316.311 0.012 0.015
Chain 1: 3300 -18521.186 0.011 0.012
Chain 1: 3400 -17995.895 0.013 0.015
Chain 1: 3500 -18608.250 0.015 0.015
Chain 1: 3600 -17914.379 0.017 0.015
Chain 1: 3700 -18301.646 0.019 0.017
Chain 1: 3800 -17260.580 0.023 0.021
Chain 1: 3900 -17256.765 0.022 0.021
Chain 1: 4000 -17374.005 0.022 0.021
Chain 1: 4100 -17287.740 0.022 0.021
Chain 1: 4200 -17103.820 0.022 0.021
Chain 1: 4300 -17242.294 0.021 0.021
Chain 1: 4400 -17198.990 0.019 0.011
Chain 1: 4500 -17101.547 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001403 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.03 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13208.714 1.000 1.000
Chain 1: 200 -9917.046 0.666 1.000
Chain 1: 300 -8509.730 0.499 0.332
Chain 1: 400 -8697.802 0.380 0.332
Chain 1: 500 -8593.406 0.306 0.165
Chain 1: 600 -8416.807 0.259 0.165
Chain 1: 700 -8550.776 0.224 0.022
Chain 1: 800 -8397.541 0.198 0.022
Chain 1: 900 -8410.460 0.176 0.021
Chain 1: 1000 -8404.684 0.159 0.021
Chain 1: 1100 -8416.527 0.059 0.018
Chain 1: 1200 -8340.173 0.027 0.016
Chain 1: 1300 -8282.847 0.011 0.012
Chain 1: 1400 -8303.085 0.009 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001393 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.93 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58433.234 1.000 1.000
Chain 1: 200 -18460.375 1.583 2.165
Chain 1: 300 -9278.430 1.385 1.000
Chain 1: 400 -8425.699 1.064 1.000
Chain 1: 500 -8845.264 0.861 0.990
Chain 1: 600 -8127.680 0.732 0.990
Chain 1: 700 -8456.562 0.633 0.101
Chain 1: 800 -8600.692 0.556 0.101
Chain 1: 900 -8013.312 0.502 0.088
Chain 1: 1000 -8081.884 0.453 0.088
Chain 1: 1100 -8222.980 0.355 0.073
Chain 1: 1200 -7937.241 0.142 0.047
Chain 1: 1300 -7753.392 0.045 0.039
Chain 1: 1400 -8106.047 0.039 0.039
Chain 1: 1500 -7646.723 0.041 0.039
Chain 1: 1600 -7920.034 0.035 0.036
Chain 1: 1700 -7800.694 0.033 0.035
Chain 1: 1800 -7798.433 0.031 0.035
Chain 1: 1900 -7750.979 0.025 0.024
Chain 1: 2000 -7810.470 0.024 0.024
Chain 1: 2100 -7677.937 0.024 0.024
Chain 1: 2200 -8106.167 0.026 0.024
Chain 1: 2300 -7782.120 0.028 0.035
Chain 1: 2400 -7832.979 0.024 0.017
Chain 1: 2500 -7707.559 0.020 0.016
Chain 1: 2600 -7652.018 0.017 0.015
Chain 1: 2700 -7586.886 0.016 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003106 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.06 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87888.912 1.000 1.000
Chain 1: 200 -14409.218 3.050 5.099
Chain 1: 300 -10574.938 2.154 1.000
Chain 1: 400 -12470.242 1.654 1.000
Chain 1: 500 -8953.960 1.401 0.393
Chain 1: 600 -8809.721 1.171 0.393
Chain 1: 700 -9392.085 1.012 0.363
Chain 1: 800 -9082.560 0.890 0.363
Chain 1: 900 -9296.306 0.794 0.152
Chain 1: 1000 -8797.614 0.720 0.152
Chain 1: 1100 -8979.563 0.622 0.062 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8765.436 0.114 0.057
Chain 1: 1300 -9025.121 0.081 0.034
Chain 1: 1400 -8971.808 0.066 0.029
Chain 1: 1500 -8995.519 0.027 0.024
Chain 1: 1600 -9081.744 0.027 0.024
Chain 1: 1700 -9139.930 0.021 0.023
Chain 1: 1800 -8672.240 0.023 0.023
Chain 1: 1900 -8786.781 0.022 0.020
Chain 1: 2000 -8805.737 0.017 0.013
Chain 1: 2100 -8895.360 0.016 0.010
Chain 1: 2200 -8665.059 0.016 0.010
Chain 1: 2300 -8835.148 0.015 0.010
Chain 1: 2400 -8696.184 0.016 0.013
Chain 1: 2500 -8756.098 0.016 0.013
Chain 1: 2600 -8662.218 0.017 0.013
Chain 1: 2700 -8697.731 0.016 0.013
Chain 1: 2800 -8658.551 0.011 0.011
Chain 1: 2900 -8765.101 0.011 0.011
Chain 1: 3000 -8672.922 0.012 0.011
Chain 1: 3100 -8639.795 0.011 0.011
Chain 1: 3200 -8608.469 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003312 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.12 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8428103.357 1.000 1.000
Chain 1: 200 -1586531.629 2.656 4.312
Chain 1: 300 -890781.004 2.031 1.000
Chain 1: 400 -458263.629 1.759 1.000
Chain 1: 500 -358404.564 1.463 0.944
Chain 1: 600 -233450.609 1.309 0.944
Chain 1: 700 -119885.171 1.257 0.944
Chain 1: 800 -87211.502 1.147 0.944
Chain 1: 900 -67599.029 1.051 0.781
Chain 1: 1000 -52450.618 0.975 0.781
Chain 1: 1100 -39973.636 0.906 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39163.807 0.477 0.375
Chain 1: 1300 -27143.769 0.443 0.375
Chain 1: 1400 -26870.453 0.350 0.312
Chain 1: 1500 -23464.084 0.337 0.312
Chain 1: 1600 -22684.763 0.287 0.290
Chain 1: 1700 -21559.476 0.197 0.289
Chain 1: 1800 -21504.874 0.160 0.145
Chain 1: 1900 -21832.009 0.132 0.052
Chain 1: 2000 -20342.221 0.111 0.052
Chain 1: 2100 -20580.646 0.081 0.034
Chain 1: 2200 -20807.776 0.080 0.034
Chain 1: 2300 -20424.128 0.037 0.019
Chain 1: 2400 -20195.826 0.038 0.019
Chain 1: 2500 -19997.879 0.024 0.015
Chain 1: 2600 -19626.956 0.022 0.015
Chain 1: 2700 -19583.622 0.017 0.012
Chain 1: 2800 -19300.019 0.019 0.015
Chain 1: 2900 -19581.733 0.019 0.014
Chain 1: 3000 -19567.772 0.011 0.012
Chain 1: 3100 -19652.965 0.011 0.011
Chain 1: 3200 -19342.924 0.011 0.014
Chain 1: 3300 -19548.239 0.010 0.011
Chain 1: 3400 -19021.925 0.012 0.014
Chain 1: 3500 -19635.652 0.014 0.015
Chain 1: 3600 -18939.831 0.016 0.015
Chain 1: 3700 -19328.449 0.018 0.016
Chain 1: 3800 -18284.367 0.022 0.020
Chain 1: 3900 -18280.392 0.020 0.020
Chain 1: 4000 -18397.707 0.021 0.020
Chain 1: 4100 -18311.287 0.021 0.020
Chain 1: 4200 -18126.679 0.020 0.020
Chain 1: 4300 -18265.670 0.020 0.020
Chain 1: 4400 -18221.787 0.018 0.010
Chain 1: 4500 -18124.177 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001315 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.15 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12768.857 1.000 1.000
Chain 1: 200 -9744.369 0.655 1.000
Chain 1: 300 -8458.286 0.487 0.310
Chain 1: 400 -8633.172 0.371 0.310
Chain 1: 500 -8599.876 0.297 0.152
Chain 1: 600 -8392.068 0.252 0.152
Chain 1: 700 -8311.465 0.217 0.025
Chain 1: 800 -8350.564 0.191 0.025
Chain 1: 900 -8379.765 0.170 0.020
Chain 1: 1000 -8331.040 0.154 0.020
Chain 1: 1100 -8467.449 0.055 0.016
Chain 1: 1200 -8317.884 0.026 0.016
Chain 1: 1300 -8264.745 0.011 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001378 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.78 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -62166.967 1.000 1.000
Chain 1: 200 -18356.009 1.693 2.387
Chain 1: 300 -9095.299 1.468 1.018
Chain 1: 400 -9634.698 1.115 1.018
Chain 1: 500 -8893.965 0.909 1.000
Chain 1: 600 -8092.831 0.774 1.000
Chain 1: 700 -7788.841 0.669 0.099
Chain 1: 800 -8120.453 0.590 0.099
Chain 1: 900 -7861.624 0.528 0.083
Chain 1: 1000 -7823.687 0.476 0.083
Chain 1: 1100 -7771.348 0.377 0.056
Chain 1: 1200 -7621.919 0.140 0.041
Chain 1: 1300 -7724.776 0.040 0.039
Chain 1: 1400 -7780.841 0.035 0.033
Chain 1: 1500 -7584.735 0.029 0.026
Chain 1: 1600 -7746.305 0.021 0.021
Chain 1: 1700 -7776.351 0.018 0.020
Chain 1: 1800 -7663.556 0.015 0.015
Chain 1: 1900 -7579.493 0.013 0.013
Chain 1: 2000 -7642.960 0.013 0.013
Chain 1: 2100 -7443.799 0.015 0.015
Chain 1: 2200 -7785.981 0.018 0.015
Chain 1: 2300 -7531.417 0.020 0.021
Chain 1: 2400 -7677.379 0.021 0.021
Chain 1: 2500 -7481.343 0.021 0.021
Chain 1: 2600 -7511.413 0.019 0.019
Chain 1: 2700 -7517.050 0.019 0.019
Chain 1: 2800 -7475.912 0.018 0.019
Chain 1: 2900 -7357.279 0.018 0.019
Chain 1: 3000 -7513.895 0.020 0.021
Chain 1: 3100 -7504.159 0.017 0.019
Chain 1: 3200 -7722.148 0.016 0.019
Chain 1: 3300 -7446.955 0.016 0.019
Chain 1: 3400 -7681.967 0.017 0.021
Chain 1: 3500 -7416.404 0.018 0.021
Chain 1: 3600 -7482.809 0.018 0.021
Chain 1: 3700 -7433.998 0.019 0.021
Chain 1: 3800 -7434.886 0.019 0.021
Chain 1: 3900 -7390.531 0.018 0.021
Chain 1: 4000 -7383.668 0.016 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003031 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.31 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87180.204 1.000 1.000
Chain 1: 200 -13962.171 3.122 5.244
Chain 1: 300 -10300.266 2.200 1.000
Chain 1: 400 -11396.142 1.674 1.000
Chain 1: 500 -9280.532 1.385 0.356
Chain 1: 600 -9011.643 1.159 0.356
Chain 1: 700 -8695.426 0.999 0.228
Chain 1: 800 -9036.163 0.878 0.228
Chain 1: 900 -9125.203 0.782 0.096
Chain 1: 1000 -8886.991 0.706 0.096
Chain 1: 1100 -8992.296 0.608 0.038 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8746.567 0.086 0.036
Chain 1: 1300 -8975.577 0.053 0.030
Chain 1: 1400 -8987.726 0.044 0.028
Chain 1: 1500 -8856.400 0.022 0.027
Chain 1: 1600 -8964.370 0.020 0.026
Chain 1: 1700 -9048.998 0.018 0.015
Chain 1: 1800 -8627.587 0.019 0.015
Chain 1: 1900 -8727.570 0.019 0.015
Chain 1: 2000 -8701.768 0.017 0.012
Chain 1: 2100 -8826.436 0.017 0.014
Chain 1: 2200 -8634.084 0.016 0.014
Chain 1: 2300 -8722.164 0.015 0.012
Chain 1: 2400 -8791.407 0.015 0.012
Chain 1: 2500 -8737.560 0.015 0.011
Chain 1: 2600 -8738.348 0.013 0.010
Chain 1: 2700 -8655.294 0.013 0.010
Chain 1: 2800 -8616.069 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003121 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.21 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8417680.231 1.000 1.000
Chain 1: 200 -1588092.859 2.650 4.300
Chain 1: 300 -893266.725 2.026 1.000
Chain 1: 400 -459097.666 1.756 1.000
Chain 1: 500 -359067.150 1.461 0.946
Chain 1: 600 -233822.306 1.306 0.946
Chain 1: 700 -119851.450 1.256 0.946
Chain 1: 800 -87015.971 1.146 0.946
Chain 1: 900 -67326.588 1.051 0.778
Chain 1: 1000 -52100.862 0.975 0.778
Chain 1: 1100 -39558.949 0.907 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38732.970 0.479 0.377
Chain 1: 1300 -26670.600 0.446 0.377
Chain 1: 1400 -26388.032 0.353 0.317
Chain 1: 1500 -22970.466 0.340 0.317
Chain 1: 1600 -22185.700 0.290 0.292
Chain 1: 1700 -21057.240 0.200 0.292
Chain 1: 1800 -21000.817 0.163 0.149
Chain 1: 1900 -21327.028 0.135 0.054
Chain 1: 2000 -19836.977 0.113 0.054
Chain 1: 2100 -20075.387 0.083 0.035
Chain 1: 2200 -20302.090 0.082 0.035
Chain 1: 2300 -19919.085 0.038 0.019
Chain 1: 2400 -19691.131 0.038 0.019
Chain 1: 2500 -19493.226 0.025 0.015
Chain 1: 2600 -19123.254 0.023 0.015
Chain 1: 2700 -19080.190 0.018 0.012
Chain 1: 2800 -18797.045 0.019 0.015
Chain 1: 2900 -19078.379 0.019 0.015
Chain 1: 3000 -19064.516 0.012 0.012
Chain 1: 3100 -19149.518 0.011 0.012
Chain 1: 3200 -18840.127 0.011 0.015
Chain 1: 3300 -19044.928 0.011 0.012
Chain 1: 3400 -18519.740 0.012 0.015
Chain 1: 3500 -19131.758 0.014 0.015
Chain 1: 3600 -18438.306 0.016 0.015
Chain 1: 3700 -18825.223 0.018 0.016
Chain 1: 3800 -17784.683 0.022 0.021
Chain 1: 3900 -17780.837 0.021 0.021
Chain 1: 4000 -17898.136 0.022 0.021
Chain 1: 4100 -17811.863 0.022 0.021
Chain 1: 4200 -17628.064 0.021 0.021
Chain 1: 4300 -17766.479 0.021 0.021
Chain 1: 4400 -17723.275 0.018 0.010
Chain 1: 4500 -17625.798 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.0013 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12596.894 1.000 1.000
Chain 1: 200 -9339.588 0.674 1.000
Chain 1: 300 -8257.683 0.493 0.349
Chain 1: 400 -8348.925 0.373 0.349
Chain 1: 500 -8236.684 0.301 0.131
Chain 1: 600 -8177.913 0.252 0.131
Chain 1: 700 -8076.693 0.218 0.014
Chain 1: 800 -8082.355 0.191 0.014
Chain 1: 900 -8109.249 0.170 0.013
Chain 1: 1000 -8176.930 0.154 0.013
Chain 1: 1100 -8223.894 0.054 0.011
Chain 1: 1200 -8084.475 0.021 0.011
Chain 1: 1300 -8059.239 0.008 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001389 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.89 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61919.905 1.000 1.000
Chain 1: 200 -18058.204 1.714 2.429
Chain 1: 300 -8980.019 1.480 1.011
Chain 1: 400 -9601.762 1.126 1.011
Chain 1: 500 -8487.746 0.927 1.000
Chain 1: 600 -8928.618 0.781 1.000
Chain 1: 700 -7646.660 0.693 0.168
Chain 1: 800 -8192.168 0.615 0.168
Chain 1: 900 -7834.170 0.552 0.131
Chain 1: 1000 -7859.890 0.497 0.131
Chain 1: 1100 -7778.211 0.398 0.067
Chain 1: 1200 -7729.151 0.156 0.065
Chain 1: 1300 -7794.205 0.055 0.049
Chain 1: 1400 -7943.723 0.051 0.046
Chain 1: 1500 -7608.520 0.042 0.044
Chain 1: 1600 -7611.789 0.037 0.019
Chain 1: 1700 -7563.180 0.021 0.011
Chain 1: 1800 -7596.288 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004336 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 43.36 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86558.985 1.000 1.000
Chain 1: 200 -13757.788 3.146 5.292
Chain 1: 300 -10112.766 2.217 1.000
Chain 1: 400 -11076.411 1.685 1.000
Chain 1: 500 -9084.465 1.392 0.360
Chain 1: 600 -8536.744 1.170 0.360
Chain 1: 700 -8655.776 1.005 0.219
Chain 1: 800 -8895.516 0.883 0.219
Chain 1: 900 -8949.949 0.785 0.087
Chain 1: 1000 -8590.743 0.711 0.087
Chain 1: 1100 -8868.075 0.614 0.064 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8427.011 0.090 0.052
Chain 1: 1300 -8773.353 0.058 0.042
Chain 1: 1400 -8759.194 0.050 0.039
Chain 1: 1500 -8654.545 0.029 0.031
Chain 1: 1600 -8764.674 0.024 0.027
Chain 1: 1700 -8832.836 0.023 0.027
Chain 1: 1800 -8411.216 0.026 0.031
Chain 1: 1900 -8510.854 0.026 0.031
Chain 1: 2000 -8485.671 0.022 0.013
Chain 1: 2100 -8611.300 0.021 0.013
Chain 1: 2200 -8414.189 0.018 0.013
Chain 1: 2300 -8506.072 0.015 0.012
Chain 1: 2400 -8574.828 0.015 0.012
Chain 1: 2500 -8521.099 0.015 0.012
Chain 1: 2600 -8522.474 0.014 0.011
Chain 1: 2700 -8439.181 0.014 0.011
Chain 1: 2800 -8399.026 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003247 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.47 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8407830.237 1.000 1.000
Chain 1: 200 -1584473.646 2.653 4.306
Chain 1: 300 -890457.027 2.029 1.000
Chain 1: 400 -457739.691 1.758 1.000
Chain 1: 500 -357966.376 1.462 0.945
Chain 1: 600 -233062.240 1.308 0.945
Chain 1: 700 -119370.007 1.257 0.945
Chain 1: 800 -86654.164 1.147 0.945
Chain 1: 900 -67014.418 1.052 0.779
Chain 1: 1000 -51826.222 0.976 0.779
Chain 1: 1100 -39317.847 0.908 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38497.789 0.479 0.378
Chain 1: 1300 -26455.631 0.447 0.378
Chain 1: 1400 -26177.471 0.354 0.318
Chain 1: 1500 -22765.805 0.341 0.318
Chain 1: 1600 -21983.674 0.291 0.293
Chain 1: 1700 -20856.602 0.201 0.293
Chain 1: 1800 -20800.983 0.163 0.150
Chain 1: 1900 -21127.312 0.136 0.054
Chain 1: 2000 -19638.195 0.114 0.054
Chain 1: 2100 -19876.459 0.083 0.036
Chain 1: 2200 -20103.218 0.082 0.036
Chain 1: 2300 -19720.132 0.039 0.019
Chain 1: 2400 -19492.137 0.039 0.019
Chain 1: 2500 -19294.410 0.025 0.015
Chain 1: 2600 -18924.230 0.023 0.015
Chain 1: 2700 -18881.129 0.018 0.012
Chain 1: 2800 -18598.007 0.019 0.015
Chain 1: 2900 -18879.344 0.019 0.015
Chain 1: 3000 -18865.457 0.012 0.012
Chain 1: 3100 -18950.476 0.011 0.012
Chain 1: 3200 -18641.025 0.012 0.015
Chain 1: 3300 -18845.892 0.011 0.012
Chain 1: 3400 -18320.650 0.012 0.015
Chain 1: 3500 -18932.793 0.015 0.015
Chain 1: 3600 -18239.118 0.016 0.015
Chain 1: 3700 -18626.179 0.018 0.017
Chain 1: 3800 -17585.398 0.023 0.021
Chain 1: 3900 -17581.552 0.021 0.021
Chain 1: 4000 -17698.831 0.022 0.021
Chain 1: 4100 -17612.569 0.022 0.021
Chain 1: 4200 -17428.714 0.021 0.021
Chain 1: 4300 -17567.144 0.021 0.021
Chain 1: 4400 -17523.863 0.018 0.011
Chain 1: 4500 -17426.404 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00127 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.7 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13053.191 1.000 1.000
Chain 1: 200 -10017.025 0.652 1.000
Chain 1: 300 -8528.246 0.493 0.303
Chain 1: 400 -8756.402 0.376 0.303
Chain 1: 500 -8600.091 0.304 0.175
Chain 1: 600 -8454.398 0.257 0.175
Chain 1: 700 -8576.759 0.222 0.026
Chain 1: 800 -8387.608 0.197 0.026
Chain 1: 900 -8460.923 0.176 0.023
Chain 1: 1000 -8389.530 0.159 0.023
Chain 1: 1100 -8434.361 0.060 0.018
Chain 1: 1200 -8361.997 0.030 0.017
Chain 1: 1300 -8296.295 0.014 0.014
Chain 1: 1400 -8328.739 0.012 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001459 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.59 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -59110.469 1.000 1.000
Chain 1: 200 -18475.162 1.600 2.199
Chain 1: 300 -9299.129 1.395 1.000
Chain 1: 400 -8545.532 1.069 1.000
Chain 1: 500 -8337.394 0.860 0.987
Chain 1: 600 -9465.264 0.736 0.987
Chain 1: 700 -7932.106 0.659 0.193
Chain 1: 800 -8526.685 0.585 0.193
Chain 1: 900 -8390.231 0.522 0.119
Chain 1: 1000 -8183.943 0.472 0.119
Chain 1: 1100 -8058.403 0.374 0.088
Chain 1: 1200 -7993.372 0.155 0.070
Chain 1: 1300 -7938.699 0.057 0.025
Chain 1: 1400 -8050.705 0.049 0.025
Chain 1: 1500 -7662.513 0.052 0.025
Chain 1: 1600 -7765.505 0.041 0.016
Chain 1: 1700 -7772.887 0.022 0.016
Chain 1: 1800 -7690.332 0.016 0.014
Chain 1: 1900 -7743.302 0.015 0.013
Chain 1: 2000 -7814.037 0.014 0.011
Chain 1: 2100 -7707.981 0.013 0.011
Chain 1: 2200 -7951.466 0.016 0.013
Chain 1: 2300 -7738.280 0.018 0.014
Chain 1: 2400 -7789.138 0.017 0.013
Chain 1: 2500 -7664.932 0.014 0.013
Chain 1: 2600 -7680.780 0.012 0.011
Chain 1: 2700 -7653.059 0.013 0.011
Chain 1: 2800 -7785.657 0.013 0.014
Chain 1: 2900 -7521.819 0.016 0.016
Chain 1: 3000 -7680.725 0.017 0.017
Chain 1: 3100 -7676.814 0.016 0.017
Chain 1: 3200 -7868.749 0.015 0.017
Chain 1: 3300 -7583.388 0.016 0.017
Chain 1: 3400 -7802.728 0.019 0.021
Chain 1: 3500 -7583.401 0.020 0.024
Chain 1: 3600 -7645.594 0.020 0.024
Chain 1: 3700 -7602.742 0.021 0.024
Chain 1: 3800 -7573.322 0.019 0.024
Chain 1: 3900 -7545.815 0.016 0.021
Chain 1: 4000 -7541.698 0.014 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003542 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.42 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86910.449 1.000 1.000
Chain 1: 200 -14283.225 3.042 5.085
Chain 1: 300 -10503.521 2.148 1.000
Chain 1: 400 -12199.865 1.646 1.000
Chain 1: 500 -9282.615 1.380 0.360
Chain 1: 600 -9749.255 1.158 0.360
Chain 1: 700 -9143.858 1.002 0.314
Chain 1: 800 -8733.137 0.882 0.314
Chain 1: 900 -8835.519 0.786 0.139
Chain 1: 1000 -9601.817 0.715 0.139
Chain 1: 1100 -8995.579 0.622 0.080 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -9411.384 0.118 0.067
Chain 1: 1300 -8790.078 0.089 0.067
Chain 1: 1400 -8976.636 0.077 0.066
Chain 1: 1500 -8931.625 0.046 0.048
Chain 1: 1600 -8889.472 0.042 0.047
Chain 1: 1700 -8765.581 0.037 0.044
Chain 1: 1800 -8819.440 0.032 0.021
Chain 1: 1900 -8758.393 0.032 0.021
Chain 1: 2000 -8768.678 0.024 0.014
Chain 1: 2100 -8828.958 0.018 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003374 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.74 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8408977.789 1.000 1.000
Chain 1: 200 -1583301.238 2.656 4.311
Chain 1: 300 -893155.306 2.028 1.000
Chain 1: 400 -459953.808 1.756 1.000
Chain 1: 500 -360383.438 1.460 0.942
Chain 1: 600 -234952.732 1.306 0.942
Chain 1: 700 -120603.568 1.255 0.942
Chain 1: 800 -87673.038 1.145 0.942
Chain 1: 900 -67899.779 1.050 0.773
Chain 1: 1000 -52619.812 0.974 0.773
Chain 1: 1100 -40023.180 0.906 0.534 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39195.227 0.477 0.376
Chain 1: 1300 -27067.107 0.444 0.376
Chain 1: 1400 -26780.623 0.351 0.315
Chain 1: 1500 -23346.164 0.338 0.315
Chain 1: 1600 -22557.259 0.288 0.291
Chain 1: 1700 -21420.362 0.199 0.290
Chain 1: 1800 -21362.421 0.161 0.147
Chain 1: 1900 -21689.235 0.134 0.053
Chain 1: 2000 -20193.698 0.112 0.053
Chain 1: 2100 -20432.342 0.082 0.035
Chain 1: 2200 -20660.204 0.081 0.035
Chain 1: 2300 -20276.025 0.038 0.019
Chain 1: 2400 -20047.779 0.038 0.019
Chain 1: 2500 -19850.123 0.024 0.015
Chain 1: 2600 -19479.173 0.023 0.015
Chain 1: 2700 -19435.826 0.018 0.012
Chain 1: 2800 -19152.504 0.019 0.015
Chain 1: 2900 -19434.241 0.019 0.014
Chain 1: 3000 -19420.218 0.011 0.012
Chain 1: 3100 -19505.355 0.011 0.011
Chain 1: 3200 -19195.443 0.011 0.014
Chain 1: 3300 -19400.649 0.010 0.011
Chain 1: 3400 -18874.658 0.012 0.014
Chain 1: 3500 -19487.971 0.014 0.015
Chain 1: 3600 -18792.834 0.016 0.015
Chain 1: 3700 -19181.075 0.018 0.016
Chain 1: 3800 -18137.956 0.022 0.020
Chain 1: 3900 -18134.086 0.021 0.020
Chain 1: 4000 -18251.343 0.021 0.020
Chain 1: 4100 -18164.972 0.021 0.020
Chain 1: 4200 -17980.602 0.021 0.020
Chain 1: 4300 -18119.381 0.020 0.020
Chain 1: 4400 -18075.718 0.018 0.010
Chain 1: 4500 -17978.195 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.0013 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48821.993 1.000 1.000
Chain 1: 200 -23440.720 1.041 1.083
Chain 1: 300 -13120.851 0.956 1.000
Chain 1: 400 -14995.613 0.749 1.000
Chain 1: 500 -17625.246 0.629 0.787
Chain 1: 600 -20380.262 0.546 0.787
Chain 1: 700 -13362.199 0.543 0.525
Chain 1: 800 -13172.062 0.477 0.525
Chain 1: 900 -16157.841 0.445 0.185
Chain 1: 1000 -31506.045 0.449 0.487
Chain 1: 1100 -17457.582 0.430 0.487
Chain 1: 1200 -20308.054 0.335 0.185
Chain 1: 1300 -11955.485 0.326 0.185
Chain 1: 1400 -10311.401 0.330 0.185
Chain 1: 1500 -10185.485 0.316 0.185
Chain 1: 1600 -34364.412 0.373 0.487
Chain 1: 1700 -13123.303 0.482 0.487
Chain 1: 1800 -9594.494 0.518 0.487 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1900 -17652.421 0.545 0.487 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2000 -17794.578 0.497 0.456
Chain 1: 2100 -10198.484 0.491 0.456
Chain 1: 2200 -16193.419 0.514 0.456 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2300 -9505.896 0.514 0.456 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2400 -10365.110 0.507 0.456 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2500 -10391.459 0.506 0.456 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2600 -9818.570 0.441 0.370
Chain 1: 2700 -14812.284 0.313 0.368
Chain 1: 2800 -10218.533 0.321 0.370
Chain 1: 2900 -9688.900 0.281 0.337
Chain 1: 3000 -12621.106 0.304 0.337
Chain 1: 3100 -16352.096 0.252 0.232
Chain 1: 3200 -9336.417 0.290 0.232
Chain 1: 3300 -15550.016 0.260 0.232
Chain 1: 3400 -9858.529 0.309 0.337
Chain 1: 3500 -13236.615 0.334 0.337
Chain 1: 3600 -14719.873 0.339 0.337
Chain 1: 3700 -9901.834 0.354 0.400
Chain 1: 3800 -10266.612 0.312 0.255
Chain 1: 3900 -9552.557 0.314 0.255
Chain 1: 4000 -9270.262 0.294 0.255
Chain 1: 4100 -8929.058 0.275 0.255
Chain 1: 4200 -11880.910 0.225 0.248
Chain 1: 4300 -8917.111 0.218 0.248
Chain 1: 4400 -12009.255 0.186 0.248
Chain 1: 4500 -12487.340 0.164 0.101
Chain 1: 4600 -14072.683 0.165 0.113
Chain 1: 4700 -9238.573 0.169 0.113
Chain 1: 4800 -13429.328 0.197 0.248
Chain 1: 4900 -9472.351 0.231 0.257
Chain 1: 5000 -13560.757 0.258 0.301
Chain 1: 5100 -8522.805 0.313 0.312
Chain 1: 5200 -9384.644 0.298 0.312
Chain 1: 5300 -15197.431 0.303 0.312
Chain 1: 5400 -11279.885 0.312 0.347
Chain 1: 5500 -9250.258 0.330 0.347
Chain 1: 5600 -9282.439 0.319 0.347
Chain 1: 5700 -11531.677 0.286 0.312
Chain 1: 5800 -9803.689 0.273 0.301
Chain 1: 5900 -9050.362 0.239 0.219
Chain 1: 6000 -9067.112 0.209 0.195
Chain 1: 6100 -8709.004 0.154 0.176
Chain 1: 6200 -8486.354 0.148 0.176
Chain 1: 6300 -8832.087 0.113 0.083
Chain 1: 6400 -8805.880 0.079 0.041
Chain 1: 6500 -8984.008 0.059 0.039
Chain 1: 6600 -12551.542 0.087 0.041
Chain 1: 6700 -8330.534 0.118 0.041
Chain 1: 6800 -9151.132 0.109 0.041
Chain 1: 6900 -11274.117 0.120 0.041
Chain 1: 7000 -8882.638 0.147 0.090
Chain 1: 7100 -8597.717 0.146 0.090
Chain 1: 7200 -9413.253 0.152 0.090
Chain 1: 7300 -11213.249 0.164 0.161
Chain 1: 7400 -10982.396 0.166 0.161
Chain 1: 7500 -8662.254 0.191 0.188
Chain 1: 7600 -12164.197 0.191 0.188
Chain 1: 7700 -8273.345 0.187 0.188
Chain 1: 7800 -13257.981 0.216 0.268
Chain 1: 7900 -8742.043 0.249 0.269
Chain 1: 8000 -8395.984 0.226 0.268
Chain 1: 8100 -8370.556 0.223 0.268
Chain 1: 8200 -9854.803 0.229 0.268
Chain 1: 8300 -8217.329 0.233 0.268
Chain 1: 8400 -8605.640 0.236 0.268
Chain 1: 8500 -8643.252 0.209 0.199
Chain 1: 8600 -10462.704 0.198 0.174
Chain 1: 8700 -8708.511 0.171 0.174
Chain 1: 8800 -8392.194 0.137 0.151
Chain 1: 8900 -8472.858 0.087 0.045
Chain 1: 9000 -10756.896 0.104 0.151
Chain 1: 9100 -8640.003 0.128 0.174
Chain 1: 9200 -11418.725 0.137 0.199
Chain 1: 9300 -10465.311 0.126 0.174
Chain 1: 9400 -11447.593 0.130 0.174
Chain 1: 9500 -8907.696 0.159 0.201
Chain 1: 9600 -8395.609 0.147 0.201
Chain 1: 9700 -8551.744 0.129 0.091
Chain 1: 9800 -11802.214 0.153 0.212
Chain 1: 9900 -11247.257 0.157 0.212
Chain 1: 10000 -8315.783 0.171 0.243
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001425 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.25 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57165.938 1.000 1.000
Chain 1: 200 -17446.334 1.638 2.277
Chain 1: 300 -8724.925 1.425 1.000
Chain 1: 400 -8335.733 1.081 1.000
Chain 1: 500 -8428.198 0.867 1.000
Chain 1: 600 -8438.910 0.723 1.000
Chain 1: 700 -7767.524 0.632 0.086
Chain 1: 800 -8272.360 0.560 0.086
Chain 1: 900 -7968.862 0.502 0.061
Chain 1: 1000 -7767.639 0.455 0.061
Chain 1: 1100 -7754.136 0.355 0.047
Chain 1: 1200 -8028.330 0.131 0.038
Chain 1: 1300 -7610.681 0.036 0.038
Chain 1: 1400 -7855.341 0.035 0.034
Chain 1: 1500 -7621.492 0.037 0.034
Chain 1: 1600 -7669.766 0.037 0.034
Chain 1: 1700 -7545.752 0.030 0.031
Chain 1: 1800 -7621.381 0.025 0.031
Chain 1: 1900 -7598.319 0.021 0.026
Chain 1: 2000 -7618.784 0.019 0.016
Chain 1: 2100 -7602.512 0.019 0.016
Chain 1: 2200 -7718.429 0.017 0.015
Chain 1: 2300 -7614.685 0.013 0.014
Chain 1: 2400 -7620.882 0.010 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003099 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.99 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85944.488 1.000 1.000
Chain 1: 200 -13522.589 3.178 5.356
Chain 1: 300 -9920.537 2.240 1.000
Chain 1: 400 -10928.603 1.703 1.000
Chain 1: 500 -8810.317 1.410 0.363
Chain 1: 600 -8401.455 1.183 0.363
Chain 1: 700 -8508.642 1.016 0.240
Chain 1: 800 -8758.615 0.893 0.240
Chain 1: 900 -8690.725 0.794 0.092
Chain 1: 1000 -8613.012 0.716 0.092
Chain 1: 1100 -8760.485 0.617 0.049 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8306.447 0.087 0.049
Chain 1: 1300 -8620.668 0.055 0.036
Chain 1: 1400 -8621.267 0.046 0.029
Chain 1: 1500 -8495.272 0.023 0.017
Chain 1: 1600 -8602.322 0.019 0.015
Chain 1: 1700 -8688.540 0.019 0.015
Chain 1: 1800 -8281.245 0.021 0.015
Chain 1: 1900 -8377.733 0.021 0.015
Chain 1: 2000 -8349.969 0.021 0.015
Chain 1: 2100 -8470.774 0.021 0.014
Chain 1: 2200 -8290.390 0.017 0.014
Chain 1: 2300 -8416.973 0.015 0.014
Chain 1: 2400 -8427.551 0.015 0.014
Chain 1: 2500 -8389.592 0.014 0.012
Chain 1: 2600 -8388.434 0.013 0.012
Chain 1: 2700 -8303.278 0.013 0.012
Chain 1: 2800 -8268.149 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00344 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.4 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8387627.417 1.000 1.000
Chain 1: 200 -1582486.425 2.650 4.300
Chain 1: 300 -890471.322 2.026 1.000
Chain 1: 400 -456986.478 1.756 1.000
Chain 1: 500 -357641.698 1.461 0.949
Chain 1: 600 -232756.103 1.307 0.949
Chain 1: 700 -119167.728 1.256 0.949
Chain 1: 800 -86404.865 1.147 0.949
Chain 1: 900 -66771.715 1.052 0.777
Chain 1: 1000 -51582.501 0.976 0.777
Chain 1: 1100 -39068.030 0.908 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38246.324 0.480 0.379
Chain 1: 1300 -26209.386 0.448 0.379
Chain 1: 1400 -25928.531 0.355 0.320
Chain 1: 1500 -22517.097 0.342 0.320
Chain 1: 1600 -21734.068 0.292 0.294
Chain 1: 1700 -20608.502 0.202 0.294
Chain 1: 1800 -20552.888 0.165 0.152
Chain 1: 1900 -20878.916 0.137 0.055
Chain 1: 2000 -19390.709 0.115 0.055
Chain 1: 2100 -19629.060 0.084 0.036
Chain 1: 2200 -19855.349 0.083 0.036
Chain 1: 2300 -19472.743 0.039 0.020
Chain 1: 2400 -19244.900 0.039 0.020
Chain 1: 2500 -19046.871 0.025 0.016
Chain 1: 2600 -18677.244 0.023 0.016
Chain 1: 2700 -18634.313 0.018 0.012
Chain 1: 2800 -18351.200 0.020 0.015
Chain 1: 2900 -18632.367 0.019 0.015
Chain 1: 3000 -18618.600 0.012 0.012
Chain 1: 3100 -18703.550 0.011 0.012
Chain 1: 3200 -18394.336 0.012 0.015
Chain 1: 3300 -18599.009 0.011 0.012
Chain 1: 3400 -18074.093 0.013 0.015
Chain 1: 3500 -18685.712 0.015 0.015
Chain 1: 3600 -17992.767 0.017 0.015
Chain 1: 3700 -18379.253 0.018 0.017
Chain 1: 3800 -17339.547 0.023 0.021
Chain 1: 3900 -17335.724 0.021 0.021
Chain 1: 4000 -17453.012 0.022 0.021
Chain 1: 4100 -17366.773 0.022 0.021
Chain 1: 4200 -17183.210 0.021 0.021
Chain 1: 4300 -17321.498 0.021 0.021
Chain 1: 4400 -17278.424 0.019 0.011
Chain 1: 4500 -17180.986 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001276 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.76 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12245.492 1.000 1.000
Chain 1: 200 -9155.835 0.669 1.000
Chain 1: 300 -8137.681 0.488 0.337
Chain 1: 400 -8149.436 0.366 0.337
Chain 1: 500 -8004.904 0.296 0.125
Chain 1: 600 -7933.150 0.249 0.125
Chain 1: 700 -7852.592 0.214 0.018
Chain 1: 800 -7883.833 0.188 0.018
Chain 1: 900 -8022.098 0.169 0.017
Chain 1: 1000 -7893.317 0.154 0.017
Chain 1: 1100 -7933.676 0.054 0.016
Chain 1: 1200 -7886.873 0.021 0.010
Chain 1: 1300 -7831.812 0.009 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001386 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.86 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56800.933 1.000 1.000
Chain 1: 200 -17271.657 1.644 2.289
Chain 1: 300 -8649.368 1.429 1.000
Chain 1: 400 -8246.123 1.084 1.000
Chain 1: 500 -8348.693 0.869 0.997
Chain 1: 600 -8601.820 0.729 0.997
Chain 1: 700 -8185.759 0.632 0.051
Chain 1: 800 -8082.998 0.555 0.051
Chain 1: 900 -7908.151 0.496 0.049
Chain 1: 1000 -7816.244 0.447 0.049
Chain 1: 1100 -7720.044 0.349 0.029
Chain 1: 1200 -7589.023 0.121 0.022
Chain 1: 1300 -7540.661 0.022 0.017
Chain 1: 1400 -7853.041 0.022 0.017
Chain 1: 1500 -7586.868 0.024 0.022
Chain 1: 1600 -7771.084 0.023 0.022
Chain 1: 1700 -7500.758 0.022 0.022
Chain 1: 1800 -7538.689 0.021 0.022
Chain 1: 1900 -7597.026 0.020 0.017
Chain 1: 2000 -7595.957 0.018 0.017
Chain 1: 2100 -7553.042 0.018 0.017
Chain 1: 2200 -7665.477 0.017 0.015
Chain 1: 2300 -7571.351 0.018 0.015
Chain 1: 2400 -7609.358 0.015 0.012
Chain 1: 2500 -7529.523 0.012 0.011
Chain 1: 2600 -7500.128 0.010 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002933 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.33 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86000.460 1.000 1.000
Chain 1: 200 -13322.469 3.228 5.455
Chain 1: 300 -9752.230 2.274 1.000
Chain 1: 400 -10507.221 1.723 1.000
Chain 1: 500 -8680.275 1.421 0.366
Chain 1: 600 -8304.878 1.191 0.366
Chain 1: 700 -8648.137 1.027 0.210
Chain 1: 800 -9138.563 0.905 0.210
Chain 1: 900 -8565.107 0.812 0.072
Chain 1: 1000 -8363.515 0.733 0.072
Chain 1: 1100 -8635.996 0.636 0.067 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8173.814 0.097 0.057
Chain 1: 1300 -8503.250 0.064 0.054
Chain 1: 1400 -8482.788 0.057 0.045
Chain 1: 1500 -8367.361 0.037 0.040
Chain 1: 1600 -8474.671 0.034 0.039
Chain 1: 1700 -8551.648 0.031 0.032
Chain 1: 1800 -8157.996 0.030 0.032
Chain 1: 1900 -8260.497 0.025 0.024
Chain 1: 2000 -8230.832 0.023 0.014
Chain 1: 2100 -8355.663 0.021 0.014
Chain 1: 2200 -8140.347 0.018 0.014
Chain 1: 2300 -8289.203 0.016 0.014
Chain 1: 2400 -8304.278 0.016 0.014
Chain 1: 2500 -8271.738 0.015 0.013
Chain 1: 2600 -8273.859 0.014 0.012
Chain 1: 2700 -8180.507 0.014 0.012
Chain 1: 2800 -8152.893 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003484 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.84 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8415748.419 1.000 1.000
Chain 1: 200 -1589117.167 2.648 4.296
Chain 1: 300 -892269.924 2.026 1.000
Chain 1: 400 -457970.370 1.756 1.000
Chain 1: 500 -357777.430 1.461 0.948
Chain 1: 600 -232625.521 1.307 0.948
Chain 1: 700 -118924.264 1.257 0.948
Chain 1: 800 -86135.956 1.147 0.948
Chain 1: 900 -66501.520 1.053 0.781
Chain 1: 1000 -51315.137 0.977 0.781
Chain 1: 1100 -38810.480 0.909 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37985.828 0.482 0.381
Chain 1: 1300 -25973.280 0.450 0.381
Chain 1: 1400 -25692.894 0.356 0.322
Chain 1: 1500 -22288.279 0.344 0.322
Chain 1: 1600 -21506.354 0.293 0.296
Chain 1: 1700 -20384.673 0.203 0.295
Chain 1: 1800 -20329.560 0.166 0.153
Chain 1: 1900 -20655.214 0.138 0.055
Chain 1: 2000 -19169.483 0.116 0.055
Chain 1: 2100 -19407.714 0.085 0.036
Chain 1: 2200 -19633.436 0.084 0.036
Chain 1: 2300 -19251.415 0.039 0.020
Chain 1: 2400 -19023.713 0.040 0.020
Chain 1: 2500 -18825.511 0.025 0.016
Chain 1: 2600 -18456.341 0.024 0.016
Chain 1: 2700 -18413.509 0.018 0.012
Chain 1: 2800 -18130.441 0.020 0.016
Chain 1: 2900 -18411.468 0.020 0.015
Chain 1: 3000 -18397.773 0.012 0.012
Chain 1: 3100 -18482.654 0.011 0.012
Chain 1: 3200 -18173.657 0.012 0.015
Chain 1: 3300 -18378.140 0.011 0.012
Chain 1: 3400 -17853.543 0.013 0.015
Chain 1: 3500 -18464.573 0.015 0.016
Chain 1: 3600 -17772.406 0.017 0.016
Chain 1: 3700 -18158.295 0.019 0.017
Chain 1: 3800 -17119.685 0.023 0.021
Chain 1: 3900 -17115.854 0.022 0.021
Chain 1: 4000 -17233.197 0.022 0.021
Chain 1: 4100 -17146.980 0.022 0.021
Chain 1: 4200 -16963.638 0.022 0.021
Chain 1: 4300 -17101.780 0.021 0.021
Chain 1: 4400 -17058.906 0.019 0.011
Chain 1: 4500 -16961.479 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001348 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.48 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48380.090 1.000 1.000
Chain 1: 200 -18866.266 1.282 1.564
Chain 1: 300 -20867.626 0.887 1.000
Chain 1: 400 -17997.377 0.705 1.000
Chain 1: 500 -14775.454 0.608 0.218
Chain 1: 600 -25840.458 0.578 0.428
Chain 1: 700 -14196.322 0.612 0.428
Chain 1: 800 -14711.732 0.540 0.428
Chain 1: 900 -10667.220 0.522 0.379
Chain 1: 1000 -10890.797 0.472 0.379
Chain 1: 1100 -11085.541 0.374 0.218
Chain 1: 1200 -9677.287 0.232 0.159
Chain 1: 1300 -11417.255 0.238 0.159
Chain 1: 1400 -11458.946 0.222 0.152
Chain 1: 1500 -14683.493 0.222 0.152
Chain 1: 1600 -10779.217 0.216 0.152
Chain 1: 1700 -9995.454 0.141 0.146
Chain 1: 1800 -9071.361 0.148 0.146
Chain 1: 1900 -10456.999 0.123 0.133
Chain 1: 2000 -11871.565 0.133 0.133
Chain 1: 2100 -17282.294 0.163 0.146
Chain 1: 2200 -10237.551 0.217 0.152
Chain 1: 2300 -8956.205 0.216 0.143
Chain 1: 2400 -10407.325 0.230 0.143
Chain 1: 2500 -12431.474 0.224 0.143
Chain 1: 2600 -8953.444 0.227 0.143
Chain 1: 2700 -8657.203 0.222 0.143
Chain 1: 2800 -8907.118 0.215 0.143
Chain 1: 2900 -9196.116 0.205 0.143
Chain 1: 3000 -8733.210 0.198 0.143
Chain 1: 3100 -8487.915 0.170 0.139
Chain 1: 3200 -13768.957 0.139 0.139
Chain 1: 3300 -9063.953 0.177 0.139
Chain 1: 3400 -14566.093 0.201 0.163
Chain 1: 3500 -8896.908 0.248 0.378
Chain 1: 3600 -9358.405 0.214 0.053
Chain 1: 3700 -8851.359 0.217 0.057
Chain 1: 3800 -9819.053 0.224 0.099
Chain 1: 3900 -8915.070 0.231 0.101
Chain 1: 4000 -8852.836 0.226 0.101
Chain 1: 4100 -9371.352 0.229 0.101
Chain 1: 4200 -10652.883 0.202 0.101
Chain 1: 4300 -15564.172 0.182 0.101
Chain 1: 4400 -9553.982 0.207 0.101
Chain 1: 4500 -8675.338 0.154 0.101
Chain 1: 4600 -12780.717 0.181 0.101
Chain 1: 4700 -9894.783 0.204 0.120
Chain 1: 4800 -8598.283 0.209 0.151
Chain 1: 4900 -8386.717 0.202 0.151
Chain 1: 5000 -15876.582 0.248 0.292
Chain 1: 5100 -8970.977 0.320 0.316
Chain 1: 5200 -13339.893 0.340 0.321
Chain 1: 5300 -15208.110 0.321 0.321
Chain 1: 5400 -14577.746 0.263 0.292
Chain 1: 5500 -10769.887 0.288 0.321
Chain 1: 5600 -9772.402 0.266 0.292
Chain 1: 5700 -10907.752 0.247 0.151
Chain 1: 5800 -8664.592 0.258 0.259
Chain 1: 5900 -9059.303 0.260 0.259
Chain 1: 6000 -8240.848 0.222 0.123
Chain 1: 6100 -10909.653 0.170 0.123
Chain 1: 6200 -8421.246 0.167 0.123
Chain 1: 6300 -8190.805 0.157 0.104
Chain 1: 6400 -13049.361 0.190 0.245
Chain 1: 6500 -10554.600 0.178 0.236
Chain 1: 6600 -9950.988 0.174 0.236
Chain 1: 6700 -9229.914 0.172 0.236
Chain 1: 6800 -12618.571 0.173 0.236
Chain 1: 6900 -12469.218 0.170 0.236
Chain 1: 7000 -8410.167 0.208 0.245
Chain 1: 7100 -8164.502 0.186 0.236
Chain 1: 7200 -10112.910 0.176 0.193
Chain 1: 7300 -8169.621 0.197 0.236
Chain 1: 7400 -8477.475 0.164 0.193
Chain 1: 7500 -7945.457 0.147 0.078
Chain 1: 7600 -8372.126 0.146 0.078
Chain 1: 7700 -8126.213 0.141 0.067
Chain 1: 7800 -8464.559 0.118 0.051
Chain 1: 7900 -8157.234 0.121 0.051
Chain 1: 8000 -8117.164 0.073 0.040
Chain 1: 8100 -8255.475 0.071 0.040
Chain 1: 8200 -8292.513 0.053 0.038
Chain 1: 8300 -10549.031 0.050 0.038
Chain 1: 8400 -12643.516 0.063 0.040
Chain 1: 8500 -8221.954 0.110 0.040
Chain 1: 8600 -8308.890 0.106 0.038
Chain 1: 8700 -8475.609 0.105 0.038
Chain 1: 8800 -8372.809 0.102 0.020
Chain 1: 8900 -8853.822 0.104 0.020
Chain 1: 9000 -10553.704 0.120 0.054
Chain 1: 9100 -8270.952 0.146 0.161
Chain 1: 9200 -8154.285 0.147 0.161
Chain 1: 9300 -9078.105 0.135 0.102
Chain 1: 9400 -7909.427 0.134 0.102
Chain 1: 9500 -7938.881 0.080 0.054
Chain 1: 9600 -8067.552 0.081 0.054
Chain 1: 9700 -8999.226 0.089 0.102
Chain 1: 9800 -10563.091 0.103 0.104
Chain 1: 9900 -8472.830 0.122 0.148
Chain 1: 10000 -9572.858 0.117 0.115
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00138 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.8 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56588.056 1.000 1.000
Chain 1: 200 -17114.727 1.653 2.306
Chain 1: 300 -8575.108 1.434 1.000
Chain 1: 400 -8401.757 1.081 1.000
Chain 1: 500 -8043.745 0.873 0.996
Chain 1: 600 -8707.528 0.741 0.996
Chain 1: 700 -7791.201 0.652 0.118
Chain 1: 800 -8028.544 0.574 0.118
Chain 1: 900 -7857.596 0.513 0.076
Chain 1: 1000 -7844.980 0.461 0.076
Chain 1: 1100 -7753.948 0.363 0.045
Chain 1: 1200 -7534.547 0.135 0.030
Chain 1: 1300 -7760.791 0.038 0.029
Chain 1: 1400 -7924.435 0.038 0.029
Chain 1: 1500 -7559.864 0.039 0.029
Chain 1: 1600 -7500.895 0.032 0.029
Chain 1: 1700 -7495.967 0.020 0.022
Chain 1: 1800 -7523.780 0.017 0.021
Chain 1: 1900 -7569.520 0.016 0.012
Chain 1: 2000 -7560.878 0.016 0.012
Chain 1: 2100 -7568.351 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003256 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.56 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86739.561 1.000 1.000
Chain 1: 200 -13157.967 3.296 5.592
Chain 1: 300 -9599.477 2.321 1.000
Chain 1: 400 -10364.682 1.759 1.000
Chain 1: 500 -8517.007 1.451 0.371
Chain 1: 600 -8161.455 1.216 0.371
Chain 1: 700 -8481.465 1.048 0.217
Chain 1: 800 -8952.852 0.923 0.217
Chain 1: 900 -8393.202 0.828 0.074
Chain 1: 1000 -8299.576 0.747 0.074
Chain 1: 1100 -8517.187 0.649 0.067 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8255.839 0.093 0.053
Chain 1: 1300 -8305.401 0.057 0.044
Chain 1: 1400 -8374.539 0.050 0.038
Chain 1: 1500 -8229.803 0.030 0.032
Chain 1: 1600 -8334.477 0.027 0.026
Chain 1: 1700 -8417.715 0.024 0.018
Chain 1: 1800 -8032.304 0.024 0.018
Chain 1: 1900 -8134.616 0.018 0.013
Chain 1: 2000 -8104.205 0.018 0.013
Chain 1: 2100 -8237.356 0.017 0.013
Chain 1: 2200 -8022.671 0.016 0.013
Chain 1: 2300 -8163.994 0.017 0.016
Chain 1: 2400 -8175.689 0.017 0.016
Chain 1: 2500 -8143.896 0.015 0.013
Chain 1: 2600 -8142.713 0.014 0.013
Chain 1: 2700 -8051.501 0.014 0.013
Chain 1: 2800 -8028.523 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003084 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.84 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8413972.497 1.000 1.000
Chain 1: 200 -1586505.006 2.652 4.303
Chain 1: 300 -891194.946 2.028 1.000
Chain 1: 400 -457571.699 1.758 1.000
Chain 1: 500 -357580.995 1.462 0.948
Chain 1: 600 -232433.954 1.308 0.948
Chain 1: 700 -118725.917 1.258 0.948
Chain 1: 800 -85953.667 1.149 0.948
Chain 1: 900 -66318.468 1.054 0.780
Chain 1: 1000 -51131.736 0.978 0.780
Chain 1: 1100 -38629.530 0.911 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37804.366 0.482 0.381
Chain 1: 1300 -25792.087 0.451 0.381
Chain 1: 1400 -25511.454 0.357 0.324
Chain 1: 1500 -22107.228 0.345 0.324
Chain 1: 1600 -21325.304 0.294 0.297
Chain 1: 1700 -20203.498 0.204 0.296
Chain 1: 1800 -20148.255 0.166 0.154
Chain 1: 1900 -20473.894 0.138 0.056
Chain 1: 2000 -18988.362 0.117 0.056
Chain 1: 2100 -19226.497 0.085 0.037
Chain 1: 2200 -19452.224 0.084 0.037
Chain 1: 2300 -19070.221 0.040 0.020
Chain 1: 2400 -18842.565 0.040 0.020
Chain 1: 2500 -18644.428 0.026 0.016
Chain 1: 2600 -18275.294 0.024 0.016
Chain 1: 2700 -18232.499 0.019 0.012
Chain 1: 2800 -17949.540 0.020 0.016
Chain 1: 2900 -18230.504 0.020 0.015
Chain 1: 3000 -18216.772 0.012 0.012
Chain 1: 3100 -18301.657 0.011 0.012
Chain 1: 3200 -17992.733 0.012 0.015
Chain 1: 3300 -18197.157 0.011 0.012
Chain 1: 3400 -17672.735 0.013 0.015
Chain 1: 3500 -18283.541 0.015 0.016
Chain 1: 3600 -17591.648 0.017 0.016
Chain 1: 3700 -17977.367 0.019 0.017
Chain 1: 3800 -16939.197 0.023 0.021
Chain 1: 3900 -16935.385 0.022 0.021
Chain 1: 4000 -17052.711 0.023 0.021
Chain 1: 4100 -16966.549 0.023 0.021
Chain 1: 4200 -16783.280 0.022 0.021
Chain 1: 4300 -16921.350 0.022 0.021
Chain 1: 4400 -16878.563 0.019 0.011
Chain 1: 4500 -16781.155 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001193 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 11.93 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49247.785 1.000 1.000
Chain 1: 200 -18320.194 1.344 1.688
Chain 1: 300 -19473.216 0.916 1.000
Chain 1: 400 -13822.933 0.789 1.000
Chain 1: 500 -20919.050 0.699 0.409
Chain 1: 600 -19815.682 0.592 0.409
Chain 1: 700 -16429.943 0.537 0.339
Chain 1: 800 -13574.248 0.496 0.339
Chain 1: 900 -15727.996 0.456 0.210
Chain 1: 1000 -12460.111 0.437 0.262
Chain 1: 1100 -16731.111 0.362 0.255
Chain 1: 1200 -19663.138 0.208 0.210
Chain 1: 1300 -11277.138 0.277 0.255
Chain 1: 1400 -12222.924 0.244 0.210
Chain 1: 1500 -10429.082 0.227 0.206
Chain 1: 1600 -10054.275 0.225 0.206
Chain 1: 1700 -9934.539 0.206 0.172
Chain 1: 1800 -16786.161 0.225 0.172
Chain 1: 1900 -11183.129 0.262 0.255
Chain 1: 2000 -9927.644 0.248 0.172
Chain 1: 2100 -16175.595 0.261 0.172
Chain 1: 2200 -22450.387 0.274 0.279
Chain 1: 2300 -11711.648 0.292 0.279
Chain 1: 2400 -10100.361 0.300 0.279
Chain 1: 2500 -10579.682 0.287 0.279
Chain 1: 2600 -13377.840 0.304 0.279
Chain 1: 2700 -9343.732 0.346 0.386
Chain 1: 2800 -10488.343 0.317 0.279
Chain 1: 2900 -10001.539 0.271 0.209
Chain 1: 3000 -9825.922 0.260 0.209
Chain 1: 3100 -9131.928 0.229 0.160
Chain 1: 3200 -11081.114 0.219 0.160
Chain 1: 3300 -9529.257 0.144 0.160
Chain 1: 3400 -9388.004 0.129 0.109
Chain 1: 3500 -9411.247 0.125 0.109
Chain 1: 3600 -9379.266 0.104 0.076
Chain 1: 3700 -13092.713 0.089 0.076
Chain 1: 3800 -16134.709 0.097 0.076
Chain 1: 3900 -9549.147 0.162 0.163
Chain 1: 4000 -9337.624 0.162 0.163
Chain 1: 4100 -9505.066 0.156 0.163
Chain 1: 4200 -14374.572 0.172 0.163
Chain 1: 4300 -9109.197 0.214 0.189
Chain 1: 4400 -9570.252 0.217 0.189
Chain 1: 4500 -15220.208 0.254 0.284
Chain 1: 4600 -10585.552 0.298 0.339
Chain 1: 4700 -10615.040 0.270 0.339
Chain 1: 4800 -8889.598 0.270 0.339
Chain 1: 4900 -9179.806 0.204 0.194
Chain 1: 5000 -16608.892 0.247 0.339
Chain 1: 5100 -9510.587 0.320 0.371
Chain 1: 5200 -9048.696 0.291 0.371
Chain 1: 5300 -15192.735 0.273 0.371
Chain 1: 5400 -15343.504 0.270 0.371
Chain 1: 5500 -11476.221 0.266 0.337
Chain 1: 5600 -13967.550 0.240 0.194
Chain 1: 5700 -15682.891 0.251 0.194
Chain 1: 5800 -9083.540 0.304 0.337
Chain 1: 5900 -13422.757 0.333 0.337
Chain 1: 6000 -8857.026 0.340 0.337
Chain 1: 6100 -8912.132 0.266 0.323
Chain 1: 6200 -10828.177 0.279 0.323
Chain 1: 6300 -15377.959 0.268 0.296
Chain 1: 6400 -12439.788 0.291 0.296
Chain 1: 6500 -12346.057 0.258 0.236
Chain 1: 6600 -10342.109 0.259 0.236
Chain 1: 6700 -9635.724 0.256 0.236
Chain 1: 6800 -9929.867 0.186 0.194
Chain 1: 6900 -9026.717 0.164 0.177
Chain 1: 7000 -9173.353 0.114 0.100
Chain 1: 7100 -8515.016 0.121 0.100
Chain 1: 7200 -10667.576 0.123 0.100
Chain 1: 7300 -11092.163 0.097 0.077
Chain 1: 7400 -9001.564 0.097 0.077
Chain 1: 7500 -8619.430 0.101 0.077
Chain 1: 7600 -9284.922 0.088 0.073
Chain 1: 7700 -9370.160 0.082 0.072
Chain 1: 7800 -9014.874 0.083 0.072
Chain 1: 7900 -12845.733 0.103 0.072
Chain 1: 8000 -10672.578 0.122 0.077
Chain 1: 8100 -9023.989 0.132 0.183
Chain 1: 8200 -9935.753 0.121 0.092
Chain 1: 8300 -8616.203 0.133 0.153
Chain 1: 8400 -9539.159 0.119 0.097
Chain 1: 8500 -8616.850 0.125 0.107
Chain 1: 8600 -10193.887 0.134 0.153
Chain 1: 8700 -9404.196 0.141 0.153
Chain 1: 8800 -10217.824 0.145 0.153
Chain 1: 8900 -9830.313 0.119 0.107
Chain 1: 9000 -8706.734 0.112 0.107
Chain 1: 9100 -10137.873 0.108 0.107
Chain 1: 9200 -8535.637 0.117 0.129
Chain 1: 9300 -8649.190 0.103 0.107
Chain 1: 9400 -8363.852 0.097 0.107
Chain 1: 9500 -8346.519 0.086 0.084
Chain 1: 9600 -11035.172 0.095 0.084
Chain 1: 9700 -8742.352 0.113 0.129
Chain 1: 9800 -11181.181 0.127 0.141
Chain 1: 9900 -10944.452 0.125 0.141
Chain 1: 10000 -9677.690 0.125 0.141
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00138 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.8 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -46319.127 1.000 1.000
Chain 1: 200 -15844.540 1.462 1.923
Chain 1: 300 -8899.417 1.235 1.000
Chain 1: 400 -8258.660 0.945 1.000
Chain 1: 500 -9120.014 0.775 0.780
Chain 1: 600 -9372.689 0.650 0.780
Chain 1: 700 -8009.906 0.582 0.170
Chain 1: 800 -7810.234 0.512 0.170
Chain 1: 900 -8073.134 0.459 0.094
Chain 1: 1000 -7668.957 0.418 0.094
Chain 1: 1100 -7834.203 0.320 0.078
Chain 1: 1200 -7687.604 0.130 0.053
Chain 1: 1300 -7669.674 0.052 0.033
Chain 1: 1400 -7853.078 0.047 0.027
Chain 1: 1500 -7614.912 0.041 0.027
Chain 1: 1600 -7801.848 0.040 0.026
Chain 1: 1700 -7717.164 0.024 0.024
Chain 1: 1800 -7634.509 0.023 0.023
Chain 1: 1900 -7605.220 0.020 0.021
Chain 1: 2000 -7728.544 0.016 0.019
Chain 1: 2100 -7619.939 0.016 0.016
Chain 1: 2200 -7762.874 0.016 0.016
Chain 1: 2300 -7611.464 0.017 0.018
Chain 1: 2400 -7633.678 0.015 0.016
Chain 1: 2500 -7669.300 0.013 0.014
Chain 1: 2600 -7576.877 0.011 0.012
Chain 1: 2700 -7505.765 0.011 0.012
Chain 1: 2800 -7575.102 0.011 0.012
Chain 1: 2900 -7449.086 0.012 0.014
Chain 1: 3000 -7568.694 0.012 0.014
Chain 1: 3100 -7572.293 0.011 0.012
Chain 1: 3200 -7774.920 0.012 0.012
Chain 1: 3300 -7493.237 0.014 0.012
Chain 1: 3400 -7720.912 0.016 0.016
Chain 1: 3500 -7481.431 0.019 0.017
Chain 1: 3600 -7547.605 0.019 0.017
Chain 1: 3700 -7497.332 0.018 0.017
Chain 1: 3800 -7495.063 0.017 0.017
Chain 1: 3900 -7461.536 0.016 0.016
Chain 1: 4000 -7458.344 0.015 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003752 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 37.52 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86616.916 1.000 1.000
Chain 1: 200 -13874.321 3.121 5.243
Chain 1: 300 -10208.710 2.201 1.000
Chain 1: 400 -11060.203 1.670 1.000
Chain 1: 500 -9206.432 1.376 0.359
Chain 1: 600 -9094.892 1.149 0.359
Chain 1: 700 -8594.786 0.993 0.201
Chain 1: 800 -9439.180 0.880 0.201
Chain 1: 900 -8952.392 0.788 0.089
Chain 1: 1000 -8888.776 0.710 0.089
Chain 1: 1100 -8989.224 0.611 0.077 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8536.408 0.092 0.058
Chain 1: 1300 -8886.317 0.060 0.054
Chain 1: 1400 -8864.606 0.053 0.053
Chain 1: 1500 -8758.056 0.034 0.039
Chain 1: 1600 -8872.209 0.034 0.039
Chain 1: 1700 -8943.905 0.029 0.013
Chain 1: 1800 -8519.259 0.025 0.013
Chain 1: 1900 -8620.359 0.021 0.012
Chain 1: 2000 -8595.194 0.020 0.012
Chain 1: 2100 -8721.253 0.021 0.013
Chain 1: 2200 -8522.366 0.018 0.013
Chain 1: 2300 -8615.467 0.015 0.012
Chain 1: 2400 -8683.995 0.015 0.012
Chain 1: 2500 -8630.290 0.015 0.012
Chain 1: 2600 -8632.046 0.014 0.011
Chain 1: 2700 -8548.560 0.014 0.011
Chain 1: 2800 -8507.917 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002896 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 28.96 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8381316.692 1.000 1.000
Chain 1: 200 -1585219.654 2.644 4.287
Chain 1: 300 -891671.983 2.022 1.000
Chain 1: 400 -457737.366 1.753 1.000
Chain 1: 500 -357873.652 1.458 0.948
Chain 1: 600 -232982.520 1.305 0.948
Chain 1: 700 -119448.381 1.254 0.948
Chain 1: 800 -86662.075 1.145 0.948
Chain 1: 900 -67061.678 1.050 0.778
Chain 1: 1000 -51897.327 0.974 0.778
Chain 1: 1100 -39396.865 0.906 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38582.845 0.479 0.378
Chain 1: 1300 -26561.970 0.447 0.378
Chain 1: 1400 -26284.867 0.353 0.317
Chain 1: 1500 -22875.763 0.340 0.317
Chain 1: 1600 -22093.349 0.290 0.292
Chain 1: 1700 -20969.802 0.200 0.292
Chain 1: 1800 -20914.835 0.163 0.149
Chain 1: 1900 -21241.105 0.135 0.054
Chain 1: 2000 -19752.969 0.113 0.054
Chain 1: 2100 -19991.661 0.083 0.035
Chain 1: 2200 -20217.716 0.082 0.035
Chain 1: 2300 -19835.211 0.038 0.019
Chain 1: 2400 -19607.244 0.039 0.019
Chain 1: 2500 -19408.954 0.025 0.015
Chain 1: 2600 -19039.340 0.023 0.015
Chain 1: 2700 -18996.432 0.018 0.012
Chain 1: 2800 -18712.982 0.019 0.015
Chain 1: 2900 -18994.326 0.019 0.015
Chain 1: 3000 -18980.676 0.012 0.012
Chain 1: 3100 -19065.590 0.011 0.012
Chain 1: 3200 -18756.294 0.011 0.015
Chain 1: 3300 -18961.032 0.011 0.012
Chain 1: 3400 -18435.769 0.012 0.015
Chain 1: 3500 -19047.821 0.014 0.015
Chain 1: 3600 -18354.377 0.016 0.015
Chain 1: 3700 -18741.160 0.018 0.016
Chain 1: 3800 -17700.555 0.023 0.021
Chain 1: 3900 -17696.656 0.021 0.021
Chain 1: 4000 -17814.017 0.022 0.021
Chain 1: 4100 -17727.643 0.022 0.021
Chain 1: 4200 -17543.911 0.021 0.021
Chain 1: 4300 -17682.361 0.021 0.021
Chain 1: 4400 -17639.145 0.018 0.010
Chain 1: 4500 -17541.622 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001292 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.92 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12544.118 1.000 1.000
Chain 1: 200 -9342.862 0.671 1.000
Chain 1: 300 -8085.221 0.499 0.343
Chain 1: 400 -8278.497 0.380 0.343
Chain 1: 500 -8217.582 0.306 0.156
Chain 1: 600 -8004.040 0.259 0.156
Chain 1: 700 -7903.310 0.224 0.027
Chain 1: 800 -7933.370 0.197 0.027
Chain 1: 900 -8015.887 0.176 0.023
Chain 1: 1000 -7964.890 0.159 0.023
Chain 1: 1100 -7993.645 0.059 0.013
Chain 1: 1200 -7908.304 0.026 0.011
Chain 1: 1300 -7849.625 0.011 0.010
Chain 1: 1400 -7879.966 0.009 0.007 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001452 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.52 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -62492.373 1.000 1.000
Chain 1: 200 -18168.499 1.720 2.440
Chain 1: 300 -9026.111 1.484 1.013
Chain 1: 400 -9656.640 1.129 1.013
Chain 1: 500 -8504.285 0.931 1.000
Chain 1: 600 -8494.751 0.776 1.000
Chain 1: 700 -8839.113 0.670 0.136
Chain 1: 800 -8077.621 0.598 0.136
Chain 1: 900 -8048.156 0.532 0.094
Chain 1: 1000 -7949.874 0.480 0.094
Chain 1: 1100 -7680.338 0.384 0.065
Chain 1: 1200 -7877.868 0.142 0.039
Chain 1: 1300 -7587.402 0.045 0.038
Chain 1: 1400 -7998.097 0.044 0.038
Chain 1: 1500 -7609.435 0.035 0.038
Chain 1: 1600 -7891.865 0.039 0.038
Chain 1: 1700 -7483.775 0.040 0.038
Chain 1: 1800 -7699.139 0.034 0.036
Chain 1: 1900 -7602.426 0.034 0.036
Chain 1: 2000 -7704.048 0.035 0.036
Chain 1: 2100 -7590.928 0.032 0.036
Chain 1: 2200 -7727.108 0.032 0.036
Chain 1: 2300 -7573.599 0.030 0.028
Chain 1: 2400 -7639.972 0.026 0.020
Chain 1: 2500 -7617.395 0.021 0.018
Chain 1: 2600 -7532.950 0.018 0.015
Chain 1: 2700 -7551.560 0.013 0.013
Chain 1: 2800 -7509.828 0.011 0.013
Chain 1: 2900 -7404.434 0.011 0.013
Chain 1: 3000 -7536.567 0.012 0.014
Chain 1: 3100 -7532.999 0.010 0.011
Chain 1: 3200 -7736.560 0.011 0.011
Chain 1: 3300 -7456.415 0.013 0.011
Chain 1: 3400 -7682.570 0.015 0.014
Chain 1: 3500 -7441.387 0.018 0.018
Chain 1: 3600 -7507.767 0.017 0.018
Chain 1: 3700 -7456.968 0.018 0.018
Chain 1: 3800 -7454.751 0.017 0.018
Chain 1: 3900 -7421.539 0.016 0.018
Chain 1: 4000 -7418.906 0.015 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003382 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.82 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86250.648 1.000 1.000
Chain 1: 200 -13777.434 3.130 5.260
Chain 1: 300 -10042.787 2.211 1.000
Chain 1: 400 -11385.429 1.688 1.000
Chain 1: 500 -8869.093 1.407 0.372
Chain 1: 600 -8390.695 1.182 0.372
Chain 1: 700 -8673.721 1.018 0.284
Chain 1: 800 -9520.824 0.902 0.284
Chain 1: 900 -8832.349 0.810 0.118
Chain 1: 1000 -8781.164 0.730 0.118
Chain 1: 1100 -8797.957 0.630 0.089 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8334.094 0.109 0.078
Chain 1: 1300 -8664.717 0.076 0.057
Chain 1: 1400 -8435.555 0.067 0.056
Chain 1: 1500 -8512.611 0.039 0.038
Chain 1: 1600 -8619.949 0.035 0.033
Chain 1: 1700 -8672.761 0.032 0.027
Chain 1: 1800 -8221.658 0.029 0.027
Chain 1: 1900 -8331.393 0.022 0.013
Chain 1: 2000 -8326.243 0.022 0.013
Chain 1: 2100 -8490.300 0.024 0.019
Chain 1: 2200 -8227.141 0.021 0.019
Chain 1: 2300 -8413.568 0.020 0.019
Chain 1: 2400 -8226.525 0.019 0.019
Chain 1: 2500 -8304.117 0.019 0.019
Chain 1: 2600 -8230.198 0.019 0.019
Chain 1: 2700 -8249.346 0.019 0.019
Chain 1: 2800 -8202.575 0.014 0.013
Chain 1: 2900 -8309.420 0.014 0.013
Chain 1: 3000 -8260.438 0.014 0.013
Chain 1: 3100 -8193.452 0.013 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003341 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.41 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8386858.529 1.000 1.000
Chain 1: 200 -1582726.844 2.649 4.299
Chain 1: 300 -890643.890 2.025 1.000
Chain 1: 400 -457876.954 1.755 1.000
Chain 1: 500 -358491.199 1.460 0.945
Chain 1: 600 -233318.810 1.306 0.945
Chain 1: 700 -119540.826 1.255 0.945
Chain 1: 800 -86794.418 1.146 0.945
Chain 1: 900 -67129.821 1.051 0.777
Chain 1: 1000 -51938.048 0.975 0.777
Chain 1: 1100 -39415.418 0.907 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38598.618 0.479 0.377
Chain 1: 1300 -26534.602 0.447 0.377
Chain 1: 1400 -26255.441 0.353 0.318
Chain 1: 1500 -22837.814 0.340 0.318
Chain 1: 1600 -22054.666 0.290 0.293
Chain 1: 1700 -20924.527 0.201 0.292
Chain 1: 1800 -20868.560 0.163 0.150
Chain 1: 1900 -21195.417 0.135 0.054
Chain 1: 2000 -19703.777 0.114 0.054
Chain 1: 2100 -19942.084 0.083 0.036
Chain 1: 2200 -20169.504 0.082 0.036
Chain 1: 2300 -19785.759 0.039 0.019
Chain 1: 2400 -19557.577 0.039 0.019
Chain 1: 2500 -19359.874 0.025 0.015
Chain 1: 2600 -18989.060 0.023 0.015
Chain 1: 2700 -18945.815 0.018 0.012
Chain 1: 2800 -18662.451 0.019 0.015
Chain 1: 2900 -18944.105 0.019 0.015
Chain 1: 3000 -18930.091 0.012 0.012
Chain 1: 3100 -19015.217 0.011 0.012
Chain 1: 3200 -18705.405 0.011 0.015
Chain 1: 3300 -18910.585 0.011 0.012
Chain 1: 3400 -18384.678 0.012 0.015
Chain 1: 3500 -18997.848 0.015 0.015
Chain 1: 3600 -18302.893 0.016 0.015
Chain 1: 3700 -18690.907 0.018 0.017
Chain 1: 3800 -17648.141 0.023 0.021
Chain 1: 3900 -17644.293 0.021 0.021
Chain 1: 4000 -17761.543 0.022 0.021
Chain 1: 4100 -17675.186 0.022 0.021
Chain 1: 4200 -17490.934 0.021 0.021
Chain 1: 4300 -17629.641 0.021 0.021
Chain 1: 4400 -17586.002 0.018 0.011
Chain 1: 4500 -17488.521 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001318 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.18 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12294.043 1.000 1.000
Chain 1: 200 -9160.030 0.671 1.000
Chain 1: 300 -7938.667 0.499 0.342
Chain 1: 400 -8062.469 0.378 0.342
Chain 1: 500 -7930.529 0.306 0.154
Chain 1: 600 -7838.418 0.257 0.154
Chain 1: 700 -7744.099 0.222 0.017
Chain 1: 800 -7788.124 0.195 0.017
Chain 1: 900 -7913.620 0.175 0.016
Chain 1: 1000 -7849.750 0.158 0.016
Chain 1: 1100 -7843.166 0.058 0.015
Chain 1: 1200 -7765.127 0.025 0.012
Chain 1: 1300 -7711.790 0.010 0.012
Chain 1: 1400 -7735.802 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001553 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.53 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56756.100 1.000 1.000
Chain 1: 200 -17256.146 1.645 2.289
Chain 1: 300 -8611.441 1.431 1.004
Chain 1: 400 -8278.027 1.083 1.004
Chain 1: 500 -8091.246 0.871 1.000
Chain 1: 600 -8738.184 0.738 1.000
Chain 1: 700 -7967.252 0.647 0.097
Chain 1: 800 -7920.113 0.567 0.097
Chain 1: 900 -7924.950 0.504 0.074
Chain 1: 1000 -7596.689 0.458 0.074
Chain 1: 1100 -7740.324 0.360 0.043
Chain 1: 1200 -7494.450 0.134 0.040
Chain 1: 1300 -7506.720 0.034 0.033
Chain 1: 1400 -7548.705 0.030 0.023
Chain 1: 1500 -7515.256 0.028 0.019
Chain 1: 1600 -7660.507 0.023 0.019
Chain 1: 1700 -7422.427 0.016 0.019
Chain 1: 1800 -7482.217 0.017 0.019
Chain 1: 1900 -7459.278 0.017 0.019
Chain 1: 2000 -7485.101 0.013 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00401 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 40.1 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86666.035 1.000 1.000
Chain 1: 200 -13339.297 3.249 5.497
Chain 1: 300 -9713.512 2.290 1.000
Chain 1: 400 -10476.766 1.736 1.000
Chain 1: 500 -8697.311 1.430 0.373
Chain 1: 600 -8160.817 1.202 0.373
Chain 1: 700 -8300.899 1.033 0.205
Chain 1: 800 -9116.458 0.915 0.205
Chain 1: 900 -8575.130 0.820 0.089
Chain 1: 1000 -8176.196 0.743 0.089
Chain 1: 1100 -8569.838 0.648 0.073 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8178.380 0.103 0.066
Chain 1: 1300 -8341.872 0.067 0.063
Chain 1: 1400 -8394.041 0.061 0.049
Chain 1: 1500 -8284.239 0.042 0.048
Chain 1: 1600 -8390.518 0.036 0.046
Chain 1: 1700 -8480.543 0.036 0.046
Chain 1: 1800 -8069.479 0.032 0.046
Chain 1: 1900 -8165.511 0.027 0.020
Chain 1: 2000 -8138.297 0.022 0.013
Chain 1: 2100 -8260.246 0.019 0.013
Chain 1: 2200 -8080.786 0.017 0.013
Chain 1: 2300 -8161.479 0.016 0.013
Chain 1: 2400 -8230.152 0.016 0.013
Chain 1: 2500 -8175.580 0.015 0.012
Chain 1: 2600 -8174.356 0.014 0.011
Chain 1: 2700 -8091.656 0.014 0.010
Chain 1: 2800 -8056.312 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002514 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.14 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8415915.292 1.000 1.000
Chain 1: 200 -1586709.277 2.652 4.304
Chain 1: 300 -890607.946 2.029 1.000
Chain 1: 400 -456981.841 1.759 1.000
Chain 1: 500 -356841.636 1.463 0.949
Chain 1: 600 -232070.854 1.309 0.949
Chain 1: 700 -118674.049 1.258 0.949
Chain 1: 800 -85979.498 1.149 0.949
Chain 1: 900 -66402.194 1.054 0.782
Chain 1: 1000 -51263.232 0.978 0.782
Chain 1: 1100 -38796.904 0.910 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37980.966 0.482 0.380
Chain 1: 1300 -25993.911 0.450 0.380
Chain 1: 1400 -25718.568 0.356 0.321
Chain 1: 1500 -22319.968 0.343 0.321
Chain 1: 1600 -21540.699 0.293 0.295
Chain 1: 1700 -20421.118 0.203 0.295
Chain 1: 1800 -20366.805 0.165 0.152
Chain 1: 1900 -20692.911 0.137 0.055
Chain 1: 2000 -19207.538 0.115 0.055
Chain 1: 2100 -19445.837 0.084 0.036
Chain 1: 2200 -19671.660 0.083 0.036
Chain 1: 2300 -19289.409 0.039 0.020
Chain 1: 2400 -19061.569 0.039 0.020
Chain 1: 2500 -18863.317 0.025 0.016
Chain 1: 2600 -18493.874 0.024 0.016
Chain 1: 2700 -18450.976 0.018 0.012
Chain 1: 2800 -18167.760 0.020 0.016
Chain 1: 2900 -18448.860 0.020 0.015
Chain 1: 3000 -18435.175 0.012 0.012
Chain 1: 3100 -18520.127 0.011 0.012
Chain 1: 3200 -18210.931 0.012 0.015
Chain 1: 3300 -18415.557 0.011 0.012
Chain 1: 3400 -17890.603 0.013 0.015
Chain 1: 3500 -18502.209 0.015 0.016
Chain 1: 3600 -17809.206 0.017 0.016
Chain 1: 3700 -18195.717 0.019 0.017
Chain 1: 3800 -17155.871 0.023 0.021
Chain 1: 3900 -17151.974 0.022 0.021
Chain 1: 4000 -17269.330 0.022 0.021
Chain 1: 4100 -17183.087 0.022 0.021
Chain 1: 4200 -16999.437 0.022 0.021
Chain 1: 4300 -17137.805 0.021 0.021
Chain 1: 4400 -17094.712 0.019 0.011
Chain 1: 4500 -16997.201 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001286 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.86 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13979.137 1.000 1.000
Chain 1: 200 -10692.386 0.654 1.000
Chain 1: 300 -8313.193 0.531 0.307
Chain 1: 400 -8558.147 0.406 0.307
Chain 1: 500 -8180.159 0.334 0.286
Chain 1: 600 -8466.511 0.284 0.286
Chain 1: 700 -8124.294 0.249 0.046
Chain 1: 800 -8227.232 0.220 0.046
Chain 1: 900 -8217.858 0.195 0.042
Chain 1: 1000 -8250.588 0.176 0.042
Chain 1: 1100 -8253.694 0.076 0.034
Chain 1: 1200 -8203.564 0.046 0.029
Chain 1: 1300 -8095.248 0.019 0.013
Chain 1: 1400 -8143.128 0.017 0.013
Chain 1: 1500 -8294.339 0.014 0.013
Chain 1: 1600 -8124.572 0.012 0.013
Chain 1: 1700 -8100.868 0.009 0.006 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001395 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.95 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -59033.955 1.000 1.000
Chain 1: 200 -18693.940 1.579 2.158
Chain 1: 300 -9203.934 1.396 1.031
Chain 1: 400 -8089.615 1.082 1.031
Chain 1: 500 -8780.475 0.881 1.000
Chain 1: 600 -8959.912 0.738 1.000
Chain 1: 700 -8371.624 0.642 0.138
Chain 1: 800 -8367.234 0.562 0.138
Chain 1: 900 -8336.998 0.500 0.079
Chain 1: 1000 -7726.375 0.458 0.079
Chain 1: 1100 -7927.494 0.360 0.079
Chain 1: 1200 -7875.384 0.145 0.070
Chain 1: 1300 -7995.037 0.044 0.025
Chain 1: 1400 -8154.247 0.032 0.020
Chain 1: 1500 -7572.621 0.032 0.020
Chain 1: 1600 -7781.193 0.032 0.025
Chain 1: 1700 -7702.786 0.026 0.020
Chain 1: 1800 -7625.030 0.027 0.020
Chain 1: 1900 -7683.890 0.028 0.020
Chain 1: 2000 -7779.124 0.021 0.015
Chain 1: 2100 -7865.609 0.020 0.012
Chain 1: 2200 -7848.480 0.019 0.012
Chain 1: 2300 -7730.596 0.019 0.012
Chain 1: 2400 -7763.119 0.018 0.011
Chain 1: 2500 -7575.044 0.012 0.011
Chain 1: 2600 -7600.788 0.010 0.010
Chain 1: 2700 -7462.651 0.011 0.011
Chain 1: 2800 -7651.775 0.012 0.012
Chain 1: 2900 -7444.881 0.014 0.015
Chain 1: 3000 -7583.552 0.015 0.018
Chain 1: 3100 -7567.661 0.014 0.018
Chain 1: 3200 -7790.639 0.017 0.019
Chain 1: 3300 -7476.870 0.019 0.025
Chain 1: 3400 -7743.410 0.022 0.025
Chain 1: 3500 -7502.977 0.023 0.028
Chain 1: 3600 -7519.252 0.023 0.028
Chain 1: 3700 -7455.421 0.022 0.028
Chain 1: 3800 -7420.347 0.020 0.028
Chain 1: 3900 -7439.141 0.018 0.018
Chain 1: 4000 -7435.811 0.016 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003156 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.56 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87799.420 1.000 1.000
Chain 1: 200 -14469.974 3.034 5.068
Chain 1: 300 -10548.851 2.146 1.000
Chain 1: 400 -13061.594 1.658 1.000
Chain 1: 500 -10261.601 1.381 0.372
Chain 1: 600 -9006.453 1.174 0.372
Chain 1: 700 -9437.007 1.013 0.273
Chain 1: 800 -9062.865 0.891 0.273
Chain 1: 900 -9319.251 0.795 0.192
Chain 1: 1000 -8761.224 0.722 0.192
Chain 1: 1100 -9172.370 0.627 0.139 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8565.483 0.127 0.071
Chain 1: 1300 -9004.901 0.095 0.064
Chain 1: 1400 -8798.724 0.078 0.049
Chain 1: 1500 -8926.632 0.052 0.046
Chain 1: 1600 -8987.046 0.039 0.045
Chain 1: 1700 -9037.940 0.035 0.041
Chain 1: 1800 -8573.142 0.036 0.045
Chain 1: 1900 -8659.977 0.034 0.045
Chain 1: 2000 -8674.772 0.028 0.023
Chain 1: 2100 -8824.767 0.025 0.017
Chain 1: 2200 -8527.435 0.022 0.017
Chain 1: 2300 -8614.120 0.018 0.014
Chain 1: 2400 -8709.070 0.017 0.011
Chain 1: 2500 -8606.625 0.016 0.011
Chain 1: 2600 -8650.164 0.016 0.011
Chain 1: 2700 -8560.861 0.017 0.011
Chain 1: 2800 -8530.726 0.012 0.010
Chain 1: 2900 -8617.418 0.012 0.010
Chain 1: 3000 -8551.854 0.012 0.010
Chain 1: 3100 -8503.540 0.011 0.010
Chain 1: 3200 -8460.138 0.008 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002852 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 28.52 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8421080.721 1.000 1.000
Chain 1: 200 -1587655.112 2.652 4.304
Chain 1: 300 -890441.573 2.029 1.000
Chain 1: 400 -458073.476 1.758 1.000
Chain 1: 500 -358252.797 1.462 0.944
Chain 1: 600 -233440.190 1.307 0.944
Chain 1: 700 -119945.608 1.256 0.944
Chain 1: 800 -87238.895 1.146 0.944
Chain 1: 900 -67654.514 1.051 0.783
Chain 1: 1000 -52524.119 0.974 0.783
Chain 1: 1100 -40050.956 0.905 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39246.087 0.477 0.375
Chain 1: 1300 -27229.022 0.443 0.375
Chain 1: 1400 -26957.407 0.350 0.311
Chain 1: 1500 -23549.832 0.336 0.311
Chain 1: 1600 -22770.083 0.286 0.289
Chain 1: 1700 -21645.379 0.197 0.288
Chain 1: 1800 -21590.860 0.159 0.145
Chain 1: 1900 -21918.479 0.132 0.052
Chain 1: 2000 -20427.549 0.110 0.052
Chain 1: 2100 -20666.402 0.080 0.034
Chain 1: 2200 -20893.537 0.080 0.034
Chain 1: 2300 -20509.659 0.037 0.019
Chain 1: 2400 -20281.219 0.037 0.019
Chain 1: 2500 -20082.980 0.024 0.015
Chain 1: 2600 -19712.013 0.022 0.015
Chain 1: 2700 -19668.595 0.017 0.012
Chain 1: 2800 -19384.701 0.019 0.015
Chain 1: 2900 -19666.543 0.019 0.014
Chain 1: 3000 -19652.727 0.011 0.012
Chain 1: 3100 -19737.950 0.011 0.011
Chain 1: 3200 -19427.714 0.011 0.014
Chain 1: 3300 -19633.122 0.010 0.011
Chain 1: 3400 -19106.355 0.012 0.014
Chain 1: 3500 -19720.754 0.014 0.015
Chain 1: 3600 -19024.018 0.016 0.015
Chain 1: 3700 -19413.335 0.018 0.016
Chain 1: 3800 -18367.759 0.022 0.020
Chain 1: 3900 -18363.675 0.020 0.020
Chain 1: 4000 -18481.048 0.021 0.020
Chain 1: 4100 -18394.563 0.021 0.020
Chain 1: 4200 -18209.588 0.020 0.020
Chain 1: 4300 -18348.890 0.020 0.020
Chain 1: 4400 -18304.774 0.018 0.010
Chain 1: 4500 -18207.040 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001416 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.16 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49071.545 1.000 1.000
Chain 1: 200 -21276.952 1.153 1.306
Chain 1: 300 -14613.221 0.921 1.000
Chain 1: 400 -12580.415 0.731 1.000
Chain 1: 500 -19418.212 0.655 0.456
Chain 1: 600 -19127.922 0.549 0.456
Chain 1: 700 -14939.354 0.510 0.352
Chain 1: 800 -14886.543 0.447 0.352
Chain 1: 900 -10822.541 0.439 0.352
Chain 1: 1000 -12813.702 0.411 0.352
Chain 1: 1100 -14698.855 0.323 0.280
Chain 1: 1200 -11709.378 0.218 0.255
Chain 1: 1300 -10149.613 0.188 0.162
Chain 1: 1400 -10738.536 0.177 0.155
Chain 1: 1500 -11098.116 0.145 0.154
Chain 1: 1600 -11768.928 0.150 0.154
Chain 1: 1700 -9851.954 0.141 0.154
Chain 1: 1800 -9895.337 0.141 0.154
Chain 1: 1900 -13033.659 0.128 0.154
Chain 1: 2000 -13746.272 0.117 0.128
Chain 1: 2100 -10995.651 0.129 0.154
Chain 1: 2200 -10619.556 0.108 0.057
Chain 1: 2300 -9444.318 0.105 0.057
Chain 1: 2400 -12616.325 0.124 0.124
Chain 1: 2500 -9079.916 0.160 0.195
Chain 1: 2600 -10087.451 0.164 0.195
Chain 1: 2700 -10625.114 0.150 0.124
Chain 1: 2800 -17047.532 0.187 0.241
Chain 1: 2900 -12089.014 0.204 0.250
Chain 1: 3000 -12672.514 0.203 0.250
Chain 1: 3100 -10068.276 0.204 0.251
Chain 1: 3200 -16578.699 0.240 0.259
Chain 1: 3300 -9543.406 0.301 0.377
Chain 1: 3400 -8977.137 0.282 0.377
Chain 1: 3500 -9253.909 0.246 0.259
Chain 1: 3600 -9658.578 0.241 0.259
Chain 1: 3700 -8611.167 0.248 0.259
Chain 1: 3800 -13475.792 0.246 0.259
Chain 1: 3900 -9514.798 0.247 0.259
Chain 1: 4000 -10118.678 0.248 0.259
Chain 1: 4100 -10175.310 0.223 0.122
Chain 1: 4200 -13055.118 0.206 0.122
Chain 1: 4300 -8871.115 0.179 0.122
Chain 1: 4400 -14103.174 0.210 0.221
Chain 1: 4500 -14496.829 0.210 0.221
Chain 1: 4600 -8427.420 0.277 0.361
Chain 1: 4700 -11521.083 0.292 0.361
Chain 1: 4800 -8983.525 0.284 0.282
Chain 1: 4900 -10615.948 0.258 0.269
Chain 1: 5000 -13212.890 0.272 0.269
Chain 1: 5100 -12833.303 0.274 0.269
Chain 1: 5200 -9375.601 0.289 0.282
Chain 1: 5300 -16113.139 0.284 0.282
Chain 1: 5400 -9846.547 0.310 0.282
Chain 1: 5500 -9962.863 0.309 0.282
Chain 1: 5600 -9077.766 0.246 0.269
Chain 1: 5700 -8452.325 0.227 0.197
Chain 1: 5800 -11364.929 0.224 0.197
Chain 1: 5900 -8619.667 0.241 0.256
Chain 1: 6000 -9385.231 0.229 0.256
Chain 1: 6100 -8499.108 0.237 0.256
Chain 1: 6200 -10971.510 0.222 0.225
Chain 1: 6300 -14364.184 0.204 0.225
Chain 1: 6400 -11504.852 0.165 0.225
Chain 1: 6500 -12895.344 0.175 0.225
Chain 1: 6600 -10072.241 0.193 0.236
Chain 1: 6700 -8781.417 0.201 0.236
Chain 1: 6800 -11589.149 0.199 0.236
Chain 1: 6900 -8631.121 0.202 0.236
Chain 1: 7000 -9546.821 0.203 0.236
Chain 1: 7100 -12457.909 0.216 0.236
Chain 1: 7200 -10994.020 0.207 0.236
Chain 1: 7300 -8822.856 0.208 0.242
Chain 1: 7400 -8511.671 0.187 0.234
Chain 1: 7500 -10975.961 0.198 0.234
Chain 1: 7600 -9598.535 0.185 0.225
Chain 1: 7700 -9472.106 0.171 0.225
Chain 1: 7800 -9259.866 0.149 0.144
Chain 1: 7900 -8901.712 0.119 0.133
Chain 1: 8000 -8318.922 0.116 0.133
Chain 1: 8100 -8584.881 0.096 0.070
Chain 1: 8200 -8826.654 0.086 0.040
Chain 1: 8300 -8558.360 0.064 0.037
Chain 1: 8400 -11332.870 0.085 0.040
Chain 1: 8500 -8755.676 0.092 0.040
Chain 1: 8600 -11390.861 0.101 0.040
Chain 1: 8700 -9416.915 0.120 0.070
Chain 1: 8800 -9101.519 0.121 0.070
Chain 1: 8900 -10092.564 0.127 0.098
Chain 1: 9000 -8730.548 0.136 0.156
Chain 1: 9100 -8159.845 0.140 0.156
Chain 1: 9200 -10471.007 0.159 0.210
Chain 1: 9300 -11329.823 0.164 0.210
Chain 1: 9400 -11632.084 0.142 0.156
Chain 1: 9500 -9996.131 0.129 0.156
Chain 1: 9600 -8285.552 0.126 0.156
Chain 1: 9700 -9961.016 0.122 0.156
Chain 1: 9800 -9362.188 0.125 0.156
Chain 1: 9900 -10511.752 0.126 0.156
Chain 1: 10000 -10695.059 0.112 0.109
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001435 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.35 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -45251.738 1.000 1.000
Chain 1: 200 -15393.788 1.470 1.940
Chain 1: 300 -8660.173 1.239 1.000
Chain 1: 400 -8565.393 0.932 1.000
Chain 1: 500 -8244.214 0.753 0.778
Chain 1: 600 -8780.315 0.638 0.778
Chain 1: 700 -8128.785 0.558 0.080
Chain 1: 800 -8029.455 0.490 0.080
Chain 1: 900 -7963.083 0.437 0.061
Chain 1: 1000 -7796.814 0.395 0.061
Chain 1: 1100 -7697.962 0.296 0.039
Chain 1: 1200 -7568.182 0.104 0.021
Chain 1: 1300 -7717.145 0.028 0.019
Chain 1: 1400 -7878.284 0.029 0.020
Chain 1: 1500 -7543.201 0.030 0.020
Chain 1: 1600 -7779.935 0.027 0.020
Chain 1: 1700 -7482.072 0.023 0.020
Chain 1: 1800 -7571.788 0.023 0.020
Chain 1: 1900 -7557.940 0.022 0.020
Chain 1: 2000 -7609.502 0.020 0.019
Chain 1: 2100 -7546.833 0.020 0.019
Chain 1: 2200 -7667.810 0.020 0.019
Chain 1: 2300 -7572.070 0.019 0.016
Chain 1: 2400 -7616.026 0.018 0.013
Chain 1: 2500 -7523.231 0.015 0.012
Chain 1: 2600 -7508.331 0.012 0.012
Chain 1: 2700 -7522.945 0.008 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003131 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.31 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85783.225 1.000 1.000
Chain 1: 200 -13543.929 3.167 5.334
Chain 1: 300 -9909.213 2.234 1.000
Chain 1: 400 -10895.279 1.698 1.000
Chain 1: 500 -8703.352 1.409 0.367
Chain 1: 600 -8369.741 1.180 0.367
Chain 1: 700 -8539.483 1.015 0.252
Chain 1: 800 -9262.727 0.898 0.252
Chain 1: 900 -8745.142 0.804 0.091
Chain 1: 1000 -8529.911 0.727 0.091
Chain 1: 1100 -8756.248 0.629 0.078 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8366.960 0.100 0.059
Chain 1: 1300 -8495.772 0.065 0.047
Chain 1: 1400 -8572.789 0.057 0.040
Chain 1: 1500 -8468.634 0.033 0.026
Chain 1: 1600 -8573.965 0.030 0.025
Chain 1: 1700 -8661.806 0.029 0.025
Chain 1: 1800 -8244.578 0.027 0.025
Chain 1: 1900 -8342.527 0.022 0.015
Chain 1: 2000 -8316.120 0.020 0.012
Chain 1: 2100 -8439.682 0.019 0.012
Chain 1: 2200 -8256.710 0.016 0.012
Chain 1: 2300 -8337.011 0.016 0.012
Chain 1: 2400 -8406.655 0.015 0.012
Chain 1: 2500 -8352.464 0.015 0.012
Chain 1: 2600 -8352.499 0.014 0.010
Chain 1: 2700 -8269.739 0.014 0.010
Chain 1: 2800 -8231.937 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003289 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.89 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8393670.291 1.000 1.000
Chain 1: 200 -1586093.216 2.646 4.292
Chain 1: 300 -891897.259 2.023 1.000
Chain 1: 400 -458130.410 1.754 1.000
Chain 1: 500 -358395.731 1.459 0.947
Chain 1: 600 -233291.371 1.305 0.947
Chain 1: 700 -119400.349 1.255 0.947
Chain 1: 800 -86564.619 1.146 0.947
Chain 1: 900 -66889.353 1.051 0.778
Chain 1: 1000 -51673.495 0.975 0.778
Chain 1: 1100 -39135.797 0.907 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38311.163 0.480 0.379
Chain 1: 1300 -26255.676 0.448 0.379
Chain 1: 1400 -25973.640 0.355 0.320
Chain 1: 1500 -22557.122 0.342 0.320
Chain 1: 1600 -21772.348 0.292 0.294
Chain 1: 1700 -20644.909 0.202 0.294
Chain 1: 1800 -20588.804 0.165 0.151
Chain 1: 1900 -20914.904 0.137 0.055
Chain 1: 2000 -19425.270 0.115 0.055
Chain 1: 2100 -19663.839 0.084 0.036
Chain 1: 2200 -19890.264 0.083 0.036
Chain 1: 2300 -19507.486 0.039 0.020
Chain 1: 2400 -19279.543 0.039 0.020
Chain 1: 2500 -19081.523 0.025 0.016
Chain 1: 2600 -18711.774 0.023 0.016
Chain 1: 2700 -18668.769 0.018 0.012
Chain 1: 2800 -18385.556 0.020 0.015
Chain 1: 2900 -18666.853 0.019 0.015
Chain 1: 3000 -18653.095 0.012 0.012
Chain 1: 3100 -18738.052 0.011 0.012
Chain 1: 3200 -18428.759 0.012 0.015
Chain 1: 3300 -18633.467 0.011 0.012
Chain 1: 3400 -18108.373 0.012 0.015
Chain 1: 3500 -18720.264 0.015 0.015
Chain 1: 3600 -18026.973 0.017 0.015
Chain 1: 3700 -18413.709 0.018 0.017
Chain 1: 3800 -17373.448 0.023 0.021
Chain 1: 3900 -17369.587 0.021 0.021
Chain 1: 4000 -17486.904 0.022 0.021
Chain 1: 4100 -17400.618 0.022 0.021
Chain 1: 4200 -17216.900 0.021 0.021
Chain 1: 4300 -17355.292 0.021 0.021
Chain 1: 4400 -17312.118 0.019 0.011
Chain 1: 4500 -17214.644 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001292 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.92 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49544.409 1.000 1.000
Chain 1: 200 -15267.058 1.623 2.245
Chain 1: 300 -20176.066 1.163 1.000
Chain 1: 400 -20167.423 0.872 1.000
Chain 1: 500 -12284.629 0.826 0.642
Chain 1: 600 -12212.309 0.689 0.642
Chain 1: 700 -16676.717 0.629 0.268
Chain 1: 800 -14535.861 0.569 0.268
Chain 1: 900 -15692.932 0.514 0.243
Chain 1: 1000 -13172.342 0.482 0.243
Chain 1: 1100 -19617.879 0.415 0.243
Chain 1: 1200 -11477.917 0.261 0.243
Chain 1: 1300 -10724.314 0.244 0.191
Chain 1: 1400 -11307.535 0.249 0.191
Chain 1: 1500 -11085.299 0.187 0.147
Chain 1: 1600 -10063.085 0.196 0.147
Chain 1: 1700 -12199.666 0.187 0.147
Chain 1: 1800 -10064.742 0.193 0.175
Chain 1: 1900 -10399.790 0.189 0.175
Chain 1: 2000 -16152.889 0.206 0.175
Chain 1: 2100 -11581.030 0.212 0.175
Chain 1: 2200 -10563.201 0.151 0.102
Chain 1: 2300 -13234.255 0.164 0.175
Chain 1: 2400 -9241.104 0.202 0.202
Chain 1: 2500 -10004.433 0.208 0.202
Chain 1: 2600 -9958.373 0.198 0.202
Chain 1: 2700 -9549.334 0.185 0.202
Chain 1: 2800 -18573.564 0.212 0.202
Chain 1: 2900 -9722.365 0.300 0.356
Chain 1: 3000 -9159.623 0.271 0.202
Chain 1: 3100 -9198.267 0.232 0.096
Chain 1: 3200 -9172.997 0.222 0.076
Chain 1: 3300 -17842.384 0.251 0.076
Chain 1: 3400 -15742.190 0.221 0.076
Chain 1: 3500 -9557.688 0.278 0.133
Chain 1: 3600 -10445.114 0.286 0.133
Chain 1: 3700 -9216.660 0.295 0.133
Chain 1: 3800 -9159.360 0.247 0.133
Chain 1: 3900 -11344.275 0.175 0.133
Chain 1: 4000 -18636.461 0.208 0.133
Chain 1: 4100 -9206.933 0.310 0.193
Chain 1: 4200 -9286.597 0.311 0.193
Chain 1: 4300 -11204.811 0.279 0.171
Chain 1: 4400 -11918.994 0.272 0.171
Chain 1: 4500 -9069.768 0.239 0.171
Chain 1: 4600 -8775.018 0.234 0.171
Chain 1: 4700 -11958.069 0.247 0.193
Chain 1: 4800 -10125.170 0.264 0.193
Chain 1: 4900 -8784.479 0.260 0.181
Chain 1: 5000 -17221.957 0.270 0.181
Chain 1: 5100 -9122.261 0.257 0.181
Chain 1: 5200 -11633.209 0.277 0.216
Chain 1: 5300 -14434.564 0.280 0.216
Chain 1: 5400 -9029.302 0.333 0.266
Chain 1: 5500 -12711.321 0.331 0.266
Chain 1: 5600 -9166.286 0.366 0.290
Chain 1: 5700 -14324.248 0.376 0.360
Chain 1: 5800 -9371.151 0.410 0.387
Chain 1: 5900 -9926.380 0.401 0.387
Chain 1: 6000 -8532.757 0.368 0.360
Chain 1: 6100 -12290.382 0.310 0.306
Chain 1: 6200 -10144.833 0.309 0.306
Chain 1: 6300 -9540.683 0.296 0.306
Chain 1: 6400 -9202.051 0.240 0.290
Chain 1: 6500 -11723.700 0.233 0.215
Chain 1: 6600 -9595.349 0.216 0.215
Chain 1: 6700 -9071.671 0.186 0.211
Chain 1: 6800 -13194.353 0.164 0.211
Chain 1: 6900 -9339.221 0.200 0.215
Chain 1: 7000 -8817.236 0.190 0.215
Chain 1: 7100 -8355.346 0.165 0.211
Chain 1: 7200 -12007.285 0.174 0.215
Chain 1: 7300 -9697.349 0.191 0.222
Chain 1: 7400 -14178.581 0.219 0.238
Chain 1: 7500 -8352.536 0.268 0.304
Chain 1: 7600 -8668.485 0.249 0.304
Chain 1: 7700 -9947.967 0.256 0.304
Chain 1: 7800 -11875.383 0.241 0.238
Chain 1: 7900 -8583.250 0.238 0.238
Chain 1: 8000 -8373.089 0.235 0.238
Chain 1: 8100 -8208.336 0.231 0.238
Chain 1: 8200 -11454.898 0.229 0.238
Chain 1: 8300 -8921.262 0.234 0.283
Chain 1: 8400 -13583.276 0.236 0.283
Chain 1: 8500 -12767.831 0.173 0.162
Chain 1: 8600 -8873.014 0.213 0.283
Chain 1: 8700 -9304.221 0.205 0.283
Chain 1: 8800 -11034.501 0.205 0.283
Chain 1: 8900 -8992.371 0.189 0.227
Chain 1: 9000 -12981.604 0.217 0.283
Chain 1: 9100 -9450.542 0.252 0.284
Chain 1: 9200 -8400.655 0.237 0.284
Chain 1: 9300 -8781.857 0.213 0.227
Chain 1: 9400 -9037.851 0.181 0.157
Chain 1: 9500 -8221.469 0.185 0.157
Chain 1: 9600 -10644.215 0.163 0.157
Chain 1: 9700 -10373.952 0.161 0.157
Chain 1: 9800 -10140.170 0.148 0.125
Chain 1: 9900 -8281.455 0.148 0.125
Chain 1: 10000 -8275.245 0.117 0.099
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001371 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.71 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -62091.766 1.000 1.000
Chain 1: 200 -18140.091 1.711 2.423
Chain 1: 300 -9024.870 1.478 1.010
Chain 1: 400 -9589.761 1.123 1.010
Chain 1: 500 -8110.047 0.935 1.000
Chain 1: 600 -9166.797 0.798 1.000
Chain 1: 700 -8128.916 0.702 0.182
Chain 1: 800 -8126.874 0.615 0.182
Chain 1: 900 -7864.128 0.550 0.128
Chain 1: 1000 -7882.117 0.495 0.128
Chain 1: 1100 -7727.838 0.397 0.115
Chain 1: 1200 -7739.541 0.155 0.059
Chain 1: 1300 -7888.906 0.056 0.033
Chain 1: 1400 -7992.534 0.051 0.020
Chain 1: 1500 -7618.271 0.038 0.020
Chain 1: 1600 -7794.384 0.029 0.020
Chain 1: 1700 -7710.784 0.017 0.019
Chain 1: 1800 -7642.830 0.018 0.019
Chain 1: 1900 -7626.014 0.015 0.013
Chain 1: 2000 -7736.417 0.016 0.014
Chain 1: 2100 -7637.186 0.015 0.013
Chain 1: 2200 -7767.101 0.017 0.014
Chain 1: 2300 -7627.403 0.017 0.014
Chain 1: 2400 -7742.045 0.017 0.015
Chain 1: 2500 -7495.812 0.015 0.015
Chain 1: 2600 -7580.805 0.014 0.014
Chain 1: 2700 -7593.252 0.013 0.014
Chain 1: 2800 -7677.770 0.014 0.014
Chain 1: 2900 -7443.136 0.017 0.015
Chain 1: 3000 -7579.660 0.017 0.017
Chain 1: 3100 -7579.260 0.016 0.017
Chain 1: 3200 -7784.026 0.017 0.018
Chain 1: 3300 -7506.799 0.018 0.018
Chain 1: 3400 -7727.510 0.020 0.026
Chain 1: 3500 -7490.869 0.020 0.026
Chain 1: 3600 -7557.634 0.019 0.026
Chain 1: 3700 -7506.607 0.020 0.026
Chain 1: 3800 -7504.171 0.019 0.026
Chain 1: 3900 -7472.373 0.016 0.018
Chain 1: 4000 -7465.562 0.014 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002637 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 26.37 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87091.314 1.000 1.000
Chain 1: 200 -13811.152 3.153 5.306
Chain 1: 300 -10079.083 2.225 1.000
Chain 1: 400 -11451.836 1.699 1.000
Chain 1: 500 -8642.034 1.424 0.370
Chain 1: 600 -8986.818 1.193 0.370
Chain 1: 700 -8642.002 1.028 0.325
Chain 1: 800 -9201.264 0.908 0.325
Chain 1: 900 -8752.956 0.812 0.120
Chain 1: 1000 -8927.102 0.733 0.120
Chain 1: 1100 -8673.537 0.636 0.061 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8423.438 0.108 0.051
Chain 1: 1300 -8684.319 0.074 0.040
Chain 1: 1400 -8704.171 0.063 0.038
Chain 1: 1500 -8571.108 0.032 0.030
Chain 1: 1600 -8683.179 0.029 0.030
Chain 1: 1700 -8744.590 0.026 0.029
Chain 1: 1800 -8301.658 0.025 0.029
Chain 1: 1900 -8408.542 0.021 0.020
Chain 1: 2000 -8392.749 0.019 0.016
Chain 1: 2100 -8519.169 0.018 0.015
Chain 1: 2200 -8307.339 0.018 0.015
Chain 1: 2300 -8402.276 0.016 0.013
Chain 1: 2400 -8469.375 0.016 0.013
Chain 1: 2500 -8417.271 0.015 0.013
Chain 1: 2600 -8429.609 0.014 0.011
Chain 1: 2700 -8337.994 0.015 0.011
Chain 1: 2800 -8286.240 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003317 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.17 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8436797.714 1.000 1.000
Chain 1: 200 -1590020.295 2.653 4.306
Chain 1: 300 -890930.843 2.030 1.000
Chain 1: 400 -457733.261 1.759 1.000
Chain 1: 500 -357649.117 1.463 0.946
Chain 1: 600 -232412.602 1.309 0.946
Chain 1: 700 -119023.409 1.258 0.946
Chain 1: 800 -86402.425 1.148 0.946
Chain 1: 900 -66832.406 1.053 0.785
Chain 1: 1000 -51724.652 0.977 0.785
Chain 1: 1100 -39282.550 0.909 0.539 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38473.875 0.480 0.378
Chain 1: 1300 -26491.600 0.447 0.378
Chain 1: 1400 -26219.729 0.353 0.317
Chain 1: 1500 -22823.811 0.340 0.317
Chain 1: 1600 -22046.824 0.290 0.293
Chain 1: 1700 -20926.666 0.200 0.292
Chain 1: 1800 -20872.883 0.163 0.149
Chain 1: 1900 -21199.628 0.135 0.054
Chain 1: 2000 -19713.488 0.113 0.054
Chain 1: 2100 -19951.472 0.083 0.035
Chain 1: 2200 -20177.978 0.082 0.035
Chain 1: 2300 -19795.049 0.038 0.019
Chain 1: 2400 -19567.047 0.039 0.019
Chain 1: 2500 -19368.996 0.025 0.015
Chain 1: 2600 -18998.639 0.023 0.015
Chain 1: 2700 -18955.552 0.018 0.012
Chain 1: 2800 -18672.146 0.019 0.015
Chain 1: 2900 -18953.568 0.019 0.015
Chain 1: 3000 -18939.664 0.012 0.012
Chain 1: 3100 -19024.755 0.011 0.012
Chain 1: 3200 -18715.090 0.011 0.015
Chain 1: 3300 -18920.139 0.011 0.012
Chain 1: 3400 -18394.414 0.012 0.015
Chain 1: 3500 -19007.169 0.015 0.015
Chain 1: 3600 -18312.697 0.016 0.015
Chain 1: 3700 -18700.293 0.018 0.017
Chain 1: 3800 -17658.191 0.023 0.021
Chain 1: 3900 -17654.300 0.021 0.021
Chain 1: 4000 -17771.600 0.022 0.021
Chain 1: 4100 -17685.269 0.022 0.021
Chain 1: 4200 -17501.159 0.021 0.021
Chain 1: 4300 -17639.798 0.021 0.021
Chain 1: 4400 -17596.270 0.018 0.011
Chain 1: 4500 -17498.772 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001186 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 11.86 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49151.430 1.000 1.000
Chain 1: 200 -16967.888 1.448 1.897
Chain 1: 300 -22320.644 1.046 1.000
Chain 1: 400 -15193.237 0.901 1.000
Chain 1: 500 -12142.248 0.771 0.469
Chain 1: 600 -16146.468 0.684 0.469
Chain 1: 700 -11709.115 0.641 0.379
Chain 1: 800 -15328.019 0.590 0.379
Chain 1: 900 -13714.011 0.538 0.251
Chain 1: 1000 -11148.048 0.507 0.251
Chain 1: 1100 -11681.959 0.411 0.248
Chain 1: 1200 -14653.918 0.242 0.240
Chain 1: 1300 -10192.654 0.262 0.248
Chain 1: 1400 -11734.441 0.228 0.236
Chain 1: 1500 -18264.968 0.239 0.236
Chain 1: 1600 -9817.814 0.300 0.236
Chain 1: 1700 -9935.402 0.263 0.230
Chain 1: 1800 -10054.525 0.241 0.203
Chain 1: 1900 -10764.847 0.236 0.203
Chain 1: 2000 -9853.497 0.222 0.131
Chain 1: 2100 -10380.087 0.222 0.131
Chain 1: 2200 -11155.290 0.209 0.092
Chain 1: 2300 -11459.755 0.168 0.069
Chain 1: 2400 -9243.612 0.179 0.069
Chain 1: 2500 -9894.764 0.149 0.066
Chain 1: 2600 -8977.888 0.074 0.066
Chain 1: 2700 -11268.145 0.093 0.069
Chain 1: 2800 -9353.192 0.112 0.092
Chain 1: 2900 -9753.157 0.110 0.092
Chain 1: 3000 -10934.197 0.111 0.102
Chain 1: 3100 -9119.750 0.126 0.108
Chain 1: 3200 -9817.810 0.126 0.108
Chain 1: 3300 -11205.932 0.136 0.124
Chain 1: 3400 -13809.551 0.131 0.124
Chain 1: 3500 -12027.985 0.139 0.148
Chain 1: 3600 -14279.254 0.145 0.158
Chain 1: 3700 -9113.999 0.181 0.158
Chain 1: 3800 -8787.801 0.164 0.148
Chain 1: 3900 -11885.423 0.186 0.158
Chain 1: 4000 -8806.492 0.210 0.189
Chain 1: 4100 -8940.941 0.192 0.158
Chain 1: 4200 -9567.971 0.191 0.158
Chain 1: 4300 -11492.003 0.196 0.167
Chain 1: 4400 -9200.559 0.202 0.167
Chain 1: 4500 -9858.922 0.194 0.167
Chain 1: 4600 -9168.017 0.185 0.167
Chain 1: 4700 -15861.969 0.171 0.167
Chain 1: 4800 -8911.180 0.245 0.249
Chain 1: 4900 -13618.504 0.254 0.249
Chain 1: 5000 -9358.726 0.264 0.249
Chain 1: 5100 -10353.776 0.272 0.249
Chain 1: 5200 -10113.367 0.268 0.249
Chain 1: 5300 -12394.625 0.270 0.249
Chain 1: 5400 -8775.716 0.286 0.346
Chain 1: 5500 -13681.546 0.315 0.359
Chain 1: 5600 -8574.989 0.367 0.412
Chain 1: 5700 -13542.476 0.362 0.367
Chain 1: 5800 -8895.993 0.336 0.367
Chain 1: 5900 -9254.980 0.305 0.367
Chain 1: 6000 -8521.420 0.268 0.359
Chain 1: 6100 -8475.937 0.259 0.359
Chain 1: 6200 -9633.767 0.269 0.359
Chain 1: 6300 -8603.464 0.263 0.359
Chain 1: 6400 -10420.967 0.239 0.174
Chain 1: 6500 -8697.990 0.223 0.174
Chain 1: 6600 -11574.784 0.188 0.174
Chain 1: 6700 -8414.169 0.189 0.174
Chain 1: 6800 -8514.562 0.138 0.120
Chain 1: 6900 -9669.779 0.146 0.120
Chain 1: 7000 -8484.770 0.151 0.140
Chain 1: 7100 -8265.738 0.153 0.140
Chain 1: 7200 -8343.955 0.142 0.140
Chain 1: 7300 -8340.857 0.130 0.140
Chain 1: 7400 -9023.713 0.121 0.119
Chain 1: 7500 -11984.976 0.125 0.119
Chain 1: 7600 -9915.569 0.121 0.119
Chain 1: 7700 -8927.730 0.095 0.111
Chain 1: 7800 -8703.890 0.096 0.111
Chain 1: 7900 -8379.426 0.088 0.076
Chain 1: 8000 -8421.849 0.075 0.039
Chain 1: 8100 -8237.762 0.074 0.039
Chain 1: 8200 -8127.070 0.075 0.039
Chain 1: 8300 -8189.899 0.076 0.039
Chain 1: 8400 -8846.548 0.075 0.039
Chain 1: 8500 -8704.317 0.052 0.026
Chain 1: 8600 -10225.621 0.046 0.026
Chain 1: 8700 -8354.119 0.058 0.026
Chain 1: 8800 -8205.436 0.057 0.022
Chain 1: 8900 -8520.317 0.057 0.022
Chain 1: 9000 -11061.203 0.079 0.037
Chain 1: 9100 -8409.116 0.108 0.074
Chain 1: 9200 -8020.727 0.112 0.074
Chain 1: 9300 -8382.234 0.116 0.074
Chain 1: 9400 -8356.782 0.108 0.048
Chain 1: 9500 -8678.794 0.110 0.048
Chain 1: 9600 -8235.138 0.101 0.048
Chain 1: 9700 -8832.009 0.085 0.048
Chain 1: 9800 -8531.236 0.087 0.048
Chain 1: 9900 -9885.931 0.097 0.054
Chain 1: 10000 -8225.538 0.094 0.054
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00138 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.8 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58028.217 1.000 1.000
Chain 1: 200 -17678.813 1.641 2.282
Chain 1: 300 -8726.707 1.436 1.026
Chain 1: 400 -8260.046 1.091 1.026
Chain 1: 500 -8217.654 0.874 1.000
Chain 1: 600 -8388.207 0.732 1.000
Chain 1: 700 -8340.946 0.628 0.056
Chain 1: 800 -8215.253 0.551 0.056
Chain 1: 900 -8093.594 0.492 0.020
Chain 1: 1000 -7790.728 0.447 0.039
Chain 1: 1100 -7729.850 0.347 0.020
Chain 1: 1200 -7681.580 0.120 0.015
Chain 1: 1300 -7694.782 0.017 0.015
Chain 1: 1400 -7697.734 0.012 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004962 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 49.62 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86812.851 1.000 1.000
Chain 1: 200 -13518.140 3.211 5.422
Chain 1: 300 -9866.781 2.264 1.000
Chain 1: 400 -10716.259 1.718 1.000
Chain 1: 500 -8854.114 1.416 0.370
Chain 1: 600 -8482.372 1.188 0.370
Chain 1: 700 -8229.715 1.022 0.210
Chain 1: 800 -8748.739 0.902 0.210
Chain 1: 900 -8701.946 0.802 0.079
Chain 1: 1000 -8343.663 0.726 0.079
Chain 1: 1100 -8724.565 0.631 0.059 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8311.813 0.094 0.050
Chain 1: 1300 -8466.704 0.058 0.044
Chain 1: 1400 -8525.124 0.051 0.044
Chain 1: 1500 -8424.930 0.031 0.043
Chain 1: 1600 -8530.862 0.028 0.031
Chain 1: 1700 -8617.513 0.026 0.018
Chain 1: 1800 -8198.768 0.025 0.018
Chain 1: 1900 -8297.525 0.026 0.018
Chain 1: 2000 -8271.379 0.022 0.012
Chain 1: 2100 -8395.410 0.019 0.012
Chain 1: 2200 -8208.843 0.016 0.012
Chain 1: 2300 -8292.061 0.015 0.012
Chain 1: 2400 -8361.565 0.016 0.012
Chain 1: 2500 -8307.497 0.015 0.012
Chain 1: 2600 -8307.830 0.014 0.010
Chain 1: 2700 -8224.992 0.014 0.010
Chain 1: 2800 -8186.569 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003034 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.34 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8417193.636 1.000 1.000
Chain 1: 200 -1587129.464 2.652 4.303
Chain 1: 300 -890530.053 2.029 1.000
Chain 1: 400 -457092.287 1.758 1.000
Chain 1: 500 -357287.557 1.463 0.948
Chain 1: 600 -232457.565 1.308 0.948
Chain 1: 700 -119014.314 1.258 0.948
Chain 1: 800 -86252.966 1.148 0.948
Chain 1: 900 -66659.398 1.053 0.782
Chain 1: 1000 -51496.761 0.977 0.782
Chain 1: 1100 -39008.812 0.909 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38191.909 0.481 0.380
Chain 1: 1300 -26188.037 0.449 0.380
Chain 1: 1400 -25911.114 0.355 0.320
Chain 1: 1500 -22507.659 0.342 0.320
Chain 1: 1600 -21726.830 0.292 0.294
Chain 1: 1700 -20605.462 0.202 0.294
Chain 1: 1800 -20550.773 0.164 0.151
Chain 1: 1900 -20876.990 0.136 0.054
Chain 1: 2000 -19390.409 0.115 0.054
Chain 1: 2100 -19628.859 0.084 0.036
Chain 1: 2200 -19854.752 0.083 0.036
Chain 1: 2300 -19472.409 0.039 0.020
Chain 1: 2400 -19244.518 0.039 0.020
Chain 1: 2500 -19046.218 0.025 0.016
Chain 1: 2600 -18676.731 0.023 0.016
Chain 1: 2700 -18633.811 0.018 0.012
Chain 1: 2800 -18350.470 0.020 0.015
Chain 1: 2900 -18631.719 0.019 0.015
Chain 1: 3000 -18618.020 0.012 0.012
Chain 1: 3100 -18702.956 0.011 0.012
Chain 1: 3200 -18393.728 0.012 0.015
Chain 1: 3300 -18598.395 0.011 0.012
Chain 1: 3400 -18073.294 0.013 0.015
Chain 1: 3500 -18685.087 0.015 0.015
Chain 1: 3600 -17991.890 0.017 0.015
Chain 1: 3700 -18378.548 0.018 0.017
Chain 1: 3800 -17338.336 0.023 0.021
Chain 1: 3900 -17334.427 0.021 0.021
Chain 1: 4000 -17451.801 0.022 0.021
Chain 1: 4100 -17365.496 0.022 0.021
Chain 1: 4200 -17181.782 0.021 0.021
Chain 1: 4300 -17320.210 0.021 0.021
Chain 1: 4400 -17277.076 0.019 0.011
Chain 1: 4500 -17179.541 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00117 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 11.7 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48785.296 1.000 1.000
Chain 1: 200 -22832.511 1.068 1.137
Chain 1: 300 -14615.144 0.900 1.000
Chain 1: 400 -15075.575 0.682 1.000
Chain 1: 500 -13949.216 0.562 0.562
Chain 1: 600 -12737.314 0.484 0.562
Chain 1: 700 -14440.479 0.432 0.118
Chain 1: 800 -13759.522 0.384 0.118
Chain 1: 900 -12585.075 0.352 0.095
Chain 1: 1000 -30907.768 0.376 0.118
Chain 1: 1100 -11818.774 0.437 0.118
Chain 1: 1200 -12135.519 0.326 0.095
Chain 1: 1300 -12192.471 0.271 0.093
Chain 1: 1400 -9886.620 0.291 0.095
Chain 1: 1500 -10056.353 0.284 0.095
Chain 1: 1600 -10291.998 0.277 0.093
Chain 1: 1700 -17516.325 0.307 0.093
Chain 1: 1800 -11634.812 0.352 0.233
Chain 1: 1900 -11191.859 0.347 0.233
Chain 1: 2000 -11147.942 0.288 0.040
Chain 1: 2100 -16056.097 0.157 0.040
Chain 1: 2200 -9520.281 0.223 0.233
Chain 1: 2300 -16560.727 0.265 0.306
Chain 1: 2400 -9758.289 0.312 0.412
Chain 1: 2500 -10022.097 0.313 0.412
Chain 1: 2600 -9163.824 0.320 0.412
Chain 1: 2700 -9295.226 0.280 0.306
Chain 1: 2800 -8767.878 0.235 0.094
Chain 1: 2900 -9352.719 0.238 0.094
Chain 1: 3000 -9413.424 0.238 0.094
Chain 1: 3100 -9482.673 0.208 0.063
Chain 1: 3200 -8896.116 0.146 0.063
Chain 1: 3300 -9540.711 0.110 0.063
Chain 1: 3400 -10300.507 0.048 0.063
Chain 1: 3500 -9194.400 0.057 0.066
Chain 1: 3600 -8687.389 0.054 0.063
Chain 1: 3700 -8763.820 0.053 0.063
Chain 1: 3800 -11152.497 0.069 0.066
Chain 1: 3900 -8627.891 0.092 0.068
Chain 1: 4000 -9887.697 0.104 0.074
Chain 1: 4100 -14205.969 0.133 0.120
Chain 1: 4200 -12057.690 0.145 0.127
Chain 1: 4300 -8792.776 0.175 0.178
Chain 1: 4400 -8973.015 0.170 0.178
Chain 1: 4500 -15628.822 0.200 0.214
Chain 1: 4600 -9095.773 0.266 0.293
Chain 1: 4700 -8481.154 0.272 0.293
Chain 1: 4800 -8741.695 0.254 0.293
Chain 1: 4900 -9453.721 0.232 0.178
Chain 1: 5000 -8528.603 0.230 0.178
Chain 1: 5100 -8403.071 0.201 0.108
Chain 1: 5200 -15419.276 0.229 0.108
Chain 1: 5300 -13227.343 0.209 0.108
Chain 1: 5400 -11086.672 0.226 0.166
Chain 1: 5500 -12407.243 0.194 0.108
Chain 1: 5600 -13928.610 0.133 0.108
Chain 1: 5700 -12539.024 0.137 0.109
Chain 1: 5800 -8622.185 0.179 0.111
Chain 1: 5900 -12689.928 0.204 0.166
Chain 1: 6000 -8183.173 0.248 0.193
Chain 1: 6100 -9115.347 0.257 0.193
Chain 1: 6200 -8326.777 0.221 0.166
Chain 1: 6300 -8476.069 0.206 0.111
Chain 1: 6400 -10344.281 0.205 0.111
Chain 1: 6500 -12524.655 0.211 0.174
Chain 1: 6600 -9028.798 0.239 0.181
Chain 1: 6700 -10484.425 0.242 0.181
Chain 1: 6800 -8190.155 0.225 0.181
Chain 1: 6900 -10098.450 0.212 0.181
Chain 1: 7000 -8399.619 0.177 0.181
Chain 1: 7100 -8061.291 0.171 0.181
Chain 1: 7200 -8844.249 0.170 0.181
Chain 1: 7300 -8458.540 0.173 0.181
Chain 1: 7400 -10412.543 0.174 0.188
Chain 1: 7500 -9234.886 0.169 0.188
Chain 1: 7600 -8390.218 0.140 0.139
Chain 1: 7700 -8628.084 0.129 0.128
Chain 1: 7800 -8931.850 0.104 0.101
Chain 1: 7900 -8397.021 0.092 0.089
Chain 1: 8000 -9646.955 0.085 0.089
Chain 1: 8100 -9728.675 0.081 0.089
Chain 1: 8200 -8090.278 0.093 0.101
Chain 1: 8300 -10449.306 0.111 0.128
Chain 1: 8400 -8459.664 0.115 0.128
Chain 1: 8500 -8348.911 0.104 0.101
Chain 1: 8600 -11882.039 0.124 0.130
Chain 1: 8700 -8547.988 0.160 0.203
Chain 1: 8800 -9490.512 0.167 0.203
Chain 1: 8900 -10832.973 0.173 0.203
Chain 1: 9000 -9946.958 0.168 0.203
Chain 1: 9100 -8039.691 0.191 0.226
Chain 1: 9200 -9685.241 0.188 0.226
Chain 1: 9300 -8087.319 0.185 0.198
Chain 1: 9400 -8686.932 0.169 0.170
Chain 1: 9500 -8148.880 0.174 0.170
Chain 1: 9600 -8815.024 0.152 0.124
Chain 1: 9700 -8115.577 0.121 0.099
Chain 1: 9800 -8166.630 0.112 0.089
Chain 1: 9900 -10480.765 0.122 0.089
Chain 1: 10000 -10533.530 0.113 0.086
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001399 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.99 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56818.013 1.000 1.000
Chain 1: 200 -17354.078 1.637 2.274
Chain 1: 300 -8680.620 1.424 1.000
Chain 1: 400 -8354.453 1.078 1.000
Chain 1: 500 -8235.956 0.865 0.999
Chain 1: 600 -8089.495 0.724 0.999
Chain 1: 700 -8026.847 0.622 0.039
Chain 1: 800 -8098.634 0.545 0.039
Chain 1: 900 -7901.460 0.487 0.025
Chain 1: 1000 -7855.907 0.439 0.025
Chain 1: 1100 -7701.705 0.341 0.020
Chain 1: 1200 -7842.578 0.116 0.018
Chain 1: 1300 -7638.173 0.018 0.018
Chain 1: 1400 -7839.654 0.017 0.018
Chain 1: 1500 -7598.318 0.019 0.020
Chain 1: 1600 -7667.058 0.018 0.020
Chain 1: 1700 -7521.150 0.019 0.020
Chain 1: 1800 -7594.140 0.019 0.020
Chain 1: 1900 -7559.437 0.017 0.019
Chain 1: 2000 -7644.455 0.018 0.019
Chain 1: 2100 -7591.475 0.016 0.018
Chain 1: 2200 -7694.420 0.016 0.013
Chain 1: 2300 -7605.122 0.014 0.012
Chain 1: 2400 -7644.311 0.012 0.011
Chain 1: 2500 -7582.591 0.010 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003163 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.63 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86519.913 1.000 1.000
Chain 1: 200 -13398.754 3.229 5.457
Chain 1: 300 -9787.869 2.275 1.000
Chain 1: 400 -10738.713 1.729 1.000
Chain 1: 500 -8706.298 1.430 0.369
Chain 1: 600 -8292.311 1.200 0.369
Chain 1: 700 -8410.183 1.030 0.233
Chain 1: 800 -9112.326 0.911 0.233
Chain 1: 900 -8592.886 0.817 0.089
Chain 1: 1000 -8402.134 0.737 0.089
Chain 1: 1100 -8630.436 0.640 0.077 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8272.544 0.098 0.060
Chain 1: 1300 -8475.369 0.064 0.050
Chain 1: 1400 -8482.815 0.055 0.043
Chain 1: 1500 -8370.557 0.033 0.026
Chain 1: 1600 -8474.672 0.029 0.024
Chain 1: 1700 -8563.380 0.029 0.024
Chain 1: 1800 -8157.128 0.026 0.024
Chain 1: 1900 -8254.571 0.021 0.023
Chain 1: 2000 -8226.403 0.020 0.013
Chain 1: 2100 -8346.587 0.018 0.013
Chain 1: 2200 -8147.963 0.016 0.013
Chain 1: 2300 -8291.261 0.016 0.013
Chain 1: 2400 -8298.014 0.016 0.013
Chain 1: 2500 -8268.097 0.015 0.012
Chain 1: 2600 -8266.635 0.014 0.012
Chain 1: 2700 -8179.318 0.014 0.012
Chain 1: 2800 -8145.195 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003196 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.96 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8393918.293 1.000 1.000
Chain 1: 200 -1583955.697 2.650 4.299
Chain 1: 300 -891069.822 2.026 1.000
Chain 1: 400 -457857.265 1.756 1.000
Chain 1: 500 -358299.513 1.460 0.946
Chain 1: 600 -233215.393 1.306 0.946
Chain 1: 700 -119274.753 1.256 0.946
Chain 1: 800 -86421.486 1.147 0.946
Chain 1: 900 -66739.739 1.052 0.778
Chain 1: 1000 -51511.584 0.976 0.778
Chain 1: 1100 -38966.699 0.909 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38139.095 0.481 0.380
Chain 1: 1300 -26081.900 0.449 0.380
Chain 1: 1400 -25798.607 0.356 0.322
Chain 1: 1500 -22381.909 0.343 0.322
Chain 1: 1600 -21596.531 0.293 0.296
Chain 1: 1700 -20469.251 0.203 0.295
Chain 1: 1800 -20412.906 0.165 0.153
Chain 1: 1900 -20738.870 0.138 0.055
Chain 1: 2000 -19249.512 0.116 0.055
Chain 1: 2100 -19488.086 0.085 0.036
Chain 1: 2200 -19714.393 0.084 0.036
Chain 1: 2300 -19331.723 0.039 0.020
Chain 1: 2400 -19103.872 0.040 0.020
Chain 1: 2500 -18905.808 0.025 0.016
Chain 1: 2600 -18536.323 0.024 0.016
Chain 1: 2700 -18493.316 0.018 0.012
Chain 1: 2800 -18210.245 0.020 0.016
Chain 1: 2900 -18491.412 0.020 0.015
Chain 1: 3000 -18477.670 0.012 0.012
Chain 1: 3100 -18562.624 0.011 0.012
Chain 1: 3200 -18253.439 0.012 0.015
Chain 1: 3300 -18458.024 0.011 0.012
Chain 1: 3400 -17933.188 0.013 0.015
Chain 1: 3500 -18544.714 0.015 0.016
Chain 1: 3600 -17851.842 0.017 0.016
Chain 1: 3700 -18238.341 0.019 0.017
Chain 1: 3800 -17198.719 0.023 0.021
Chain 1: 3900 -17194.850 0.022 0.021
Chain 1: 4000 -17312.171 0.022 0.021
Chain 1: 4100 -17225.981 0.022 0.021
Chain 1: 4200 -17042.337 0.022 0.021
Chain 1: 4300 -17180.665 0.021 0.021
Chain 1: 4400 -17137.631 0.019 0.011
Chain 1: 4500 -17040.155 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001358 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.58 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12677.997 1.000 1.000
Chain 1: 200 -9598.614 0.660 1.000
Chain 1: 300 -8103.526 0.502 0.321
Chain 1: 400 -8341.583 0.383 0.321
Chain 1: 500 -8238.064 0.309 0.184
Chain 1: 600 -8069.730 0.261 0.184
Chain 1: 700 -7971.094 0.226 0.029
Chain 1: 800 -8049.993 0.199 0.029
Chain 1: 900 -7936.899 0.178 0.021
Chain 1: 1000 -8119.662 0.163 0.023
Chain 1: 1100 -8304.467 0.065 0.022
Chain 1: 1200 -7994.504 0.037 0.022
Chain 1: 1300 -7966.455 0.019 0.021
Chain 1: 1400 -7955.086 0.016 0.014
Chain 1: 1500 -8051.399 0.016 0.014
Chain 1: 1600 -7995.436 0.014 0.012
Chain 1: 1700 -7945.585 0.014 0.012
Chain 1: 1800 -7918.781 0.013 0.012
Chain 1: 1900 -7941.370 0.012 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001594 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.94 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -51243.051 1.000 1.000
Chain 1: 200 -16576.622 1.546 2.091
Chain 1: 300 -8867.257 1.320 1.000
Chain 1: 400 -10227.879 1.023 1.000
Chain 1: 500 -8653.894 0.855 0.869
Chain 1: 600 -8462.818 0.716 0.869
Chain 1: 700 -8172.695 0.619 0.182
Chain 1: 800 -8230.340 0.543 0.182
Chain 1: 900 -7557.947 0.492 0.133
Chain 1: 1000 -7767.815 0.446 0.133
Chain 1: 1100 -7790.965 0.346 0.089
Chain 1: 1200 -7766.536 0.137 0.035
Chain 1: 1300 -7687.923 0.051 0.027
Chain 1: 1400 -7762.839 0.039 0.023
Chain 1: 1500 -7571.295 0.023 0.023
Chain 1: 1600 -7798.260 0.024 0.025
Chain 1: 1700 -7667.283 0.022 0.017
Chain 1: 1800 -7554.559 0.023 0.017
Chain 1: 1900 -7656.056 0.015 0.015
Chain 1: 2000 -7589.813 0.013 0.013
Chain 1: 2100 -7551.224 0.014 0.013
Chain 1: 2200 -7749.929 0.016 0.015
Chain 1: 2300 -7565.681 0.017 0.017
Chain 1: 2400 -7624.929 0.017 0.017
Chain 1: 2500 -7552.760 0.016 0.015
Chain 1: 2600 -7496.838 0.013 0.013
Chain 1: 2700 -7494.609 0.012 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003664 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 36.64 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86629.647 1.000 1.000
Chain 1: 200 -13823.882 3.133 5.267
Chain 1: 300 -10076.553 2.213 1.000
Chain 1: 400 -11592.368 1.692 1.000
Chain 1: 500 -8797.409 1.417 0.372
Chain 1: 600 -8409.759 1.189 0.372
Chain 1: 700 -8660.703 1.023 0.318
Chain 1: 800 -9480.708 0.906 0.318
Chain 1: 900 -8773.285 0.814 0.131
Chain 1: 1000 -8446.124 0.737 0.131
Chain 1: 1100 -8837.520 0.641 0.086 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8412.281 0.120 0.081
Chain 1: 1300 -8795.334 0.087 0.051
Chain 1: 1400 -8558.815 0.076 0.046
Chain 1: 1500 -8593.498 0.045 0.044
Chain 1: 1600 -8691.565 0.042 0.044
Chain 1: 1700 -8751.253 0.039 0.044
Chain 1: 1800 -8313.388 0.036 0.044
Chain 1: 1900 -8417.158 0.029 0.039
Chain 1: 2000 -8397.763 0.026 0.028
Chain 1: 2100 -8521.248 0.023 0.014
Chain 1: 2200 -8315.629 0.020 0.014
Chain 1: 2300 -8409.211 0.017 0.012
Chain 1: 2400 -8476.406 0.015 0.011
Chain 1: 2500 -8425.185 0.015 0.011
Chain 1: 2600 -8437.385 0.014 0.011
Chain 1: 2700 -8345.900 0.014 0.011
Chain 1: 2800 -8294.433 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004042 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 40.42 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8395948.553 1.000 1.000
Chain 1: 200 -1582709.730 2.652 4.305
Chain 1: 300 -892010.680 2.026 1.000
Chain 1: 400 -458539.326 1.756 1.000
Chain 1: 500 -359260.105 1.460 0.945
Chain 1: 600 -234019.716 1.306 0.945
Chain 1: 700 -119926.087 1.255 0.945
Chain 1: 800 -87039.143 1.146 0.945
Chain 1: 900 -67311.352 1.051 0.774
Chain 1: 1000 -52063.118 0.975 0.774
Chain 1: 1100 -39492.817 0.907 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38667.003 0.479 0.378
Chain 1: 1300 -26570.283 0.447 0.378
Chain 1: 1400 -26285.901 0.353 0.318
Chain 1: 1500 -22858.737 0.341 0.318
Chain 1: 1600 -22071.355 0.291 0.293
Chain 1: 1700 -20938.644 0.201 0.293
Chain 1: 1800 -20881.466 0.163 0.150
Chain 1: 1900 -21208.076 0.136 0.054
Chain 1: 2000 -19714.752 0.114 0.054
Chain 1: 2100 -19953.470 0.083 0.036
Chain 1: 2200 -20180.756 0.082 0.036
Chain 1: 2300 -19797.096 0.039 0.019
Chain 1: 2400 -19568.944 0.039 0.019
Chain 1: 2500 -19371.070 0.025 0.015
Chain 1: 2600 -19000.649 0.023 0.015
Chain 1: 2700 -18957.404 0.018 0.012
Chain 1: 2800 -18674.109 0.019 0.015
Chain 1: 2900 -18955.638 0.019 0.015
Chain 1: 3000 -18941.776 0.012 0.012
Chain 1: 3100 -19026.866 0.011 0.012
Chain 1: 3200 -18717.150 0.011 0.015
Chain 1: 3300 -18922.161 0.011 0.012
Chain 1: 3400 -18396.473 0.012 0.015
Chain 1: 3500 -19009.298 0.015 0.015
Chain 1: 3600 -18314.746 0.016 0.015
Chain 1: 3700 -18702.536 0.018 0.017
Chain 1: 3800 -17660.309 0.023 0.021
Chain 1: 3900 -17656.412 0.021 0.021
Chain 1: 4000 -17773.715 0.022 0.021
Chain 1: 4100 -17687.386 0.022 0.021
Chain 1: 4200 -17503.201 0.021 0.021
Chain 1: 4300 -17641.897 0.021 0.021
Chain 1: 4400 -17598.394 0.018 0.011
Chain 1: 4500 -17500.855 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001452 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.52 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12520.437 1.000 1.000
Chain 1: 200 -9396.844 0.666 1.000
Chain 1: 300 -8084.680 0.498 0.332
Chain 1: 400 -8267.508 0.379 0.332
Chain 1: 500 -8243.579 0.304 0.162
Chain 1: 600 -8028.592 0.258 0.162
Chain 1: 700 -7928.560 0.223 0.027
Chain 1: 800 -7962.844 0.195 0.027
Chain 1: 900 -8080.467 0.175 0.022
Chain 1: 1000 -7977.578 0.159 0.022
Chain 1: 1100 -8013.632 0.060 0.015
Chain 1: 1200 -7977.463 0.027 0.013
Chain 1: 1300 -7896.430 0.012 0.013
Chain 1: 1400 -7920.932 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002227 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 22.27 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -46596.662 1.000 1.000
Chain 1: 200 -15712.851 1.483 1.966
Chain 1: 300 -8736.623 1.255 1.000
Chain 1: 400 -8614.190 0.945 1.000
Chain 1: 500 -8693.508 0.757 0.799
Chain 1: 600 -8298.251 0.639 0.799
Chain 1: 700 -7936.485 0.554 0.048
Chain 1: 800 -8129.397 0.488 0.048
Chain 1: 900 -7963.178 0.436 0.046
Chain 1: 1000 -7640.800 0.397 0.046
Chain 1: 1100 -7615.442 0.297 0.042
Chain 1: 1200 -7863.644 0.104 0.032
Chain 1: 1300 -7812.825 0.024 0.024
Chain 1: 1400 -7875.638 0.024 0.024
Chain 1: 1500 -7531.807 0.028 0.032
Chain 1: 1600 -7762.969 0.026 0.030
Chain 1: 1700 -7432.596 0.026 0.030
Chain 1: 1800 -7628.839 0.026 0.030
Chain 1: 1900 -7427.579 0.026 0.030
Chain 1: 2000 -7584.550 0.024 0.027
Chain 1: 2100 -7598.546 0.024 0.027
Chain 1: 2200 -7689.384 0.022 0.026
Chain 1: 2300 -7553.853 0.023 0.026
Chain 1: 2400 -7609.074 0.023 0.026
Chain 1: 2500 -7637.324 0.019 0.021
Chain 1: 2600 -7475.641 0.018 0.021
Chain 1: 2700 -7475.596 0.014 0.018
Chain 1: 2800 -7539.883 0.012 0.012
Chain 1: 2900 -7363.888 0.012 0.012
Chain 1: 3000 -7492.394 0.011 0.012
Chain 1: 3100 -7484.371 0.011 0.012
Chain 1: 3200 -7682.118 0.013 0.017
Chain 1: 3300 -7413.888 0.015 0.017
Chain 1: 3400 -7625.632 0.017 0.022
Chain 1: 3500 -7397.157 0.019 0.024
Chain 1: 3600 -7461.827 0.018 0.024
Chain 1: 3700 -7411.288 0.019 0.024
Chain 1: 3800 -7413.879 0.018 0.024
Chain 1: 3900 -7379.819 0.016 0.017
Chain 1: 4000 -7374.701 0.014 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003706 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 37.06 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86379.720 1.000 1.000
Chain 1: 200 -13621.678 3.171 5.341
Chain 1: 300 -9959.739 2.236 1.000
Chain 1: 400 -10791.810 1.697 1.000
Chain 1: 500 -8959.007 1.398 0.368
Chain 1: 600 -8408.033 1.176 0.368
Chain 1: 700 -8382.036 1.008 0.205
Chain 1: 800 -8748.797 0.888 0.205
Chain 1: 900 -8685.022 0.790 0.077
Chain 1: 1000 -8729.916 0.711 0.077
Chain 1: 1100 -8771.667 0.612 0.066 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8256.086 0.084 0.062
Chain 1: 1300 -8637.120 0.052 0.044
Chain 1: 1400 -8635.241 0.044 0.042
Chain 1: 1500 -8502.904 0.025 0.016
Chain 1: 1600 -8611.047 0.020 0.013
Chain 1: 1700 -8684.688 0.020 0.013
Chain 1: 1800 -8257.981 0.021 0.013
Chain 1: 1900 -8360.297 0.022 0.013
Chain 1: 2000 -8335.054 0.022 0.013
Chain 1: 2100 -8462.012 0.023 0.015
Chain 1: 2200 -8261.286 0.019 0.015
Chain 1: 2300 -8355.522 0.015 0.013
Chain 1: 2400 -8423.548 0.016 0.013
Chain 1: 2500 -8369.755 0.015 0.012
Chain 1: 2600 -8372.082 0.014 0.011
Chain 1: 2700 -8288.352 0.014 0.011
Chain 1: 2800 -8247.021 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003459 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.59 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8396537.964 1.000 1.000
Chain 1: 200 -1583049.165 2.652 4.304
Chain 1: 300 -890780.286 2.027 1.000
Chain 1: 400 -457672.157 1.757 1.000
Chain 1: 500 -358150.279 1.461 0.946
Chain 1: 600 -233199.179 1.307 0.946
Chain 1: 700 -119402.215 1.256 0.946
Chain 1: 800 -86616.335 1.147 0.946
Chain 1: 900 -66951.765 1.052 0.777
Chain 1: 1000 -51744.198 0.976 0.777
Chain 1: 1100 -39214.425 0.908 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38391.177 0.480 0.379
Chain 1: 1300 -26336.236 0.448 0.379
Chain 1: 1400 -26055.553 0.354 0.320
Chain 1: 1500 -22639.497 0.342 0.320
Chain 1: 1600 -21855.231 0.292 0.294
Chain 1: 1700 -20727.298 0.202 0.294
Chain 1: 1800 -20671.184 0.164 0.151
Chain 1: 1900 -20997.484 0.136 0.054
Chain 1: 2000 -19507.380 0.115 0.054
Chain 1: 2100 -19745.940 0.084 0.036
Chain 1: 2200 -19972.626 0.083 0.036
Chain 1: 2300 -19589.570 0.039 0.020
Chain 1: 2400 -19361.522 0.039 0.020
Chain 1: 2500 -19163.616 0.025 0.016
Chain 1: 2600 -18793.660 0.023 0.016
Chain 1: 2700 -18750.542 0.018 0.012
Chain 1: 2800 -18467.363 0.019 0.015
Chain 1: 2900 -18748.681 0.019 0.015
Chain 1: 3000 -18734.891 0.012 0.012
Chain 1: 3100 -18819.902 0.011 0.012
Chain 1: 3200 -18510.489 0.012 0.015
Chain 1: 3300 -18715.268 0.011 0.012
Chain 1: 3400 -18190.042 0.012 0.015
Chain 1: 3500 -18802.201 0.015 0.015
Chain 1: 3600 -18108.472 0.017 0.015
Chain 1: 3700 -18495.578 0.018 0.017
Chain 1: 3800 -17454.724 0.023 0.021
Chain 1: 3900 -17450.837 0.021 0.021
Chain 1: 4000 -17568.150 0.022 0.021
Chain 1: 4100 -17481.890 0.022 0.021
Chain 1: 4200 -17297.982 0.021 0.021
Chain 1: 4300 -17436.495 0.021 0.021
Chain 1: 4400 -17393.197 0.018 0.011
Chain 1: 4500 -17295.691 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001357 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.57 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48790.108 1.000 1.000
Chain 1: 200 -26956.129 0.905 1.000
Chain 1: 300 -16629.369 0.810 0.810
Chain 1: 400 -20757.505 0.657 0.810
Chain 1: 500 -14174.359 0.619 0.621
Chain 1: 600 -13607.717 0.523 0.621
Chain 1: 700 -11187.376 0.479 0.464
Chain 1: 800 -14583.303 0.448 0.464
Chain 1: 900 -11124.986 0.433 0.311
Chain 1: 1000 -13180.747 0.405 0.311
Chain 1: 1100 -17284.947 0.329 0.237
Chain 1: 1200 -10793.911 0.308 0.237
Chain 1: 1300 -21921.679 0.297 0.237
Chain 1: 1400 -13176.373 0.343 0.311
Chain 1: 1500 -11431.880 0.312 0.237
Chain 1: 1600 -10308.799 0.319 0.237
Chain 1: 1700 -12546.242 0.315 0.237
Chain 1: 1800 -11196.702 0.304 0.237
Chain 1: 1900 -10334.562 0.281 0.178
Chain 1: 2000 -9416.485 0.275 0.178
Chain 1: 2100 -9801.620 0.255 0.153
Chain 1: 2200 -10068.948 0.198 0.121
Chain 1: 2300 -10443.444 0.151 0.109
Chain 1: 2400 -9634.780 0.093 0.097
Chain 1: 2500 -10735.205 0.088 0.097
Chain 1: 2600 -9910.112 0.085 0.084
Chain 1: 2700 -9411.240 0.073 0.083
Chain 1: 2800 -14296.892 0.095 0.083
Chain 1: 2900 -16133.126 0.098 0.084
Chain 1: 3000 -10839.210 0.137 0.084
Chain 1: 3100 -8675.859 0.158 0.103
Chain 1: 3200 -8645.909 0.156 0.103
Chain 1: 3300 -8979.405 0.156 0.103
Chain 1: 3400 -8580.441 0.152 0.103
Chain 1: 3500 -9558.461 0.152 0.102
Chain 1: 3600 -9202.373 0.147 0.102
Chain 1: 3700 -8523.144 0.150 0.102
Chain 1: 3800 -12501.776 0.148 0.102
Chain 1: 3900 -8441.827 0.184 0.102
Chain 1: 4000 -8685.281 0.138 0.080
Chain 1: 4100 -9207.661 0.119 0.057
Chain 1: 4200 -8697.672 0.125 0.059
Chain 1: 4300 -9643.658 0.131 0.080
Chain 1: 4400 -9640.221 0.126 0.080
Chain 1: 4500 -12481.245 0.139 0.080
Chain 1: 4600 -12721.687 0.137 0.080
Chain 1: 4700 -8707.921 0.175 0.098
Chain 1: 4800 -8728.941 0.143 0.059
Chain 1: 4900 -12308.936 0.124 0.059
Chain 1: 5000 -9157.786 0.156 0.098
Chain 1: 5100 -8440.565 0.159 0.098
Chain 1: 5200 -10239.696 0.170 0.176
Chain 1: 5300 -12512.204 0.179 0.182
Chain 1: 5400 -9814.028 0.206 0.228
Chain 1: 5500 -10437.210 0.189 0.182
Chain 1: 5600 -8970.579 0.204 0.182
Chain 1: 5700 -9275.856 0.161 0.176
Chain 1: 5800 -8682.620 0.168 0.176
Chain 1: 5900 -8945.699 0.142 0.163
Chain 1: 6000 -8557.644 0.112 0.085
Chain 1: 6100 -8171.148 0.108 0.068
Chain 1: 6200 -8213.657 0.091 0.060
Chain 1: 6300 -8199.565 0.073 0.047
Chain 1: 6400 -11637.127 0.075 0.047
Chain 1: 6500 -11416.927 0.071 0.045
Chain 1: 6600 -8434.610 0.090 0.045
Chain 1: 6700 -9580.198 0.099 0.047
Chain 1: 6800 -8592.000 0.103 0.047
Chain 1: 6900 -9839.986 0.113 0.115
Chain 1: 7000 -8643.332 0.122 0.120
Chain 1: 7100 -12572.696 0.149 0.127
Chain 1: 7200 -10580.363 0.167 0.138
Chain 1: 7300 -8364.483 0.193 0.188
Chain 1: 7400 -8546.662 0.166 0.138
Chain 1: 7500 -10667.201 0.184 0.188
Chain 1: 7600 -8312.743 0.177 0.188
Chain 1: 7700 -10876.581 0.189 0.199
Chain 1: 7800 -9660.629 0.190 0.199
Chain 1: 7900 -8311.843 0.193 0.199
Chain 1: 8000 -10631.867 0.201 0.218
Chain 1: 8100 -8476.670 0.195 0.218
Chain 1: 8200 -8693.239 0.179 0.218
Chain 1: 8300 -8070.659 0.160 0.199
Chain 1: 8400 -8209.529 0.160 0.199
Chain 1: 8500 -8166.302 0.140 0.162
Chain 1: 8600 -10974.241 0.138 0.162
Chain 1: 8700 -9623.426 0.128 0.140
Chain 1: 8800 -8434.431 0.130 0.141
Chain 1: 8900 -8321.825 0.115 0.140
Chain 1: 9000 -8366.789 0.093 0.077
Chain 1: 9100 -8012.690 0.072 0.044
Chain 1: 9200 -10144.763 0.091 0.077
Chain 1: 9300 -8236.323 0.106 0.140
Chain 1: 9400 -11477.211 0.133 0.141
Chain 1: 9500 -10499.546 0.142 0.141
Chain 1: 9600 -10186.541 0.119 0.140
Chain 1: 9700 -8775.359 0.121 0.141
Chain 1: 9800 -8709.895 0.108 0.093
Chain 1: 9900 -10350.546 0.122 0.159
Chain 1: 10000 -8096.943 0.150 0.161
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001382 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.82 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61742.055 1.000 1.000
Chain 1: 200 -17551.239 1.759 2.518
Chain 1: 300 -8731.019 1.509 1.010
Chain 1: 400 -9038.993 1.141 1.010
Chain 1: 500 -7896.588 0.941 1.000
Chain 1: 600 -8440.659 0.795 1.000
Chain 1: 700 -8235.254 0.685 0.145
Chain 1: 800 -8132.889 0.601 0.145
Chain 1: 900 -7863.450 0.538 0.064
Chain 1: 1000 -7845.089 0.485 0.064
Chain 1: 1100 -7694.603 0.386 0.034
Chain 1: 1200 -7644.298 0.135 0.034
Chain 1: 1300 -7707.348 0.035 0.025
Chain 1: 1400 -7888.274 0.034 0.023
Chain 1: 1500 -7622.350 0.023 0.023
Chain 1: 1600 -7547.701 0.018 0.020
Chain 1: 1700 -7557.420 0.015 0.013
Chain 1: 1800 -7580.931 0.014 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003457 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.57 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -84950.707 1.000 1.000
Chain 1: 200 -13196.191 3.219 5.438
Chain 1: 300 -9681.046 2.267 1.000
Chain 1: 400 -10501.856 1.720 1.000
Chain 1: 500 -8621.457 1.419 0.363
Chain 1: 600 -8228.781 1.191 0.363
Chain 1: 700 -8342.515 1.023 0.218
Chain 1: 800 -8964.696 0.903 0.218
Chain 1: 900 -8522.036 0.809 0.078
Chain 1: 1000 -8286.558 0.731 0.078
Chain 1: 1100 -8555.214 0.634 0.069 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8209.232 0.094 0.052
Chain 1: 1300 -8427.985 0.061 0.048
Chain 1: 1400 -8436.021 0.053 0.042
Chain 1: 1500 -8316.341 0.033 0.031
Chain 1: 1600 -8410.297 0.029 0.028
Chain 1: 1700 -8502.992 0.029 0.028
Chain 1: 1800 -8118.925 0.026 0.028
Chain 1: 1900 -8220.854 0.023 0.026
Chain 1: 2000 -8190.705 0.020 0.014
Chain 1: 2100 -8327.511 0.019 0.014
Chain 1: 2200 -8110.285 0.017 0.014
Chain 1: 2300 -8251.941 0.016 0.014
Chain 1: 2400 -8261.085 0.016 0.014
Chain 1: 2500 -8229.407 0.015 0.012
Chain 1: 2600 -8226.644 0.014 0.012
Chain 1: 2700 -8136.785 0.014 0.012
Chain 1: 2800 -8115.963 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003723 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 37.23 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8402641.893 1.000 1.000
Chain 1: 200 -1582223.460 2.655 4.311
Chain 1: 300 -890018.110 2.029 1.000
Chain 1: 400 -457375.705 1.759 1.000
Chain 1: 500 -357849.876 1.462 0.946
Chain 1: 600 -232862.890 1.308 0.946
Chain 1: 700 -118991.309 1.258 0.946
Chain 1: 800 -86212.221 1.148 0.946
Chain 1: 900 -66524.317 1.054 0.778
Chain 1: 1000 -51295.940 0.978 0.778
Chain 1: 1100 -38760.173 0.910 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37928.373 0.481 0.380
Chain 1: 1300 -25876.223 0.450 0.380
Chain 1: 1400 -25592.161 0.357 0.323
Chain 1: 1500 -22178.773 0.344 0.323
Chain 1: 1600 -21394.680 0.294 0.297
Chain 1: 1700 -20267.678 0.204 0.296
Chain 1: 1800 -20211.512 0.166 0.154
Chain 1: 1900 -20537.044 0.138 0.056
Chain 1: 2000 -19049.425 0.117 0.056
Chain 1: 2100 -19287.395 0.085 0.037
Chain 1: 2200 -19513.671 0.084 0.037
Chain 1: 2300 -19131.270 0.040 0.020
Chain 1: 2400 -18903.602 0.040 0.020
Chain 1: 2500 -18705.865 0.026 0.016
Chain 1: 2600 -18336.383 0.024 0.016
Chain 1: 2700 -18293.512 0.019 0.012
Chain 1: 2800 -18010.736 0.020 0.016
Chain 1: 2900 -18291.748 0.020 0.015
Chain 1: 3000 -18277.874 0.012 0.012
Chain 1: 3100 -18362.773 0.011 0.012
Chain 1: 3200 -18053.813 0.012 0.015
Chain 1: 3300 -18258.296 0.011 0.012
Chain 1: 3400 -17733.932 0.013 0.015
Chain 1: 3500 -18344.750 0.015 0.016
Chain 1: 3600 -17652.860 0.017 0.016
Chain 1: 3700 -18038.603 0.019 0.017
Chain 1: 3800 -17000.554 0.023 0.021
Chain 1: 3900 -16996.820 0.022 0.021
Chain 1: 4000 -17114.065 0.022 0.021
Chain 1: 4100 -17027.938 0.022 0.021
Chain 1: 4200 -16844.719 0.022 0.021
Chain 1: 4300 -16982.696 0.022 0.021
Chain 1: 4400 -16939.900 0.019 0.011
Chain 1: 4500 -16842.583 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001299 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.99 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12423.729 1.000 1.000
Chain 1: 200 -9257.514 0.671 1.000
Chain 1: 300 -8168.459 0.492 0.342
Chain 1: 400 -8249.532 0.371 0.342
Chain 1: 500 -8231.579 0.297 0.133
Chain 1: 600 -8028.963 0.252 0.133
Chain 1: 700 -7922.009 0.218 0.025
Chain 1: 800 -7943.181 0.191 0.025
Chain 1: 900 -7914.281 0.170 0.014
Chain 1: 1000 -8093.543 0.155 0.022
Chain 1: 1100 -8030.336 0.056 0.014
Chain 1: 1200 -7945.893 0.023 0.011
Chain 1: 1300 -7914.372 0.010 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001642 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 16.42 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -62011.158 1.000 1.000
Chain 1: 200 -17868.921 1.735 2.470
Chain 1: 300 -8890.004 1.493 1.010
Chain 1: 400 -8406.922 1.134 1.010
Chain 1: 500 -8498.713 0.910 1.000
Chain 1: 600 -8655.404 0.761 1.000
Chain 1: 700 -8176.983 0.661 0.059
Chain 1: 800 -8169.281 0.578 0.059
Chain 1: 900 -7904.820 0.518 0.057
Chain 1: 1000 -7893.333 0.466 0.057
Chain 1: 1100 -7876.015 0.366 0.033
Chain 1: 1200 -7611.947 0.123 0.033
Chain 1: 1300 -7805.817 0.024 0.025
Chain 1: 1400 -7640.393 0.021 0.022
Chain 1: 1500 -7570.683 0.021 0.022
Chain 1: 1600 -7758.325 0.021 0.024
Chain 1: 1700 -7484.505 0.019 0.024
Chain 1: 1800 -7633.472 0.021 0.024
Chain 1: 1900 -7629.485 0.017 0.022
Chain 1: 2000 -7581.061 0.018 0.022
Chain 1: 2100 -7570.698 0.018 0.022
Chain 1: 2200 -7698.811 0.016 0.020
Chain 1: 2300 -7588.655 0.015 0.017
Chain 1: 2400 -7629.852 0.013 0.015
Chain 1: 2500 -7560.047 0.013 0.015
Chain 1: 2600 -7534.282 0.011 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003528 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.28 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86198.441 1.000 1.000
Chain 1: 200 -13545.747 3.182 5.364
Chain 1: 300 -9932.796 2.242 1.000
Chain 1: 400 -10849.584 1.703 1.000
Chain 1: 500 -8891.005 1.406 0.364
Chain 1: 600 -8626.351 1.177 0.364
Chain 1: 700 -8446.803 1.012 0.220
Chain 1: 800 -8877.233 0.892 0.220
Chain 1: 900 -8694.271 0.795 0.084
Chain 1: 1000 -8457.086 0.718 0.084
Chain 1: 1100 -8696.175 0.621 0.048 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8418.390 0.088 0.033
Chain 1: 1300 -8626.175 0.054 0.031
Chain 1: 1400 -8629.659 0.045 0.028
Chain 1: 1500 -8484.206 0.025 0.027
Chain 1: 1600 -8597.406 0.023 0.024
Chain 1: 1700 -8681.616 0.022 0.024
Chain 1: 1800 -8269.553 0.022 0.024
Chain 1: 1900 -8365.563 0.021 0.024
Chain 1: 2000 -8338.782 0.019 0.017
Chain 1: 2100 -8461.101 0.018 0.014
Chain 1: 2200 -8280.912 0.017 0.014
Chain 1: 2300 -8360.724 0.015 0.013
Chain 1: 2400 -8430.320 0.016 0.013
Chain 1: 2500 -8375.633 0.015 0.011
Chain 1: 2600 -8374.905 0.013 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003514 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.14 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8414803.887 1.000 1.000
Chain 1: 200 -1583152.936 2.658 4.315
Chain 1: 300 -889546.194 2.032 1.000
Chain 1: 400 -457539.080 1.760 1.000
Chain 1: 500 -357685.060 1.464 0.944
Chain 1: 600 -232799.816 1.309 0.944
Chain 1: 700 -119114.948 1.258 0.944
Chain 1: 800 -86400.576 1.148 0.944
Chain 1: 900 -66763.914 1.054 0.780
Chain 1: 1000 -51577.497 0.978 0.780
Chain 1: 1100 -39077.737 0.910 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38256.284 0.480 0.379
Chain 1: 1300 -26229.643 0.448 0.379
Chain 1: 1400 -25951.433 0.355 0.320
Chain 1: 1500 -22544.067 0.342 0.320
Chain 1: 1600 -21762.779 0.292 0.294
Chain 1: 1700 -20637.966 0.202 0.294
Chain 1: 1800 -20582.814 0.164 0.151
Chain 1: 1900 -20908.883 0.136 0.055
Chain 1: 2000 -19421.379 0.115 0.055
Chain 1: 2100 -19659.487 0.084 0.036
Chain 1: 2200 -19885.907 0.083 0.036
Chain 1: 2300 -19503.201 0.039 0.020
Chain 1: 2400 -19275.319 0.039 0.020
Chain 1: 2500 -19077.480 0.025 0.016
Chain 1: 2600 -18707.517 0.023 0.016
Chain 1: 2700 -18664.509 0.018 0.012
Chain 1: 2800 -18381.413 0.019 0.015
Chain 1: 2900 -18662.661 0.019 0.015
Chain 1: 3000 -18648.787 0.012 0.012
Chain 1: 3100 -18733.779 0.011 0.012
Chain 1: 3200 -18424.448 0.012 0.015
Chain 1: 3300 -18629.223 0.011 0.012
Chain 1: 3400 -18104.179 0.012 0.015
Chain 1: 3500 -18715.966 0.015 0.015
Chain 1: 3600 -18022.749 0.017 0.015
Chain 1: 3700 -18409.448 0.018 0.017
Chain 1: 3800 -17369.339 0.023 0.021
Chain 1: 3900 -17365.511 0.021 0.021
Chain 1: 4000 -17482.797 0.022 0.021
Chain 1: 4100 -17396.544 0.022 0.021
Chain 1: 4200 -17212.867 0.021 0.021
Chain 1: 4300 -17351.193 0.021 0.021
Chain 1: 4400 -17308.020 0.019 0.011
Chain 1: 4500 -17210.591 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001267 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.67 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49265.325 1.000 1.000
Chain 1: 200 -19494.679 1.264 1.527
Chain 1: 300 -18335.159 0.863 1.000
Chain 1: 400 -21229.105 0.682 1.000
Chain 1: 500 -13059.709 0.670 0.626
Chain 1: 600 -14629.267 0.577 0.626
Chain 1: 700 -13851.497 0.502 0.136
Chain 1: 800 -10991.270 0.472 0.260
Chain 1: 900 -14486.771 0.446 0.241
Chain 1: 1000 -10467.834 0.440 0.260
Chain 1: 1100 -11833.836 0.352 0.241
Chain 1: 1200 -10522.311 0.211 0.136
Chain 1: 1300 -11512.261 0.214 0.136
Chain 1: 1400 -20890.004 0.245 0.241
Chain 1: 1500 -11637.786 0.262 0.241
Chain 1: 1600 -12725.403 0.260 0.241
Chain 1: 1700 -17220.039 0.280 0.260
Chain 1: 1800 -10968.614 0.311 0.261
Chain 1: 1900 -10962.402 0.287 0.261
Chain 1: 2000 -21214.252 0.297 0.261
Chain 1: 2100 -10254.537 0.392 0.449
Chain 1: 2200 -10624.156 0.383 0.449
Chain 1: 2300 -11560.554 0.383 0.449
Chain 1: 2400 -9661.259 0.358 0.261
Chain 1: 2500 -9831.997 0.280 0.197
Chain 1: 2600 -9883.039 0.272 0.197
Chain 1: 2700 -12151.544 0.264 0.187
Chain 1: 2800 -10881.714 0.219 0.117
Chain 1: 2900 -9383.049 0.235 0.160
Chain 1: 3000 -9405.873 0.187 0.117
Chain 1: 3100 -10265.513 0.088 0.084
Chain 1: 3200 -12644.914 0.104 0.117
Chain 1: 3300 -13143.376 0.099 0.117
Chain 1: 3400 -10257.927 0.108 0.117
Chain 1: 3500 -9680.960 0.112 0.117
Chain 1: 3600 -10151.091 0.116 0.117
Chain 1: 3700 -11305.299 0.108 0.102
Chain 1: 3800 -9245.263 0.118 0.102
Chain 1: 3900 -15953.704 0.144 0.102
Chain 1: 4000 -14267.071 0.156 0.118
Chain 1: 4100 -9354.985 0.200 0.188
Chain 1: 4200 -9605.522 0.184 0.118
Chain 1: 4300 -10192.256 0.186 0.118
Chain 1: 4400 -13671.837 0.183 0.118
Chain 1: 4500 -9408.705 0.223 0.223
Chain 1: 4600 -9010.385 0.222 0.223
Chain 1: 4700 -8990.657 0.212 0.223
Chain 1: 4800 -8739.017 0.193 0.118
Chain 1: 4900 -8940.271 0.153 0.058
Chain 1: 5000 -13517.913 0.175 0.058
Chain 1: 5100 -10117.788 0.156 0.058
Chain 1: 5200 -18613.227 0.199 0.255
Chain 1: 5300 -9576.245 0.288 0.336
Chain 1: 5400 -10683.192 0.273 0.336
Chain 1: 5500 -13834.430 0.250 0.228
Chain 1: 5600 -9968.739 0.285 0.336
Chain 1: 5700 -15320.803 0.319 0.339
Chain 1: 5800 -8869.914 0.389 0.349
Chain 1: 5900 -12843.975 0.418 0.349
Chain 1: 6000 -8965.157 0.427 0.388
Chain 1: 6100 -10416.955 0.408 0.388
Chain 1: 6200 -8808.642 0.380 0.349
Chain 1: 6300 -8985.354 0.288 0.309
Chain 1: 6400 -8900.897 0.279 0.309
Chain 1: 6500 -9158.137 0.259 0.309
Chain 1: 6600 -8766.030 0.224 0.183
Chain 1: 6700 -8612.707 0.191 0.139
Chain 1: 6800 -11186.618 0.141 0.139
Chain 1: 6900 -9198.207 0.132 0.139
Chain 1: 7000 -13683.757 0.122 0.139
Chain 1: 7100 -8957.202 0.160 0.183
Chain 1: 7200 -8731.915 0.145 0.045
Chain 1: 7300 -8850.570 0.144 0.045
Chain 1: 7400 -8519.750 0.147 0.045
Chain 1: 7500 -9544.970 0.155 0.107
Chain 1: 7600 -8879.920 0.158 0.107
Chain 1: 7700 -8718.331 0.158 0.107
Chain 1: 7800 -9061.241 0.139 0.075
Chain 1: 7900 -11474.889 0.138 0.075
Chain 1: 8000 -8607.767 0.139 0.075
Chain 1: 8100 -8811.340 0.088 0.039
Chain 1: 8200 -8413.273 0.090 0.047
Chain 1: 8300 -8577.371 0.091 0.047
Chain 1: 8400 -16988.037 0.137 0.075
Chain 1: 8500 -8488.055 0.226 0.075
Chain 1: 8600 -10931.996 0.241 0.210
Chain 1: 8700 -10328.710 0.245 0.210
Chain 1: 8800 -11703.488 0.253 0.210
Chain 1: 8900 -9451.386 0.256 0.224
Chain 1: 9000 -11711.582 0.242 0.193
Chain 1: 9100 -8986.552 0.270 0.224
Chain 1: 9200 -8475.984 0.271 0.224
Chain 1: 9300 -8265.689 0.272 0.224
Chain 1: 9400 -12505.094 0.256 0.224
Chain 1: 9500 -8425.846 0.204 0.224
Chain 1: 9600 -10926.014 0.205 0.229
Chain 1: 9700 -12263.847 0.210 0.229
Chain 1: 9800 -8699.706 0.239 0.238
Chain 1: 9900 -8963.912 0.218 0.229
Chain 1: 10000 -8459.650 0.205 0.229
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00164 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 16.4 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -59009.830 1.000 1.000
Chain 1: 200 -18343.658 1.608 2.217
Chain 1: 300 -9009.537 1.418 1.036
Chain 1: 400 -8244.453 1.086 1.036
Chain 1: 500 -8618.820 0.878 1.000
Chain 1: 600 -9274.394 0.743 1.000
Chain 1: 700 -8562.389 0.649 0.093
Chain 1: 800 -8446.613 0.570 0.093
Chain 1: 900 -7970.333 0.513 0.083
Chain 1: 1000 -7888.092 0.463 0.083
Chain 1: 1100 -7865.044 0.363 0.071
Chain 1: 1200 -7868.581 0.141 0.060
Chain 1: 1300 -7797.267 0.039 0.043
Chain 1: 1400 -7989.050 0.032 0.024
Chain 1: 1500 -7634.964 0.032 0.024
Chain 1: 1600 -7924.351 0.029 0.024
Chain 1: 1700 -7707.392 0.023 0.024
Chain 1: 1800 -7669.566 0.022 0.024
Chain 1: 1900 -7778.266 0.018 0.014
Chain 1: 2000 -7789.672 0.017 0.014
Chain 1: 2100 -7666.641 0.018 0.016
Chain 1: 2200 -7874.004 0.021 0.024
Chain 1: 2300 -7671.409 0.022 0.026
Chain 1: 2400 -7754.691 0.021 0.026
Chain 1: 2500 -7722.113 0.017 0.016
Chain 1: 2600 -7624.567 0.015 0.014
Chain 1: 2700 -7614.540 0.012 0.013
Chain 1: 2800 -7611.181 0.011 0.013
Chain 1: 2900 -7461.462 0.012 0.013
Chain 1: 3000 -7615.000 0.014 0.016
Chain 1: 3100 -7614.590 0.012 0.013
Chain 1: 3200 -7836.228 0.012 0.013
Chain 1: 3300 -7553.977 0.014 0.013
Chain 1: 3400 -7796.507 0.016 0.020
Chain 1: 3500 -7534.548 0.019 0.020
Chain 1: 3600 -7593.627 0.018 0.020
Chain 1: 3700 -7540.714 0.019 0.020
Chain 1: 3800 -7558.054 0.019 0.020
Chain 1: 3900 -7517.578 0.017 0.020
Chain 1: 4000 -7490.542 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003849 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 38.49 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86262.590 1.000 1.000
Chain 1: 200 -14047.983 3.070 5.141
Chain 1: 300 -10228.804 2.171 1.000
Chain 1: 400 -12421.759 1.673 1.000
Chain 1: 500 -8606.604 1.427 0.443
Chain 1: 600 -8885.247 1.194 0.443
Chain 1: 700 -8921.447 1.024 0.373
Chain 1: 800 -8972.411 0.897 0.373
Chain 1: 900 -8771.946 0.800 0.177
Chain 1: 1000 -8620.103 0.722 0.177
Chain 1: 1100 -8738.124 0.623 0.031 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8423.444 0.113 0.031
Chain 1: 1300 -8874.308 0.080 0.031
Chain 1: 1400 -8695.354 0.065 0.023
Chain 1: 1500 -8656.412 0.021 0.021
Chain 1: 1600 -8736.784 0.019 0.018
Chain 1: 1700 -8792.931 0.019 0.018
Chain 1: 1800 -8337.505 0.024 0.021
Chain 1: 1900 -8438.813 0.023 0.018
Chain 1: 2000 -8459.088 0.021 0.014
Chain 1: 2100 -8561.602 0.021 0.012
Chain 1: 2200 -8317.881 0.020 0.012
Chain 1: 2300 -8506.846 0.017 0.012
Chain 1: 2400 -8353.231 0.017 0.012
Chain 1: 2500 -8409.752 0.017 0.012
Chain 1: 2600 -8315.974 0.018 0.012
Chain 1: 2700 -8351.353 0.017 0.012
Chain 1: 2800 -8311.571 0.012 0.012
Chain 1: 2900 -8418.548 0.012 0.012
Chain 1: 3000 -8326.438 0.013 0.012
Chain 1: 3100 -8293.405 0.012 0.011
Chain 1: 3200 -8261.946 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003309 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.09 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8393387.736 1.000 1.000
Chain 1: 200 -1579822.673 2.656 4.313
Chain 1: 300 -889923.483 2.029 1.000
Chain 1: 400 -458274.533 1.758 1.000
Chain 1: 500 -358989.842 1.461 0.942
Chain 1: 600 -234026.376 1.307 0.942
Chain 1: 700 -120046.697 1.256 0.942
Chain 1: 800 -87240.307 1.146 0.942
Chain 1: 900 -67539.295 1.051 0.775
Chain 1: 1000 -52317.164 0.975 0.775
Chain 1: 1100 -39765.540 0.906 0.534 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38946.356 0.477 0.376
Chain 1: 1300 -26848.363 0.445 0.376
Chain 1: 1400 -26567.176 0.352 0.316
Chain 1: 1500 -23140.539 0.339 0.316
Chain 1: 1600 -22354.952 0.289 0.292
Chain 1: 1700 -21220.365 0.199 0.291
Chain 1: 1800 -21163.443 0.162 0.148
Chain 1: 1900 -21490.647 0.134 0.053
Chain 1: 2000 -19995.981 0.113 0.053
Chain 1: 2100 -20234.484 0.082 0.035
Chain 1: 2200 -20462.533 0.081 0.035
Chain 1: 2300 -20078.112 0.038 0.019
Chain 1: 2400 -19849.761 0.038 0.019
Chain 1: 2500 -19652.167 0.024 0.015
Chain 1: 2600 -19280.868 0.023 0.015
Chain 1: 2700 -19237.434 0.018 0.012
Chain 1: 2800 -18954.005 0.019 0.015
Chain 1: 2900 -19235.803 0.019 0.015
Chain 1: 3000 -19221.797 0.012 0.012
Chain 1: 3100 -19306.998 0.011 0.012
Chain 1: 3200 -18996.892 0.011 0.015
Chain 1: 3300 -19202.250 0.010 0.012
Chain 1: 3400 -18675.929 0.012 0.015
Chain 1: 3500 -19289.804 0.014 0.015
Chain 1: 3600 -18593.891 0.016 0.015
Chain 1: 3700 -18982.667 0.018 0.016
Chain 1: 3800 -17938.452 0.022 0.020
Chain 1: 3900 -17934.569 0.021 0.020
Chain 1: 4000 -18051.810 0.021 0.020
Chain 1: 4100 -17965.417 0.021 0.020
Chain 1: 4200 -17780.804 0.021 0.020
Chain 1: 4300 -17919.761 0.021 0.020
Chain 1: 4400 -17875.864 0.018 0.010
Chain 1: 4500 -17778.331 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001382 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.82 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12490.906 1.000 1.000
Chain 1: 200 -9312.548 0.671 1.000
Chain 1: 300 -8124.501 0.496 0.341
Chain 1: 400 -8275.792 0.376 0.341
Chain 1: 500 -8262.306 0.301 0.146
Chain 1: 600 -8035.857 0.256 0.146
Chain 1: 700 -7934.641 0.221 0.028
Chain 1: 800 -7964.674 0.194 0.028
Chain 1: 900 -8070.242 0.174 0.018
Chain 1: 1000 -7984.299 0.158 0.018
Chain 1: 1100 -7977.599 0.058 0.013
Chain 1: 1200 -7953.094 0.024 0.013
Chain 1: 1300 -7907.501 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001645 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 16.45 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -46098.047 1.000 1.000
Chain 1: 200 -15749.229 1.464 1.927
Chain 1: 300 -8789.533 1.240 1.000
Chain 1: 400 -8506.390 0.938 1.000
Chain 1: 500 -8720.405 0.755 0.792
Chain 1: 600 -8917.697 0.633 0.792
Chain 1: 700 -8023.019 0.559 0.112
Chain 1: 800 -8265.476 0.492 0.112
Chain 1: 900 -7703.431 0.446 0.073
Chain 1: 1000 -7908.244 0.404 0.073
Chain 1: 1100 -7799.005 0.305 0.033
Chain 1: 1200 -7851.159 0.113 0.029
Chain 1: 1300 -7819.008 0.034 0.026
Chain 1: 1400 -7900.702 0.032 0.025
Chain 1: 1500 -7623.697 0.033 0.026
Chain 1: 1600 -7659.267 0.032 0.026
Chain 1: 1700 -7587.859 0.021 0.014
Chain 1: 1800 -7616.289 0.019 0.010
Chain 1: 1900 -7624.032 0.012 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003303 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.03 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86372.676 1.000 1.000
Chain 1: 200 -13691.361 3.154 5.309
Chain 1: 300 -10001.837 2.226 1.000
Chain 1: 400 -11107.167 1.694 1.000
Chain 1: 500 -8994.440 1.402 0.369
Chain 1: 600 -8452.677 1.179 0.369
Chain 1: 700 -8670.749 1.014 0.235
Chain 1: 800 -9025.898 0.893 0.235
Chain 1: 900 -8798.185 0.796 0.100
Chain 1: 1000 -8748.119 0.717 0.100
Chain 1: 1100 -8585.545 0.619 0.064 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8380.587 0.091 0.039
Chain 1: 1300 -8686.688 0.057 0.035
Chain 1: 1400 -8634.234 0.048 0.026
Chain 1: 1500 -8519.074 0.026 0.025
Chain 1: 1600 -8625.775 0.021 0.024
Chain 1: 1700 -8699.774 0.019 0.019
Chain 1: 1800 -8267.074 0.020 0.019
Chain 1: 1900 -8371.171 0.019 0.014
Chain 1: 2000 -8346.574 0.019 0.014
Chain 1: 2100 -8325.285 0.017 0.012
Chain 1: 2200 -8289.576 0.015 0.012
Chain 1: 2300 -8418.532 0.013 0.012
Chain 1: 2400 -8273.431 0.014 0.012
Chain 1: 2500 -8341.956 0.014 0.012
Chain 1: 2600 -8261.310 0.013 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002965 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.65 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8392547.310 1.000 1.000
Chain 1: 200 -1582440.623 2.652 4.304
Chain 1: 300 -891410.700 2.026 1.000
Chain 1: 400 -458485.767 1.756 1.000
Chain 1: 500 -358966.077 1.460 0.944
Chain 1: 600 -233716.786 1.306 0.944
Chain 1: 700 -119685.546 1.256 0.944
Chain 1: 800 -86845.299 1.146 0.944
Chain 1: 900 -67134.326 1.051 0.775
Chain 1: 1000 -51896.049 0.975 0.775
Chain 1: 1100 -39340.290 0.907 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38515.735 0.479 0.378
Chain 1: 1300 -26432.081 0.447 0.378
Chain 1: 1400 -26148.908 0.354 0.319
Chain 1: 1500 -22726.123 0.341 0.319
Chain 1: 1600 -21940.222 0.291 0.294
Chain 1: 1700 -20808.790 0.201 0.294
Chain 1: 1800 -20752.057 0.164 0.151
Chain 1: 1900 -21078.491 0.136 0.054
Chain 1: 2000 -19586.639 0.114 0.054
Chain 1: 2100 -19825.043 0.084 0.036
Chain 1: 2200 -20052.205 0.083 0.036
Chain 1: 2300 -19668.756 0.039 0.019
Chain 1: 2400 -19440.692 0.039 0.019
Chain 1: 2500 -19242.919 0.025 0.015
Chain 1: 2600 -18872.486 0.023 0.015
Chain 1: 2700 -18829.367 0.018 0.012
Chain 1: 2800 -18546.131 0.019 0.015
Chain 1: 2900 -18827.650 0.019 0.015
Chain 1: 3000 -18813.702 0.012 0.012
Chain 1: 3100 -18898.725 0.011 0.012
Chain 1: 3200 -18589.160 0.012 0.015
Chain 1: 3300 -18794.133 0.011 0.012
Chain 1: 3400 -18268.630 0.012 0.015
Chain 1: 3500 -18881.168 0.015 0.015
Chain 1: 3600 -18187.083 0.016 0.015
Chain 1: 3700 -18574.447 0.018 0.017
Chain 1: 3800 -17532.941 0.023 0.021
Chain 1: 3900 -17529.112 0.021 0.021
Chain 1: 4000 -17646.379 0.022 0.021
Chain 1: 4100 -17560.044 0.022 0.021
Chain 1: 4200 -17376.074 0.021 0.021
Chain 1: 4300 -17514.592 0.021 0.021
Chain 1: 4400 -17471.193 0.018 0.011
Chain 1: 4500 -17373.737 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001278 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.78 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13196.765 1.000 1.000
Chain 1: 200 -9888.373 0.667 1.000
Chain 1: 300 -8975.796 0.479 0.335
Chain 1: 400 -8773.671 0.365 0.335
Chain 1: 500 -8644.291 0.295 0.102
Chain 1: 600 -8274.856 0.253 0.102
Chain 1: 700 -8186.114 0.219 0.045
Chain 1: 800 -8200.912 0.191 0.045
Chain 1: 900 -8267.890 0.171 0.023
Chain 1: 1000 -8240.075 0.154 0.023
Chain 1: 1100 -8231.605 0.054 0.015
Chain 1: 1200 -8182.417 0.022 0.011
Chain 1: 1300 -8133.169 0.012 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001436 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.36 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -51830.010 1.000 1.000
Chain 1: 200 -16999.256 1.524 2.049
Chain 1: 300 -9095.762 1.306 1.000
Chain 1: 400 -9045.157 0.981 1.000
Chain 1: 500 -8345.193 0.801 0.869
Chain 1: 600 -8315.061 0.668 0.869
Chain 1: 700 -8753.741 0.580 0.084
Chain 1: 800 -7812.479 0.523 0.120
Chain 1: 900 -7858.266 0.465 0.084
Chain 1: 1000 -7851.406 0.419 0.084
Chain 1: 1100 -7708.460 0.321 0.050
Chain 1: 1200 -7757.814 0.116 0.019
Chain 1: 1300 -7754.611 0.030 0.006 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003076 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.76 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87430.524 1.000 1.000
Chain 1: 200 -14223.318 3.073 5.147
Chain 1: 300 -10398.802 2.172 1.000
Chain 1: 400 -12250.068 1.666 1.000
Chain 1: 500 -9126.359 1.402 0.368
Chain 1: 600 -9890.840 1.181 0.368
Chain 1: 700 -8955.107 1.027 0.342
Chain 1: 800 -9052.208 0.900 0.342
Chain 1: 900 -9032.593 0.800 0.151
Chain 1: 1000 -9183.853 0.722 0.151
Chain 1: 1100 -8887.568 0.625 0.104 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8615.119 0.114 0.077
Chain 1: 1300 -9010.403 0.081 0.044
Chain 1: 1400 -8904.294 0.067 0.033
Chain 1: 1500 -8886.158 0.033 0.032
Chain 1: 1600 -8932.377 0.026 0.016
Chain 1: 1700 -8989.126 0.016 0.012
Chain 1: 1800 -8521.896 0.021 0.016
Chain 1: 1900 -8643.752 0.022 0.016
Chain 1: 2000 -8664.123 0.021 0.014
Chain 1: 2100 -8750.126 0.018 0.012
Chain 1: 2200 -8525.892 0.018 0.012
Chain 1: 2300 -8733.817 0.016 0.012
Chain 1: 2400 -8534.034 0.017 0.014
Chain 1: 2500 -8612.735 0.018 0.014
Chain 1: 2600 -8518.361 0.018 0.014
Chain 1: 2700 -8557.349 0.018 0.014
Chain 1: 2800 -8509.749 0.013 0.011
Chain 1: 2900 -8623.226 0.013 0.011
Chain 1: 3000 -8532.029 0.014 0.011
Chain 1: 3100 -8499.816 0.013 0.011
Chain 1: 3200 -8470.485 0.011 0.011
Chain 1: 3300 -8735.464 0.012 0.011
Chain 1: 3400 -8783.375 0.010 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00369 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 36.9 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8426897.812 1.000 1.000
Chain 1: 200 -1587124.316 2.655 4.310
Chain 1: 300 -892745.081 2.029 1.000
Chain 1: 400 -459033.522 1.758 1.000
Chain 1: 500 -359135.442 1.462 0.945
Chain 1: 600 -233978.989 1.308 0.945
Chain 1: 700 -120097.729 1.256 0.945
Chain 1: 800 -87250.904 1.146 0.945
Chain 1: 900 -67578.203 1.051 0.778
Chain 1: 1000 -52375.714 0.975 0.778
Chain 1: 1100 -39845.218 0.907 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39028.971 0.478 0.376
Chain 1: 1300 -26969.742 0.445 0.376
Chain 1: 1400 -26691.073 0.351 0.314
Chain 1: 1500 -23272.919 0.338 0.314
Chain 1: 1600 -22488.901 0.288 0.291
Chain 1: 1700 -21360.428 0.199 0.290
Chain 1: 1800 -21304.515 0.161 0.147
Chain 1: 1900 -21631.527 0.134 0.053
Chain 1: 2000 -20139.567 0.112 0.053
Chain 1: 2100 -20378.371 0.082 0.035
Chain 1: 2200 -20605.428 0.081 0.035
Chain 1: 2300 -20221.844 0.038 0.019
Chain 1: 2400 -19993.584 0.038 0.019
Chain 1: 2500 -19795.427 0.024 0.015
Chain 1: 2600 -19424.816 0.023 0.015
Chain 1: 2700 -19381.558 0.018 0.012
Chain 1: 2800 -19097.911 0.019 0.015
Chain 1: 2900 -19379.634 0.019 0.015
Chain 1: 3000 -19365.801 0.011 0.012
Chain 1: 3100 -19450.909 0.011 0.011
Chain 1: 3200 -19140.976 0.011 0.015
Chain 1: 3300 -19346.186 0.010 0.011
Chain 1: 3400 -18819.960 0.012 0.015
Chain 1: 3500 -19433.450 0.014 0.015
Chain 1: 3600 -18738.043 0.016 0.015
Chain 1: 3700 -19126.411 0.018 0.016
Chain 1: 3800 -18082.740 0.022 0.020
Chain 1: 3900 -18078.758 0.021 0.020
Chain 1: 4000 -18196.124 0.021 0.020
Chain 1: 4100 -18109.653 0.021 0.020
Chain 1: 4200 -17925.180 0.021 0.020
Chain 1: 4300 -18064.117 0.020 0.020
Chain 1: 4400 -18020.370 0.018 0.010
Chain 1: 4500 -17922.733 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001447 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.47 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49551.658 1.000 1.000
Chain 1: 200 -22934.376 1.080 1.161
Chain 1: 300 -19946.109 0.770 1.000
Chain 1: 400 -36020.588 0.689 1.000
Chain 1: 500 -13715.574 0.877 1.000
Chain 1: 600 -19860.600 0.782 1.000
Chain 1: 700 -12621.770 0.752 0.574
Chain 1: 800 -13477.047 0.666 0.574
Chain 1: 900 -11036.596 0.617 0.446
Chain 1: 1000 -24656.944 0.610 0.552
Chain 1: 1100 -19341.178 0.538 0.446 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -10612.357 0.504 0.446 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1300 -22031.732 0.541 0.518 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1400 -20123.513 0.506 0.518 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1500 -10609.040 0.433 0.518 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1600 -12229.329 0.415 0.518 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1700 -9736.893 0.383 0.275
Chain 1: 1800 -12942.566 0.402 0.275
Chain 1: 1900 -11356.249 0.394 0.275
Chain 1: 2000 -10363.990 0.348 0.256
Chain 1: 2100 -10420.582 0.321 0.248
Chain 1: 2200 -14824.680 0.268 0.248
Chain 1: 2300 -10168.380 0.262 0.248
Chain 1: 2400 -9226.768 0.263 0.248
Chain 1: 2500 -9552.065 0.177 0.140
Chain 1: 2600 -9881.122 0.167 0.140
Chain 1: 2700 -9325.350 0.147 0.102
Chain 1: 2800 -9906.584 0.128 0.096
Chain 1: 2900 -10198.812 0.117 0.060
Chain 1: 3000 -18285.212 0.152 0.060
Chain 1: 3100 -9431.983 0.245 0.102
Chain 1: 3200 -8996.315 0.220 0.060
Chain 1: 3300 -16134.963 0.219 0.060
Chain 1: 3400 -9943.995 0.271 0.060
Chain 1: 3500 -9178.552 0.276 0.083
Chain 1: 3600 -10273.540 0.283 0.107
Chain 1: 3700 -9273.360 0.288 0.108
Chain 1: 3800 -9225.417 0.283 0.108
Chain 1: 3900 -9419.956 0.282 0.108
Chain 1: 4000 -8695.344 0.246 0.107
Chain 1: 4100 -8919.603 0.155 0.083
Chain 1: 4200 -9863.018 0.159 0.096
Chain 1: 4300 -13117.408 0.140 0.096
Chain 1: 4400 -9207.050 0.120 0.096
Chain 1: 4500 -8932.965 0.115 0.096
Chain 1: 4600 -8631.630 0.108 0.083
Chain 1: 4700 -9412.764 0.105 0.083
Chain 1: 4800 -12517.074 0.129 0.083
Chain 1: 4900 -10602.699 0.145 0.096
Chain 1: 5000 -10961.403 0.140 0.096
Chain 1: 5100 -15964.332 0.169 0.181
Chain 1: 5200 -10472.410 0.212 0.248
Chain 1: 5300 -10844.157 0.191 0.181
Chain 1: 5400 -14714.387 0.174 0.181
Chain 1: 5500 -10233.580 0.215 0.248
Chain 1: 5600 -14466.463 0.241 0.263
Chain 1: 5700 -10064.853 0.276 0.293
Chain 1: 5800 -9058.744 0.263 0.293
Chain 1: 5900 -15689.560 0.287 0.313
Chain 1: 6000 -10098.752 0.339 0.423
Chain 1: 6100 -12152.625 0.325 0.423
Chain 1: 6200 -8385.389 0.317 0.423
Chain 1: 6300 -10236.132 0.332 0.423
Chain 1: 6400 -9922.708 0.309 0.423
Chain 1: 6500 -8722.692 0.279 0.293
Chain 1: 6600 -8447.842 0.253 0.181
Chain 1: 6700 -11892.362 0.238 0.181
Chain 1: 6800 -9650.627 0.250 0.232
Chain 1: 6900 -8639.520 0.219 0.181
Chain 1: 7000 -13900.047 0.202 0.181
Chain 1: 7100 -8258.182 0.253 0.232
Chain 1: 7200 -8571.237 0.212 0.181
Chain 1: 7300 -8554.444 0.194 0.138
Chain 1: 7400 -8608.148 0.192 0.138
Chain 1: 7500 -8685.245 0.179 0.117
Chain 1: 7600 -8822.378 0.177 0.117
Chain 1: 7700 -9940.189 0.159 0.112
Chain 1: 7800 -8416.322 0.154 0.112
Chain 1: 7900 -11146.235 0.167 0.112
Chain 1: 8000 -9349.958 0.148 0.112
Chain 1: 8100 -8614.434 0.089 0.085
Chain 1: 8200 -9178.149 0.091 0.085
Chain 1: 8300 -8341.587 0.101 0.100
Chain 1: 8400 -8406.275 0.101 0.100
Chain 1: 8500 -9417.540 0.111 0.107
Chain 1: 8600 -10019.280 0.115 0.107
Chain 1: 8700 -8514.392 0.122 0.107
Chain 1: 8800 -9094.698 0.110 0.100
Chain 1: 8900 -9593.358 0.091 0.085
Chain 1: 9000 -11610.739 0.089 0.085
Chain 1: 9100 -9172.181 0.107 0.100
Chain 1: 9200 -8517.144 0.108 0.100
Chain 1: 9300 -8645.624 0.100 0.077
Chain 1: 9400 -8473.388 0.101 0.077
Chain 1: 9500 -8232.735 0.093 0.064
Chain 1: 9600 -8406.733 0.089 0.064
Chain 1: 9700 -8502.542 0.073 0.052
Chain 1: 9800 -9109.978 0.073 0.052
Chain 1: 9900 -9183.565 0.069 0.029
Chain 1: 10000 -10507.297 0.064 0.029
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001547 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.47 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58553.654 1.000 1.000
Chain 1: 200 -17897.106 1.636 2.272
Chain 1: 300 -8815.090 1.434 1.030
Chain 1: 400 -8192.184 1.095 1.030
Chain 1: 500 -8134.447 0.877 1.000
Chain 1: 600 -8434.401 0.737 1.000
Chain 1: 700 -8211.316 0.635 0.076
Chain 1: 800 -8319.196 0.558 0.076
Chain 1: 900 -8121.189 0.498 0.036
Chain 1: 1000 -7909.870 0.451 0.036
Chain 1: 1100 -7730.954 0.354 0.027
Chain 1: 1200 -7655.938 0.127 0.027
Chain 1: 1300 -7845.583 0.027 0.024
Chain 1: 1400 -7864.076 0.019 0.024
Chain 1: 1500 -7661.891 0.021 0.024
Chain 1: 1600 -7840.748 0.020 0.024
Chain 1: 1700 -7581.784 0.021 0.024
Chain 1: 1800 -7634.834 0.020 0.024
Chain 1: 1900 -7671.958 0.018 0.023
Chain 1: 2000 -7709.634 0.016 0.023
Chain 1: 2100 -7651.654 0.014 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003458 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.58 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86363.472 1.000 1.000
Chain 1: 200 -13705.237 3.151 5.301
Chain 1: 300 -10006.708 2.224 1.000
Chain 1: 400 -11059.301 1.692 1.000
Chain 1: 500 -8795.629 1.405 0.370
Chain 1: 600 -8344.747 1.180 0.370
Chain 1: 700 -8462.851 1.013 0.257
Chain 1: 800 -8733.755 0.890 0.257
Chain 1: 900 -8686.038 0.792 0.095
Chain 1: 1000 -8724.427 0.713 0.095
Chain 1: 1100 -8553.544 0.615 0.054 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8363.493 0.087 0.031
Chain 1: 1300 -8646.499 0.054 0.031
Chain 1: 1400 -8625.005 0.044 0.023
Chain 1: 1500 -8488.913 0.020 0.020
Chain 1: 1600 -8605.863 0.016 0.016
Chain 1: 1700 -8667.528 0.016 0.016
Chain 1: 1800 -8229.008 0.018 0.016
Chain 1: 1900 -8333.546 0.018 0.016
Chain 1: 2000 -8312.824 0.018 0.016
Chain 1: 2100 -8459.032 0.018 0.016
Chain 1: 2200 -8235.533 0.018 0.016
Chain 1: 2300 -8405.896 0.017 0.016
Chain 1: 2400 -8237.424 0.019 0.017
Chain 1: 2500 -8307.108 0.018 0.017
Chain 1: 2600 -8219.143 0.018 0.017
Chain 1: 2700 -8252.682 0.018 0.017
Chain 1: 2800 -8211.197 0.013 0.013
Chain 1: 2900 -8307.342 0.013 0.012
Chain 1: 3000 -8146.003 0.014 0.017
Chain 1: 3100 -8295.362 0.015 0.018
Chain 1: 3200 -8166.298 0.013 0.016
Chain 1: 3300 -8178.396 0.012 0.012
Chain 1: 3400 -8354.024 0.012 0.012
Chain 1: 3500 -8357.780 0.011 0.012
Chain 1: 3600 -8122.318 0.013 0.016
Chain 1: 3700 -8270.705 0.014 0.018
Chain 1: 3800 -8128.114 0.015 0.018
Chain 1: 3900 -8061.826 0.015 0.018
Chain 1: 4000 -8144.635 0.014 0.018
Chain 1: 4100 -8133.909 0.012 0.016
Chain 1: 4200 -8119.182 0.011 0.010
Chain 1: 4300 -8152.472 0.011 0.010
Chain 1: 4400 -8109.799 0.010 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003558 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.58 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8431115.647 1.000 1.000
Chain 1: 200 -1588575.814 2.654 4.307
Chain 1: 300 -889350.980 2.031 1.000
Chain 1: 400 -456546.889 1.760 1.000
Chain 1: 500 -356690.568 1.464 0.948
Chain 1: 600 -231964.891 1.310 0.948
Chain 1: 700 -118826.450 1.259 0.948
Chain 1: 800 -86217.319 1.149 0.948
Chain 1: 900 -66687.355 1.054 0.786
Chain 1: 1000 -51586.178 0.978 0.786
Chain 1: 1100 -39154.717 0.909 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38347.488 0.481 0.378
Chain 1: 1300 -26383.376 0.447 0.378
Chain 1: 1400 -26112.600 0.354 0.317
Chain 1: 1500 -22720.346 0.341 0.317
Chain 1: 1600 -21944.175 0.290 0.293
Chain 1: 1700 -20826.556 0.200 0.293
Chain 1: 1800 -20773.286 0.163 0.149
Chain 1: 1900 -21099.967 0.135 0.054
Chain 1: 2000 -19614.894 0.113 0.054
Chain 1: 2100 -19853.069 0.083 0.035
Chain 1: 2200 -20079.143 0.082 0.035
Chain 1: 2300 -19696.536 0.039 0.019
Chain 1: 2400 -19468.532 0.039 0.019
Chain 1: 2500 -19270.270 0.025 0.015
Chain 1: 2600 -18900.249 0.023 0.015
Chain 1: 2700 -18857.203 0.018 0.012
Chain 1: 2800 -18573.699 0.019 0.015
Chain 1: 2900 -18855.070 0.019 0.015
Chain 1: 3000 -18841.284 0.012 0.012
Chain 1: 3100 -18926.341 0.011 0.012
Chain 1: 3200 -18616.789 0.012 0.015
Chain 1: 3300 -18821.726 0.011 0.012
Chain 1: 3400 -18296.080 0.012 0.015
Chain 1: 3500 -18908.683 0.015 0.015
Chain 1: 3600 -18214.353 0.016 0.015
Chain 1: 3700 -18601.822 0.018 0.017
Chain 1: 3800 -17559.934 0.023 0.021
Chain 1: 3900 -17555.999 0.021 0.021
Chain 1: 4000 -17673.352 0.022 0.021
Chain 1: 4100 -17586.995 0.022 0.021
Chain 1: 4200 -17402.926 0.021 0.021
Chain 1: 4300 -17541.591 0.021 0.021
Chain 1: 4400 -17498.112 0.018 0.011
Chain 1: 4500 -17400.558 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001281 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.81 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48871.449 1.000 1.000
Chain 1: 200 -15370.307 1.590 2.180
Chain 1: 300 -18907.050 1.122 1.000
Chain 1: 400 -18512.747 0.847 1.000
Chain 1: 500 -16623.035 0.700 0.187
Chain 1: 600 -15304.315 0.598 0.187
Chain 1: 700 -16182.077 0.520 0.114
Chain 1: 800 -13173.488 0.484 0.187
Chain 1: 900 -13913.544 0.436 0.114
Chain 1: 1000 -12070.569 0.408 0.153
Chain 1: 1100 -10141.108 0.327 0.153
Chain 1: 1200 -12822.482 0.130 0.153
Chain 1: 1300 -19318.612 0.145 0.153
Chain 1: 1400 -12998.503 0.191 0.190
Chain 1: 1500 -10880.780 0.199 0.195
Chain 1: 1600 -11760.160 0.198 0.195
Chain 1: 1700 -12102.983 0.195 0.195
Chain 1: 1800 -20065.885 0.212 0.195
Chain 1: 1900 -10801.385 0.293 0.209
Chain 1: 2000 -12924.260 0.294 0.209
Chain 1: 2100 -10026.463 0.304 0.289
Chain 1: 2200 -9484.393 0.289 0.289
Chain 1: 2300 -9803.615 0.258 0.195
Chain 1: 2400 -9620.829 0.211 0.164
Chain 1: 2500 -9890.624 0.195 0.075
Chain 1: 2600 -9845.624 0.188 0.057
Chain 1: 2700 -12800.422 0.208 0.164
Chain 1: 2800 -11296.124 0.182 0.133
Chain 1: 2900 -15786.836 0.124 0.133
Chain 1: 3000 -9027.392 0.183 0.133
Chain 1: 3100 -8917.094 0.155 0.057
Chain 1: 3200 -10919.527 0.168 0.133
Chain 1: 3300 -9387.933 0.181 0.163
Chain 1: 3400 -9248.146 0.180 0.163
Chain 1: 3500 -9860.323 0.184 0.163
Chain 1: 3600 -10283.548 0.187 0.163
Chain 1: 3700 -9456.070 0.173 0.133
Chain 1: 3800 -13762.487 0.191 0.163
Chain 1: 3900 -8806.398 0.219 0.163
Chain 1: 4000 -8770.490 0.144 0.088
Chain 1: 4100 -9251.530 0.148 0.088
Chain 1: 4200 -9682.440 0.135 0.062
Chain 1: 4300 -9832.065 0.120 0.052
Chain 1: 4400 -11053.134 0.129 0.062
Chain 1: 4500 -15135.208 0.150 0.088
Chain 1: 4600 -13345.184 0.159 0.110
Chain 1: 4700 -8839.932 0.202 0.134
Chain 1: 4800 -12846.739 0.201 0.134
Chain 1: 4900 -9548.258 0.180 0.134
Chain 1: 5000 -9575.463 0.180 0.134
Chain 1: 5100 -12031.560 0.195 0.204
Chain 1: 5200 -13026.641 0.198 0.204
Chain 1: 5300 -8659.042 0.247 0.270
Chain 1: 5400 -8875.057 0.238 0.270
Chain 1: 5500 -9146.690 0.214 0.204
Chain 1: 5600 -13218.052 0.232 0.308
Chain 1: 5700 -12868.592 0.183 0.204
Chain 1: 5800 -9013.330 0.195 0.204
Chain 1: 5900 -8410.336 0.168 0.076
Chain 1: 6000 -12458.047 0.200 0.204
Chain 1: 6100 -8832.958 0.220 0.308
Chain 1: 6200 -8444.614 0.217 0.308
Chain 1: 6300 -9108.475 0.174 0.073
Chain 1: 6400 -13283.214 0.203 0.308
Chain 1: 6500 -10701.908 0.224 0.308
Chain 1: 6600 -8951.628 0.213 0.241
Chain 1: 6700 -8805.522 0.212 0.241
Chain 1: 6800 -10972.968 0.189 0.198
Chain 1: 6900 -9324.115 0.200 0.198
Chain 1: 7000 -12277.701 0.191 0.198
Chain 1: 7100 -8303.813 0.198 0.198
Chain 1: 7200 -8995.045 0.201 0.198
Chain 1: 7300 -8430.817 0.200 0.198
Chain 1: 7400 -8822.259 0.173 0.196
Chain 1: 7500 -8654.178 0.151 0.177
Chain 1: 7600 -8442.110 0.134 0.077
Chain 1: 7700 -8285.513 0.135 0.077
Chain 1: 7800 -8731.198 0.120 0.067
Chain 1: 7900 -9346.855 0.109 0.066
Chain 1: 8000 -8312.196 0.097 0.066
Chain 1: 8100 -8780.528 0.055 0.053
Chain 1: 8200 -8674.527 0.048 0.051
Chain 1: 8300 -10520.591 0.059 0.051
Chain 1: 8400 -10360.838 0.056 0.051
Chain 1: 8500 -8435.989 0.077 0.053
Chain 1: 8600 -12888.634 0.109 0.066
Chain 1: 8700 -8341.752 0.162 0.124
Chain 1: 8800 -8691.931 0.161 0.124
Chain 1: 8900 -9736.585 0.165 0.124
Chain 1: 9000 -8385.100 0.168 0.161
Chain 1: 9100 -8925.053 0.169 0.161
Chain 1: 9200 -8372.952 0.174 0.161
Chain 1: 9300 -11107.336 0.182 0.161
Chain 1: 9400 -11485.461 0.183 0.161
Chain 1: 9500 -8449.332 0.196 0.161
Chain 1: 9600 -9976.612 0.177 0.153
Chain 1: 9700 -8369.846 0.142 0.153
Chain 1: 9800 -10522.908 0.158 0.161
Chain 1: 9900 -11086.179 0.153 0.161
Chain 1: 10000 -8271.426 0.171 0.192
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001525 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.25 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57001.043 1.000 1.000
Chain 1: 200 -17507.846 1.628 2.256
Chain 1: 300 -8755.176 1.418 1.000
Chain 1: 400 -8409.954 1.074 1.000
Chain 1: 500 -8141.657 0.866 1.000
Chain 1: 600 -8907.246 0.736 1.000
Chain 1: 700 -8041.596 0.646 0.108
Chain 1: 800 -7871.526 0.568 0.108
Chain 1: 900 -8165.822 0.509 0.086
Chain 1: 1000 -7874.610 0.462 0.086
Chain 1: 1100 -7801.045 0.363 0.041
Chain 1: 1200 -7661.047 0.139 0.037
Chain 1: 1300 -7689.925 0.039 0.036
Chain 1: 1400 -7847.418 0.037 0.033
Chain 1: 1500 -7646.490 0.037 0.026
Chain 1: 1600 -7791.503 0.030 0.022
Chain 1: 1700 -7547.412 0.022 0.022
Chain 1: 1800 -7597.034 0.021 0.020
Chain 1: 1900 -7633.598 0.018 0.019
Chain 1: 2000 -7666.290 0.014 0.018
Chain 1: 2100 -7697.215 0.014 0.018
Chain 1: 2200 -7740.234 0.013 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003311 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.11 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86668.781 1.000 1.000
Chain 1: 200 -13515.678 3.206 5.412
Chain 1: 300 -9902.667 2.259 1.000
Chain 1: 400 -10717.475 1.713 1.000
Chain 1: 500 -8860.080 1.413 0.365
Chain 1: 600 -8411.733 1.186 0.365
Chain 1: 700 -8662.337 1.021 0.210
Chain 1: 800 -9401.273 0.903 0.210
Chain 1: 900 -8774.067 0.811 0.079
Chain 1: 1000 -8492.335 0.733 0.079
Chain 1: 1100 -8697.708 0.635 0.076 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8270.279 0.099 0.071
Chain 1: 1300 -8584.368 0.066 0.053
Chain 1: 1400 -8600.403 0.059 0.052
Chain 1: 1500 -8482.993 0.039 0.037
Chain 1: 1600 -8590.473 0.035 0.033
Chain 1: 1700 -8675.850 0.033 0.033
Chain 1: 1800 -8268.486 0.030 0.033
Chain 1: 1900 -8365.055 0.024 0.024
Chain 1: 2000 -8337.399 0.021 0.014
Chain 1: 2100 -8458.266 0.020 0.014
Chain 1: 2200 -8281.751 0.017 0.014
Chain 1: 2300 -8403.408 0.015 0.014
Chain 1: 2400 -8414.098 0.015 0.014
Chain 1: 2500 -8376.247 0.014 0.013
Chain 1: 2600 -8375.733 0.013 0.012
Chain 1: 2700 -8290.733 0.013 0.012
Chain 1: 2800 -8255.531 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003128 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.28 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8409672.474 1.000 1.000
Chain 1: 200 -1587468.522 2.649 4.298
Chain 1: 300 -891531.976 2.026 1.000
Chain 1: 400 -458007.601 1.756 1.000
Chain 1: 500 -358095.123 1.461 0.947
Chain 1: 600 -233013.429 1.307 0.947
Chain 1: 700 -119234.519 1.256 0.947
Chain 1: 800 -86450.735 1.147 0.947
Chain 1: 900 -66792.441 1.052 0.781
Chain 1: 1000 -51587.713 0.976 0.781
Chain 1: 1100 -39065.844 0.908 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38239.902 0.481 0.379
Chain 1: 1300 -26200.272 0.449 0.379
Chain 1: 1400 -25918.770 0.355 0.321
Chain 1: 1500 -22507.522 0.342 0.321
Chain 1: 1600 -21724.268 0.292 0.295
Chain 1: 1700 -20598.707 0.202 0.294
Chain 1: 1800 -20542.849 0.165 0.152
Chain 1: 1900 -20868.870 0.137 0.055
Chain 1: 2000 -19380.690 0.115 0.055
Chain 1: 2100 -19618.979 0.084 0.036
Chain 1: 2200 -19845.346 0.083 0.036
Chain 1: 2300 -19462.671 0.039 0.020
Chain 1: 2400 -19234.811 0.039 0.020
Chain 1: 2500 -19036.843 0.025 0.016
Chain 1: 2600 -18667.194 0.024 0.016
Chain 1: 2700 -18624.192 0.018 0.012
Chain 1: 2800 -18341.137 0.020 0.015
Chain 1: 2900 -18622.321 0.019 0.015
Chain 1: 3000 -18608.497 0.012 0.012
Chain 1: 3100 -18693.459 0.011 0.012
Chain 1: 3200 -18384.251 0.012 0.015
Chain 1: 3300 -18588.894 0.011 0.012
Chain 1: 3400 -18064.001 0.013 0.015
Chain 1: 3500 -18675.603 0.015 0.015
Chain 1: 3600 -17982.661 0.017 0.015
Chain 1: 3700 -18369.177 0.019 0.017
Chain 1: 3800 -17329.465 0.023 0.021
Chain 1: 3900 -17325.622 0.021 0.021
Chain 1: 4000 -17442.926 0.022 0.021
Chain 1: 4100 -17356.720 0.022 0.021
Chain 1: 4200 -17173.072 0.021 0.021
Chain 1: 4300 -17311.386 0.021 0.021
Chain 1: 4400 -17268.320 0.019 0.011
Chain 1: 4500 -17170.861 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001365 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.65 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12951.228 1.000 1.000
Chain 1: 200 -9540.310 0.679 1.000
Chain 1: 300 -8098.648 0.512 0.358
Chain 1: 400 -8270.125 0.389 0.358
Chain 1: 500 -8144.199 0.314 0.178
Chain 1: 600 -7900.078 0.267 0.178
Chain 1: 700 -8011.247 0.231 0.031
Chain 1: 800 -7809.343 0.205 0.031
Chain 1: 900 -7823.647 0.183 0.026
Chain 1: 1000 -7878.282 0.165 0.026
Chain 1: 1100 -7990.988 0.067 0.021
Chain 1: 1200 -7864.113 0.032 0.016
Chain 1: 1300 -7815.833 0.015 0.015
Chain 1: 1400 -7841.613 0.013 0.014
Chain 1: 1500 -7929.366 0.013 0.014
Chain 1: 1600 -7901.352 0.010 0.011
Chain 1: 1700 -7822.868 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00144 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.4 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56755.959 1.000 1.000
Chain 1: 200 -17273.972 1.643 2.286
Chain 1: 300 -8571.771 1.434 1.015
Chain 1: 400 -7880.483 1.097 1.015
Chain 1: 500 -8304.272 0.888 1.000
Chain 1: 600 -9185.264 0.756 1.000
Chain 1: 700 -7775.394 0.674 0.181
Chain 1: 800 -8072.655 0.594 0.181
Chain 1: 900 -7952.900 0.530 0.096
Chain 1: 1000 -7718.410 0.480 0.096
Chain 1: 1100 -7792.452 0.381 0.088
Chain 1: 1200 -7550.724 0.155 0.051
Chain 1: 1300 -7724.945 0.056 0.037
Chain 1: 1400 -7798.618 0.048 0.032
Chain 1: 1500 -7610.626 0.046 0.030
Chain 1: 1600 -7570.214 0.037 0.025
Chain 1: 1700 -7500.021 0.020 0.023
Chain 1: 1800 -7568.697 0.017 0.015
Chain 1: 1900 -7622.641 0.016 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003015 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.15 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86704.372 1.000 1.000
Chain 1: 200 -13348.234 3.248 5.496
Chain 1: 300 -9764.722 2.288 1.000
Chain 1: 400 -10465.454 1.732 1.000
Chain 1: 500 -8697.265 1.427 0.367
Chain 1: 600 -8280.619 1.197 0.367
Chain 1: 700 -8324.539 1.027 0.203
Chain 1: 800 -9124.075 0.910 0.203
Chain 1: 900 -8658.117 0.814 0.088
Chain 1: 1000 -8303.957 0.737 0.088
Chain 1: 1100 -8595.682 0.641 0.067 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8318.665 0.094 0.054
Chain 1: 1300 -8498.050 0.060 0.050
Chain 1: 1400 -8500.864 0.053 0.043
Chain 1: 1500 -8362.867 0.034 0.034
Chain 1: 1600 -8471.638 0.031 0.033
Chain 1: 1700 -8557.653 0.031 0.033
Chain 1: 1800 -8162.303 0.027 0.033
Chain 1: 1900 -8262.630 0.023 0.021
Chain 1: 2000 -8233.469 0.019 0.017
Chain 1: 2100 -8354.807 0.017 0.015
Chain 1: 2200 -8133.452 0.017 0.015
Chain 1: 2300 -8291.550 0.016 0.015
Chain 1: 2400 -8304.181 0.017 0.015
Chain 1: 2500 -8275.287 0.015 0.013
Chain 1: 2600 -8277.936 0.014 0.012
Chain 1: 2700 -8184.120 0.014 0.012
Chain 1: 2800 -8154.579 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003386 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.86 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8424398.871 1.000 1.000
Chain 1: 200 -1587570.862 2.653 4.306
Chain 1: 300 -890995.456 2.029 1.000
Chain 1: 400 -457668.332 1.759 1.000
Chain 1: 500 -357523.498 1.463 0.947
Chain 1: 600 -232331.926 1.309 0.947
Chain 1: 700 -118768.931 1.259 0.947
Chain 1: 800 -86040.604 1.149 0.947
Chain 1: 900 -66429.447 1.054 0.782
Chain 1: 1000 -51266.998 0.978 0.782
Chain 1: 1100 -38791.591 0.910 0.539 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37970.118 0.482 0.380
Chain 1: 1300 -25983.014 0.450 0.380
Chain 1: 1400 -25705.816 0.356 0.322
Chain 1: 1500 -22308.032 0.343 0.322
Chain 1: 1600 -21528.583 0.293 0.296
Chain 1: 1700 -20409.576 0.203 0.295
Chain 1: 1800 -20355.213 0.165 0.152
Chain 1: 1900 -20680.992 0.137 0.055
Chain 1: 2000 -19196.627 0.115 0.055
Chain 1: 2100 -19434.738 0.085 0.036
Chain 1: 2200 -19660.343 0.083 0.036
Chain 1: 2300 -19278.379 0.039 0.020
Chain 1: 2400 -19050.668 0.039 0.020
Chain 1: 2500 -18852.434 0.025 0.016
Chain 1: 2600 -18483.219 0.024 0.016
Chain 1: 2700 -18440.394 0.018 0.012
Chain 1: 2800 -18157.307 0.020 0.016
Chain 1: 2900 -18438.301 0.020 0.015
Chain 1: 3000 -18424.579 0.012 0.012
Chain 1: 3100 -18509.496 0.011 0.012
Chain 1: 3200 -18200.476 0.012 0.015
Chain 1: 3300 -18404.969 0.011 0.012
Chain 1: 3400 -17880.332 0.013 0.015
Chain 1: 3500 -18491.438 0.015 0.016
Chain 1: 3600 -17799.094 0.017 0.016
Chain 1: 3700 -18185.119 0.019 0.017
Chain 1: 3800 -17146.275 0.023 0.021
Chain 1: 3900 -17142.412 0.022 0.021
Chain 1: 4000 -17259.762 0.022 0.021
Chain 1: 4100 -17173.568 0.022 0.021
Chain 1: 4200 -16990.135 0.022 0.021
Chain 1: 4300 -17128.333 0.021 0.021
Chain 1: 4400 -17085.419 0.019 0.011
Chain 1: 4500 -16987.959 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001508 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.08 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13015.433 1.000 1.000
Chain 1: 200 -10116.092 0.643 1.000
Chain 1: 300 -8703.474 0.483 0.287
Chain 1: 400 -8396.292 0.371 0.287
Chain 1: 500 -8267.878 0.300 0.162
Chain 1: 600 -8256.309 0.250 0.162
Chain 1: 700 -8275.206 0.215 0.037
Chain 1: 800 -8226.952 0.189 0.037
Chain 1: 900 -8176.546 0.169 0.016
Chain 1: 1000 -8256.053 0.153 0.016
Chain 1: 1100 -8303.388 0.053 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001433 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.33 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -52724.164 1.000 1.000
Chain 1: 200 -17174.804 1.535 2.070
Chain 1: 300 -9219.179 1.311 1.000
Chain 1: 400 -8643.046 1.000 1.000
Chain 1: 500 -9098.447 0.810 0.863
Chain 1: 600 -8362.405 0.690 0.863
Chain 1: 700 -7841.640 0.601 0.088
Chain 1: 800 -8601.280 0.537 0.088
Chain 1: 900 -8149.646 0.483 0.088
Chain 1: 1000 -7990.752 0.437 0.088
Chain 1: 1100 -7861.049 0.338 0.067
Chain 1: 1200 -7982.270 0.133 0.066
Chain 1: 1300 -7798.993 0.049 0.055
Chain 1: 1400 -7811.636 0.042 0.050
Chain 1: 1500 -7679.910 0.039 0.024
Chain 1: 1600 -7782.298 0.032 0.020
Chain 1: 1700 -7785.489 0.025 0.017
Chain 1: 1800 -7727.549 0.017 0.016
Chain 1: 1900 -7692.360 0.012 0.015
Chain 1: 2000 -7922.218 0.013 0.015
Chain 1: 2100 -7763.958 0.013 0.015
Chain 1: 2200 -8106.781 0.016 0.017
Chain 1: 2300 -7715.097 0.019 0.017
Chain 1: 2400 -7891.178 0.021 0.020
Chain 1: 2500 -7722.755 0.021 0.022
Chain 1: 2600 -7683.143 0.020 0.022
Chain 1: 2700 -7592.034 0.022 0.022
Chain 1: 2800 -7803.684 0.024 0.022
Chain 1: 2900 -7539.213 0.027 0.027
Chain 1: 3000 -7686.111 0.026 0.022
Chain 1: 3100 -7689.296 0.024 0.022
Chain 1: 3200 -7893.369 0.022 0.022
Chain 1: 3300 -7597.444 0.021 0.022
Chain 1: 3400 -7815.924 0.021 0.026
Chain 1: 3500 -7592.765 0.022 0.027
Chain 1: 3600 -7658.040 0.022 0.027
Chain 1: 3700 -7611.569 0.022 0.027
Chain 1: 3800 -7585.804 0.019 0.026
Chain 1: 3900 -7559.601 0.016 0.019
Chain 1: 4000 -7555.599 0.014 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003228 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.28 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86555.672 1.000 1.000
Chain 1: 200 -14235.423 3.040 5.080
Chain 1: 300 -10413.218 2.149 1.000
Chain 1: 400 -12535.058 1.654 1.000
Chain 1: 500 -8784.918 1.409 0.427
Chain 1: 600 -8770.927 1.174 0.427
Chain 1: 700 -9316.282 1.015 0.367
Chain 1: 800 -9198.542 0.890 0.367
Chain 1: 900 -9060.979 0.792 0.169
Chain 1: 1000 -9435.718 0.717 0.169
Chain 1: 1100 -9128.710 0.620 0.059 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8736.807 0.117 0.045
Chain 1: 1300 -8990.381 0.083 0.040
Chain 1: 1400 -8834.128 0.068 0.034
Chain 1: 1500 -8856.446 0.025 0.028
Chain 1: 1600 -8926.541 0.026 0.028
Chain 1: 1700 -8987.947 0.021 0.018
Chain 1: 1800 -8525.114 0.025 0.028
Chain 1: 1900 -8638.996 0.025 0.028
Chain 1: 2000 -8656.363 0.021 0.018
Chain 1: 2100 -8748.418 0.019 0.013
Chain 1: 2200 -8517.702 0.017 0.013
Chain 1: 2300 -8703.992 0.016 0.013
Chain 1: 2400 -8542.920 0.016 0.013
Chain 1: 2500 -8607.336 0.017 0.013
Chain 1: 2600 -8515.087 0.017 0.013
Chain 1: 2700 -8549.933 0.017 0.013
Chain 1: 2800 -8505.065 0.012 0.011
Chain 1: 2900 -8616.231 0.012 0.011
Chain 1: 3000 -8524.487 0.013 0.011
Chain 1: 3100 -8492.291 0.012 0.011
Chain 1: 3200 -8461.814 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003445 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.45 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8430167.934 1.000 1.000
Chain 1: 200 -1587811.308 2.655 4.309
Chain 1: 300 -892514.313 2.029 1.000
Chain 1: 400 -458881.777 1.758 1.000
Chain 1: 500 -359127.242 1.462 0.945
Chain 1: 600 -233870.345 1.308 0.945
Chain 1: 700 -120020.871 1.256 0.945
Chain 1: 800 -87243.431 1.146 0.945
Chain 1: 900 -67576.405 1.051 0.779
Chain 1: 1000 -52389.922 0.975 0.779
Chain 1: 1100 -39870.736 0.907 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39054.141 0.478 0.376
Chain 1: 1300 -26994.003 0.445 0.376
Chain 1: 1400 -26716.154 0.351 0.314
Chain 1: 1500 -23299.396 0.338 0.314
Chain 1: 1600 -22516.156 0.288 0.291
Chain 1: 1700 -21386.933 0.198 0.290
Chain 1: 1800 -21331.140 0.161 0.147
Chain 1: 1900 -21658.189 0.133 0.053
Chain 1: 2000 -20166.261 0.112 0.053
Chain 1: 2100 -20404.666 0.082 0.035
Chain 1: 2200 -20632.134 0.081 0.035
Chain 1: 2300 -20248.278 0.038 0.019
Chain 1: 2400 -20020.017 0.038 0.019
Chain 1: 2500 -19822.140 0.024 0.015
Chain 1: 2600 -19451.178 0.023 0.015
Chain 1: 2700 -19407.856 0.018 0.012
Chain 1: 2800 -19124.323 0.019 0.015
Chain 1: 2900 -19406.034 0.019 0.015
Chain 1: 3000 -19392.139 0.011 0.012
Chain 1: 3100 -19477.275 0.011 0.011
Chain 1: 3200 -19167.262 0.011 0.015
Chain 1: 3300 -19372.534 0.010 0.011
Chain 1: 3400 -18846.258 0.012 0.015
Chain 1: 3500 -19459.926 0.014 0.015
Chain 1: 3600 -18764.303 0.016 0.015
Chain 1: 3700 -19152.803 0.018 0.016
Chain 1: 3800 -18108.926 0.022 0.020
Chain 1: 3900 -18105.000 0.021 0.020
Chain 1: 4000 -18222.305 0.021 0.020
Chain 1: 4100 -18135.883 0.021 0.020
Chain 1: 4200 -17951.364 0.021 0.020
Chain 1: 4300 -18090.279 0.020 0.020
Chain 1: 4400 -18046.430 0.018 0.010
Chain 1: 4500 -17948.900 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001379 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.79 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48481.376 1.000 1.000
Chain 1: 200 -18766.727 1.292 1.583
Chain 1: 300 -29495.027 0.982 1.000
Chain 1: 400 -17512.015 0.908 1.000
Chain 1: 500 -16486.744 0.739 0.684
Chain 1: 600 -13558.504 0.652 0.684
Chain 1: 700 -11332.795 0.587 0.364
Chain 1: 800 -13431.633 0.533 0.364
Chain 1: 900 -21060.750 0.514 0.362
Chain 1: 1000 -11888.022 0.540 0.364
Chain 1: 1100 -15890.749 0.465 0.362
Chain 1: 1200 -12758.217 0.331 0.252
Chain 1: 1300 -11218.264 0.308 0.246
Chain 1: 1400 -10711.715 0.245 0.216
Chain 1: 1500 -12816.045 0.255 0.216
Chain 1: 1600 -12470.856 0.236 0.196
Chain 1: 1700 -9556.113 0.247 0.246
Chain 1: 1800 -11706.080 0.250 0.246
Chain 1: 1900 -9684.303 0.234 0.209
Chain 1: 2000 -18761.827 0.206 0.209
Chain 1: 2100 -10505.688 0.259 0.209
Chain 1: 2200 -11528.198 0.243 0.184
Chain 1: 2300 -11278.268 0.232 0.184
Chain 1: 2400 -9115.390 0.251 0.209
Chain 1: 2500 -9487.771 0.238 0.209
Chain 1: 2600 -9040.389 0.240 0.209
Chain 1: 2700 -10461.752 0.223 0.184
Chain 1: 2800 -10156.808 0.208 0.136
Chain 1: 2900 -9179.293 0.198 0.106
Chain 1: 3000 -12096.106 0.174 0.106
Chain 1: 3100 -10613.545 0.109 0.106
Chain 1: 3200 -8863.904 0.120 0.136
Chain 1: 3300 -9143.341 0.121 0.136
Chain 1: 3400 -8649.760 0.103 0.106
Chain 1: 3500 -14145.128 0.138 0.136
Chain 1: 3600 -8799.271 0.193 0.140
Chain 1: 3700 -15717.035 0.224 0.197
Chain 1: 3800 -8704.941 0.301 0.241
Chain 1: 3900 -9855.180 0.302 0.241
Chain 1: 4000 -8956.259 0.288 0.197
Chain 1: 4100 -9788.058 0.283 0.197
Chain 1: 4200 -12274.779 0.283 0.203
Chain 1: 4300 -8433.681 0.326 0.388
Chain 1: 4400 -12443.969 0.352 0.388
Chain 1: 4500 -8318.212 0.363 0.440
Chain 1: 4600 -15166.087 0.348 0.440
Chain 1: 4700 -14464.169 0.308 0.322
Chain 1: 4800 -8519.808 0.298 0.322
Chain 1: 4900 -14137.896 0.326 0.397
Chain 1: 5000 -12641.622 0.327 0.397
Chain 1: 5100 -8137.867 0.374 0.452
Chain 1: 5200 -13841.714 0.395 0.452
Chain 1: 5300 -12328.185 0.362 0.412
Chain 1: 5400 -14288.012 0.343 0.412
Chain 1: 5500 -8277.148 0.367 0.412
Chain 1: 5600 -9219.875 0.332 0.397
Chain 1: 5700 -12619.795 0.354 0.397
Chain 1: 5800 -9958.943 0.311 0.269
Chain 1: 5900 -14464.122 0.302 0.269
Chain 1: 6000 -8827.444 0.354 0.311
Chain 1: 6100 -8166.562 0.307 0.269
Chain 1: 6200 -8288.430 0.267 0.267
Chain 1: 6300 -13127.421 0.292 0.269
Chain 1: 6400 -11410.554 0.293 0.269
Chain 1: 6500 -10592.074 0.228 0.267
Chain 1: 6600 -9116.260 0.234 0.267
Chain 1: 6700 -11164.021 0.225 0.183
Chain 1: 6800 -9688.467 0.214 0.162
Chain 1: 6900 -12797.548 0.207 0.162
Chain 1: 7000 -11125.135 0.158 0.152
Chain 1: 7100 -8834.382 0.176 0.162
Chain 1: 7200 -8757.597 0.176 0.162
Chain 1: 7300 -9564.490 0.147 0.152
Chain 1: 7400 -11437.327 0.148 0.162
Chain 1: 7500 -8739.677 0.172 0.164
Chain 1: 7600 -8364.860 0.160 0.164
Chain 1: 7700 -8551.652 0.144 0.152
Chain 1: 7800 -11403.935 0.153 0.164
Chain 1: 7900 -8121.393 0.170 0.164
Chain 1: 8000 -9551.240 0.170 0.164
Chain 1: 8100 -7941.609 0.164 0.164
Chain 1: 8200 -9381.773 0.178 0.164
Chain 1: 8300 -9862.074 0.175 0.164
Chain 1: 8400 -8845.473 0.170 0.154
Chain 1: 8500 -7920.171 0.151 0.150
Chain 1: 8600 -9678.732 0.164 0.154
Chain 1: 8700 -9810.495 0.164 0.154
Chain 1: 8800 -7803.198 0.164 0.154
Chain 1: 8900 -9032.891 0.137 0.150
Chain 1: 9000 -9158.555 0.124 0.136
Chain 1: 9100 -7961.222 0.119 0.136
Chain 1: 9200 -8654.801 0.111 0.117
Chain 1: 9300 -7996.755 0.115 0.117
Chain 1: 9400 -11729.799 0.135 0.136
Chain 1: 9500 -11776.953 0.124 0.136
Chain 1: 9600 -8069.510 0.152 0.136
Chain 1: 9700 -8845.060 0.159 0.136
Chain 1: 9800 -9404.507 0.139 0.088
Chain 1: 9900 -10775.679 0.138 0.088
Chain 1: 10000 -10205.922 0.142 0.088
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003298 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.98 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -62895.085 1.000 1.000
Chain 1: 200 -17690.869 1.778 2.555
Chain 1: 300 -8594.630 1.538 1.058
Chain 1: 400 -8142.271 1.167 1.058
Chain 1: 500 -8527.791 0.943 1.000
Chain 1: 600 -8553.229 0.786 1.000
Chain 1: 700 -8290.474 0.678 0.056
Chain 1: 800 -7929.284 0.599 0.056
Chain 1: 900 -7943.389 0.533 0.046
Chain 1: 1000 -7684.644 0.483 0.046
Chain 1: 1100 -7706.701 0.383 0.045
Chain 1: 1200 -7635.604 0.129 0.034
Chain 1: 1300 -7753.026 0.024 0.032
Chain 1: 1400 -7802.760 0.019 0.015
Chain 1: 1500 -7659.294 0.017 0.015
Chain 1: 1600 -7562.717 0.018 0.015
Chain 1: 1700 -7536.929 0.015 0.013
Chain 1: 1800 -7595.706 0.011 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003335 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.35 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85881.781 1.000 1.000
Chain 1: 200 -13016.423 3.299 5.598
Chain 1: 300 -9533.150 2.321 1.000
Chain 1: 400 -10266.590 1.759 1.000
Chain 1: 500 -8387.898 1.452 0.365
Chain 1: 600 -8660.776 1.215 0.365
Chain 1: 700 -8184.833 1.050 0.224
Chain 1: 800 -8624.160 0.925 0.224
Chain 1: 900 -8449.338 0.824 0.071
Chain 1: 1000 -8187.952 0.745 0.071
Chain 1: 1100 -8294.027 0.646 0.058 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8128.304 0.089 0.051
Chain 1: 1300 -8181.899 0.053 0.032
Chain 1: 1400 -8248.957 0.047 0.032
Chain 1: 1500 -8204.480 0.025 0.021
Chain 1: 1600 -8208.514 0.022 0.020
Chain 1: 1700 -8157.626 0.016 0.013
Chain 1: 1800 -8034.869 0.013 0.013
Chain 1: 1900 -8145.682 0.012 0.013
Chain 1: 2000 -8110.321 0.009 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003584 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.84 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8402164.048 1.000 1.000
Chain 1: 200 -1586306.209 2.648 4.297
Chain 1: 300 -890783.019 2.026 1.000
Chain 1: 400 -457548.681 1.756 1.000
Chain 1: 500 -357655.826 1.461 0.947
Chain 1: 600 -232508.251 1.307 0.947
Chain 1: 700 -118663.744 1.257 0.947
Chain 1: 800 -85898.779 1.148 0.947
Chain 1: 900 -66227.734 1.053 0.781
Chain 1: 1000 -51013.323 0.978 0.781
Chain 1: 1100 -38496.059 0.910 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37662.927 0.483 0.381
Chain 1: 1300 -25639.267 0.452 0.381
Chain 1: 1400 -25356.211 0.358 0.325
Chain 1: 1500 -21950.267 0.346 0.325
Chain 1: 1600 -21167.655 0.296 0.298
Chain 1: 1700 -20044.431 0.205 0.297
Chain 1: 1800 -19988.757 0.167 0.155
Chain 1: 1900 -20313.934 0.139 0.056
Chain 1: 2000 -18828.644 0.117 0.056
Chain 1: 2100 -19066.718 0.086 0.037
Chain 1: 2200 -19292.398 0.085 0.037
Chain 1: 2300 -18910.507 0.040 0.020
Chain 1: 2400 -18682.939 0.040 0.020
Chain 1: 2500 -18485.089 0.026 0.016
Chain 1: 2600 -18116.270 0.024 0.016
Chain 1: 2700 -18073.439 0.019 0.012
Chain 1: 2800 -17790.830 0.020 0.016
Chain 1: 2900 -18071.566 0.020 0.016
Chain 1: 3000 -18057.771 0.012 0.012
Chain 1: 3100 -18142.674 0.011 0.012
Chain 1: 3200 -17833.995 0.012 0.016
Chain 1: 3300 -18038.180 0.011 0.012
Chain 1: 3400 -17514.361 0.013 0.016
Chain 1: 3500 -18124.433 0.015 0.016
Chain 1: 3600 -17433.347 0.017 0.016
Chain 1: 3700 -17818.524 0.019 0.017
Chain 1: 3800 -16781.849 0.024 0.022
Chain 1: 3900 -16778.067 0.022 0.022
Chain 1: 4000 -16895.344 0.023 0.022
Chain 1: 4100 -16809.363 0.023 0.022
Chain 1: 4200 -16626.317 0.022 0.022
Chain 1: 4300 -16764.180 0.022 0.022
Chain 1: 4400 -16721.632 0.019 0.011
Chain 1: 4500 -16624.277 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001381 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.81 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13019.383 1.000 1.000
Chain 1: 200 -9826.280 0.662 1.000
Chain 1: 300 -8529.427 0.492 0.325
Chain 1: 400 -8694.383 0.374 0.325
Chain 1: 500 -8642.282 0.300 0.152
Chain 1: 600 -8448.921 0.254 0.152
Chain 1: 700 -8380.451 0.219 0.023
Chain 1: 800 -8386.526 0.192 0.023
Chain 1: 900 -8322.128 0.171 0.019
Chain 1: 1000 -8504.286 0.156 0.021
Chain 1: 1100 -8489.534 0.056 0.019
Chain 1: 1200 -8394.046 0.025 0.011
Chain 1: 1300 -8297.961 0.011 0.011
Chain 1: 1400 -8313.455 0.009 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001402 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.02 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58883.116 1.000 1.000
Chain 1: 200 -18333.073 1.606 2.212
Chain 1: 300 -8987.335 1.417 1.040
Chain 1: 400 -8042.281 1.092 1.040
Chain 1: 500 -9300.009 0.901 1.000
Chain 1: 600 -8818.649 0.760 1.000
Chain 1: 700 -8937.429 0.653 0.135
Chain 1: 800 -8464.648 0.579 0.135
Chain 1: 900 -8122.728 0.519 0.118
Chain 1: 1000 -7882.925 0.470 0.118
Chain 1: 1100 -7787.657 0.371 0.056
Chain 1: 1200 -7607.488 0.152 0.055
Chain 1: 1300 -7732.981 0.050 0.042
Chain 1: 1400 -7764.207 0.039 0.030
Chain 1: 1500 -7536.883 0.028 0.030
Chain 1: 1600 -7749.770 0.026 0.027
Chain 1: 1700 -7606.208 0.026 0.027
Chain 1: 1800 -7547.740 0.021 0.024
Chain 1: 1900 -7592.438 0.018 0.019
Chain 1: 2000 -7635.093 0.015 0.016
Chain 1: 2100 -7572.334 0.015 0.016
Chain 1: 2200 -7853.266 0.016 0.016
Chain 1: 2300 -7608.654 0.018 0.019
Chain 1: 2400 -7612.711 0.017 0.019
Chain 1: 2500 -7589.442 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004033 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 40.33 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86701.933 1.000 1.000
Chain 1: 200 -14128.891 3.068 5.136
Chain 1: 300 -10425.181 2.164 1.000
Chain 1: 400 -11895.557 1.654 1.000
Chain 1: 500 -9381.478 1.377 0.355
Chain 1: 600 -8774.138 1.159 0.355
Chain 1: 700 -9432.073 1.003 0.268
Chain 1: 800 -9811.611 0.883 0.268
Chain 1: 900 -9102.458 0.793 0.124
Chain 1: 1000 -8752.323 0.718 0.124
Chain 1: 1100 -9165.476 0.622 0.078 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8774.374 0.113 0.070
Chain 1: 1300 -9128.157 0.082 0.069
Chain 1: 1400 -8907.880 0.072 0.045
Chain 1: 1500 -8935.273 0.045 0.045
Chain 1: 1600 -9043.373 0.039 0.040
Chain 1: 1700 -9111.013 0.033 0.039
Chain 1: 1800 -8671.656 0.034 0.040
Chain 1: 1900 -8776.829 0.028 0.039
Chain 1: 2000 -8755.781 0.024 0.025
Chain 1: 2100 -8896.011 0.021 0.016
Chain 1: 2200 -8683.934 0.019 0.016
Chain 1: 2300 -8843.306 0.017 0.016
Chain 1: 2400 -8680.913 0.016 0.016
Chain 1: 2500 -8752.392 0.017 0.016
Chain 1: 2600 -8664.244 0.017 0.016
Chain 1: 2700 -8697.921 0.016 0.016
Chain 1: 2800 -8657.583 0.012 0.012
Chain 1: 2900 -8751.576 0.012 0.011
Chain 1: 3000 -8586.790 0.013 0.016
Chain 1: 3100 -8740.349 0.014 0.018
Chain 1: 3200 -8612.056 0.013 0.015
Chain 1: 3300 -8621.254 0.011 0.011
Chain 1: 3400 -8784.368 0.011 0.011
Chain 1: 3500 -8796.344 0.010 0.011
Chain 1: 3600 -8568.406 0.012 0.015
Chain 1: 3700 -8715.335 0.013 0.017
Chain 1: 3800 -8574.643 0.014 0.017
Chain 1: 3900 -8508.868 0.014 0.017
Chain 1: 4000 -8586.609 0.013 0.016
Chain 1: 4100 -8580.233 0.011 0.015
Chain 1: 4200 -8564.726 0.010 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003304 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.04 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8412306.651 1.000 1.000
Chain 1: 200 -1585847.309 2.652 4.305
Chain 1: 300 -891441.554 2.028 1.000
Chain 1: 400 -458453.657 1.757 1.000
Chain 1: 500 -358596.192 1.461 0.944
Chain 1: 600 -233515.326 1.307 0.944
Chain 1: 700 -119795.333 1.256 0.944
Chain 1: 800 -87048.479 1.146 0.944
Chain 1: 900 -67402.621 1.051 0.779
Chain 1: 1000 -52221.555 0.975 0.779
Chain 1: 1100 -39715.070 0.906 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38896.618 0.478 0.376
Chain 1: 1300 -26857.332 0.445 0.376
Chain 1: 1400 -26579.393 0.352 0.315
Chain 1: 1500 -23167.441 0.339 0.315
Chain 1: 1600 -22385.353 0.288 0.291
Chain 1: 1700 -21258.682 0.199 0.291
Chain 1: 1800 -21203.196 0.161 0.147
Chain 1: 1900 -21529.771 0.134 0.053
Chain 1: 2000 -20040.114 0.112 0.053
Chain 1: 2100 -20278.519 0.082 0.035
Chain 1: 2200 -20505.341 0.081 0.035
Chain 1: 2300 -20122.102 0.038 0.019
Chain 1: 2400 -19894.018 0.038 0.019
Chain 1: 2500 -19696.106 0.024 0.015
Chain 1: 2600 -19325.817 0.023 0.015
Chain 1: 2700 -19282.658 0.018 0.012
Chain 1: 2800 -18999.343 0.019 0.015
Chain 1: 2900 -19280.761 0.019 0.015
Chain 1: 3000 -19266.942 0.012 0.012
Chain 1: 3100 -19352.007 0.011 0.011
Chain 1: 3200 -19042.397 0.011 0.015
Chain 1: 3300 -19247.354 0.010 0.011
Chain 1: 3400 -18721.786 0.012 0.015
Chain 1: 3500 -19334.430 0.014 0.015
Chain 1: 3600 -18640.080 0.016 0.015
Chain 1: 3700 -19027.618 0.018 0.016
Chain 1: 3800 -17985.806 0.022 0.020
Chain 1: 3900 -17981.915 0.021 0.020
Chain 1: 4000 -18099.216 0.021 0.020
Chain 1: 4100 -18012.905 0.021 0.020
Chain 1: 4200 -17828.822 0.021 0.020
Chain 1: 4300 -17967.457 0.020 0.020
Chain 1: 4400 -17923.990 0.018 0.010
Chain 1: 4500 -17826.471 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001214 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.14 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12382.293 1.000 1.000
Chain 1: 200 -9284.616 0.667 1.000
Chain 1: 300 -7979.054 0.499 0.334
Chain 1: 400 -8248.339 0.382 0.334
Chain 1: 500 -8040.408 0.311 0.164
Chain 1: 600 -7964.194 0.261 0.164
Chain 1: 700 -7866.646 0.225 0.033
Chain 1: 800 -7871.563 0.197 0.033
Chain 1: 900 -7798.520 0.176 0.026
Chain 1: 1000 -7984.766 0.161 0.026
Chain 1: 1100 -8010.237 0.061 0.023
Chain 1: 1200 -7881.407 0.030 0.016
Chain 1: 1300 -7853.153 0.014 0.012
Chain 1: 1400 -7861.372 0.011 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001568 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.68 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61918.239 1.000 1.000
Chain 1: 200 -17910.492 1.729 2.457
Chain 1: 300 -8823.089 1.496 1.030
Chain 1: 400 -9406.741 1.137 1.030
Chain 1: 500 -8302.944 0.936 1.000
Chain 1: 600 -7866.090 0.790 1.000
Chain 1: 700 -7839.401 0.677 0.133
Chain 1: 800 -8184.103 0.598 0.133
Chain 1: 900 -7917.249 0.535 0.062
Chain 1: 1000 -7988.648 0.483 0.062
Chain 1: 1100 -7789.100 0.385 0.056
Chain 1: 1200 -7675.731 0.141 0.042
Chain 1: 1300 -7744.878 0.039 0.034
Chain 1: 1400 -7640.199 0.034 0.026
Chain 1: 1500 -7535.717 0.022 0.015
Chain 1: 1600 -7463.939 0.017 0.014
Chain 1: 1700 -7551.519 0.018 0.014
Chain 1: 1800 -7598.868 0.015 0.014
Chain 1: 1900 -7585.153 0.012 0.012
Chain 1: 2000 -7536.741 0.011 0.012
Chain 1: 2100 -7576.574 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003001 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.01 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86798.934 1.000 1.000
Chain 1: 200 -13533.220 3.207 5.414
Chain 1: 300 -9881.806 2.261 1.000
Chain 1: 400 -10810.230 1.717 1.000
Chain 1: 500 -8651.888 1.424 0.370
Chain 1: 600 -8285.301 1.194 0.370
Chain 1: 700 -8381.494 1.025 0.249
Chain 1: 800 -8832.736 0.903 0.249
Chain 1: 900 -8622.480 0.806 0.086
Chain 1: 1000 -8437.522 0.727 0.086
Chain 1: 1100 -8712.590 0.630 0.051 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8317.628 0.094 0.047
Chain 1: 1300 -8556.069 0.060 0.044
Chain 1: 1400 -8582.213 0.051 0.032
Chain 1: 1500 -8426.093 0.028 0.028
Chain 1: 1600 -8540.424 0.025 0.024
Chain 1: 1700 -8616.140 0.025 0.024
Chain 1: 1800 -8192.923 0.025 0.024
Chain 1: 1900 -8294.001 0.024 0.022
Chain 1: 2000 -8268.417 0.022 0.019
Chain 1: 2100 -8393.994 0.020 0.015
Chain 1: 2200 -8197.029 0.018 0.015
Chain 1: 2300 -8288.788 0.016 0.013
Chain 1: 2400 -8357.565 0.017 0.013
Chain 1: 2500 -8303.783 0.015 0.012
Chain 1: 2600 -8305.151 0.014 0.011
Chain 1: 2700 -8221.891 0.014 0.011
Chain 1: 2800 -8181.752 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00371 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 37.1 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8402956.830 1.000 1.000
Chain 1: 200 -1584566.512 2.652 4.303
Chain 1: 300 -890013.444 2.028 1.000
Chain 1: 400 -456486.548 1.758 1.000
Chain 1: 500 -356913.715 1.462 0.950
Chain 1: 600 -231905.273 1.309 0.950
Chain 1: 700 -118730.601 1.258 0.950
Chain 1: 800 -86064.845 1.148 0.950
Chain 1: 900 -66521.947 1.053 0.780
Chain 1: 1000 -51414.008 0.977 0.780
Chain 1: 1100 -38971.387 0.909 0.539 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38163.763 0.481 0.380
Chain 1: 1300 -26193.083 0.449 0.380
Chain 1: 1400 -25920.935 0.355 0.319
Chain 1: 1500 -22525.846 0.342 0.319
Chain 1: 1600 -21748.548 0.291 0.294
Chain 1: 1700 -20630.289 0.202 0.294
Chain 1: 1800 -20576.767 0.164 0.151
Chain 1: 1900 -20903.101 0.136 0.054
Chain 1: 2000 -19418.240 0.114 0.054
Chain 1: 2100 -19656.512 0.084 0.036
Chain 1: 2200 -19882.277 0.083 0.036
Chain 1: 2300 -19500.036 0.039 0.020
Chain 1: 2400 -19272.137 0.039 0.020
Chain 1: 2500 -19073.824 0.025 0.016
Chain 1: 2600 -18704.213 0.023 0.016
Chain 1: 2700 -18661.360 0.018 0.012
Chain 1: 2800 -18377.951 0.019 0.015
Chain 1: 2900 -18659.188 0.019 0.015
Chain 1: 3000 -18645.448 0.012 0.012
Chain 1: 3100 -18730.420 0.011 0.012
Chain 1: 3200 -18421.134 0.012 0.015
Chain 1: 3300 -18625.904 0.011 0.012
Chain 1: 3400 -18100.683 0.013 0.015
Chain 1: 3500 -18712.657 0.015 0.015
Chain 1: 3600 -18019.217 0.017 0.015
Chain 1: 3700 -18405.986 0.018 0.017
Chain 1: 3800 -17365.489 0.023 0.021
Chain 1: 3900 -17361.606 0.021 0.021
Chain 1: 4000 -17478.933 0.022 0.021
Chain 1: 4100 -17392.609 0.022 0.021
Chain 1: 4200 -17208.912 0.021 0.021
Chain 1: 4300 -17347.339 0.021 0.021
Chain 1: 4400 -17304.120 0.019 0.011
Chain 1: 4500 -17206.611 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001343 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.43 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12120.377 1.000 1.000
Chain 1: 200 -9054.156 0.669 1.000
Chain 1: 300 -7867.590 0.496 0.339
Chain 1: 400 -8049.118 0.378 0.339
Chain 1: 500 -7951.404 0.305 0.151
Chain 1: 600 -7813.944 0.257 0.151
Chain 1: 700 -7736.916 0.222 0.023
Chain 1: 800 -7746.797 0.194 0.023
Chain 1: 900 -7655.047 0.174 0.018
Chain 1: 1000 -7764.017 0.158 0.018
Chain 1: 1100 -7777.090 0.058 0.014
Chain 1: 1200 -7745.935 0.025 0.012
Chain 1: 1300 -7754.993 0.010 0.012 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001471 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.71 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56627.847 1.000 1.000
Chain 1: 200 -17088.240 1.657 2.314
Chain 1: 300 -8569.976 1.436 1.000
Chain 1: 400 -7825.526 1.101 1.000
Chain 1: 500 -8322.842 0.893 0.994
Chain 1: 600 -8610.539 0.749 0.994
Chain 1: 700 -8228.179 0.649 0.095
Chain 1: 800 -8219.446 0.568 0.095
Chain 1: 900 -7729.083 0.512 0.063
Chain 1: 1000 -7724.756 0.461 0.063
Chain 1: 1100 -7611.294 0.362 0.060
Chain 1: 1200 -7552.061 0.132 0.046
Chain 1: 1300 -7600.463 0.033 0.033
Chain 1: 1400 -7753.650 0.025 0.020
Chain 1: 1500 -7547.348 0.022 0.020
Chain 1: 1600 -7509.339 0.019 0.015
Chain 1: 1700 -7478.915 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003503 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.03 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86207.637 1.000 1.000
Chain 1: 200 -13198.409 3.266 5.532
Chain 1: 300 -9632.304 2.301 1.000
Chain 1: 400 -10380.201 1.743 1.000
Chain 1: 500 -8530.520 1.438 0.370
Chain 1: 600 -8186.592 1.205 0.370
Chain 1: 700 -8344.261 1.036 0.217
Chain 1: 800 -8933.691 0.915 0.217
Chain 1: 900 -8509.256 0.819 0.072
Chain 1: 1000 -8198.625 0.741 0.072
Chain 1: 1100 -8532.965 0.644 0.066 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8059.651 0.097 0.059
Chain 1: 1300 -8234.525 0.062 0.050
Chain 1: 1400 -8364.403 0.057 0.042
Chain 1: 1500 -8252.529 0.036 0.039
Chain 1: 1600 -8361.683 0.033 0.038
Chain 1: 1700 -8442.387 0.032 0.038
Chain 1: 1800 -8049.985 0.031 0.038
Chain 1: 1900 -8152.943 0.027 0.021
Chain 1: 2000 -8122.850 0.024 0.016
Chain 1: 2100 -8249.892 0.021 0.015
Chain 1: 2200 -8036.060 0.018 0.015
Chain 1: 2300 -8181.478 0.018 0.015
Chain 1: 2400 -8196.788 0.016 0.014
Chain 1: 2500 -8163.273 0.015 0.013
Chain 1: 2600 -8165.019 0.014 0.013
Chain 1: 2700 -8072.036 0.014 0.013
Chain 1: 2800 -8045.435 0.010 0.012 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003397 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.97 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8393288.755 1.000 1.000
Chain 1: 200 -1581194.945 2.654 4.308
Chain 1: 300 -889586.382 2.029 1.000
Chain 1: 400 -456617.357 1.758 1.000
Chain 1: 500 -357137.758 1.462 0.948
Chain 1: 600 -232168.245 1.308 0.948
Chain 1: 700 -118688.230 1.258 0.948
Chain 1: 800 -85956.222 1.148 0.948
Chain 1: 900 -66345.570 1.054 0.777
Chain 1: 1000 -51174.376 0.978 0.777
Chain 1: 1100 -38682.390 0.910 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37861.235 0.482 0.381
Chain 1: 1300 -25851.725 0.450 0.381
Chain 1: 1400 -25572.123 0.357 0.323
Chain 1: 1500 -22168.068 0.344 0.323
Chain 1: 1600 -21386.879 0.294 0.296
Chain 1: 1700 -20264.868 0.204 0.296
Chain 1: 1800 -20209.925 0.166 0.154
Chain 1: 1900 -20535.699 0.138 0.055
Chain 1: 2000 -19049.937 0.116 0.055
Chain 1: 2100 -19288.037 0.085 0.037
Chain 1: 2200 -19513.887 0.084 0.037
Chain 1: 2300 -19131.755 0.040 0.020
Chain 1: 2400 -18904.053 0.040 0.020
Chain 1: 2500 -18705.926 0.025 0.016
Chain 1: 2600 -18336.634 0.024 0.016
Chain 1: 2700 -18293.850 0.019 0.012
Chain 1: 2800 -18010.840 0.020 0.016
Chain 1: 2900 -18291.835 0.020 0.015
Chain 1: 3000 -18278.069 0.012 0.012
Chain 1: 3100 -18362.983 0.011 0.012
Chain 1: 3200 -18053.987 0.012 0.015
Chain 1: 3300 -18258.491 0.011 0.012
Chain 1: 3400 -17733.938 0.013 0.015
Chain 1: 3500 -18344.992 0.015 0.016
Chain 1: 3600 -17652.768 0.017 0.016
Chain 1: 3700 -18038.710 0.019 0.017
Chain 1: 3800 -17000.096 0.023 0.021
Chain 1: 3900 -16996.300 0.022 0.021
Chain 1: 4000 -17113.588 0.022 0.021
Chain 1: 4100 -17027.406 0.023 0.021
Chain 1: 4200 -16844.093 0.022 0.021
Chain 1: 4300 -16982.209 0.022 0.021
Chain 1: 4400 -16939.336 0.019 0.011
Chain 1: 4500 -16841.932 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001286 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.86 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48782.720 1.000 1.000
Chain 1: 200 -17841.537 1.367 1.734
Chain 1: 300 -19765.315 0.944 1.000
Chain 1: 400 -32434.736 0.806 1.000
Chain 1: 500 -18434.966 0.796 0.759
Chain 1: 600 -11725.223 0.759 0.759
Chain 1: 700 -15535.385 0.686 0.572
Chain 1: 800 -10800.244 0.655 0.572
Chain 1: 900 -10315.094 0.587 0.438
Chain 1: 1000 -11861.493 0.541 0.438
Chain 1: 1100 -27021.856 0.498 0.438
Chain 1: 1200 -10149.435 0.490 0.438
Chain 1: 1300 -21681.381 0.534 0.532 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1400 -12247.520 0.572 0.561 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1500 -9737.580 0.522 0.532 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1600 -11625.656 0.481 0.438
Chain 1: 1700 -10126.466 0.471 0.438
Chain 1: 1800 -10137.493 0.427 0.258
Chain 1: 1900 -10355.138 0.425 0.258
Chain 1: 2000 -15714.053 0.446 0.341
Chain 1: 2100 -9956.025 0.447 0.341
Chain 1: 2200 -9654.976 0.284 0.258
Chain 1: 2300 -17321.393 0.275 0.258
Chain 1: 2400 -8893.233 0.293 0.258
Chain 1: 2500 -10193.949 0.280 0.162
Chain 1: 2600 -9827.168 0.268 0.148
Chain 1: 2700 -8940.765 0.263 0.128
Chain 1: 2800 -9404.563 0.268 0.128
Chain 1: 2900 -14834.861 0.302 0.341
Chain 1: 3000 -13641.661 0.277 0.128
Chain 1: 3100 -9610.993 0.261 0.128
Chain 1: 3200 -16501.581 0.299 0.366
Chain 1: 3300 -9226.349 0.334 0.366
Chain 1: 3400 -12182.651 0.264 0.243
Chain 1: 3500 -8917.290 0.287 0.366
Chain 1: 3600 -9711.236 0.292 0.366
Chain 1: 3700 -9997.131 0.285 0.366
Chain 1: 3800 -8672.080 0.295 0.366
Chain 1: 3900 -9674.606 0.269 0.243
Chain 1: 4000 -8563.471 0.273 0.243
Chain 1: 4100 -12871.316 0.265 0.243
Chain 1: 4200 -10619.755 0.244 0.212
Chain 1: 4300 -9736.487 0.174 0.153
Chain 1: 4400 -9030.039 0.158 0.130
Chain 1: 4500 -9326.486 0.124 0.104
Chain 1: 4600 -13172.104 0.145 0.130
Chain 1: 4700 -8780.296 0.193 0.153
Chain 1: 4800 -8791.551 0.177 0.130
Chain 1: 4900 -9070.649 0.170 0.130
Chain 1: 5000 -8502.777 0.164 0.091
Chain 1: 5100 -10701.237 0.151 0.091
Chain 1: 5200 -11914.502 0.140 0.091
Chain 1: 5300 -9262.428 0.159 0.102
Chain 1: 5400 -13542.084 0.183 0.205
Chain 1: 5500 -14571.800 0.187 0.205
Chain 1: 5600 -8831.150 0.223 0.205
Chain 1: 5700 -13860.680 0.209 0.205
Chain 1: 5800 -8587.379 0.270 0.286
Chain 1: 5900 -10992.326 0.289 0.286
Chain 1: 6000 -8382.899 0.314 0.311
Chain 1: 6100 -8840.376 0.298 0.311
Chain 1: 6200 -8355.957 0.294 0.311
Chain 1: 6300 -8560.499 0.268 0.311
Chain 1: 6400 -12544.367 0.268 0.311
Chain 1: 6500 -9053.843 0.299 0.318
Chain 1: 6600 -8446.252 0.242 0.311
Chain 1: 6700 -8289.740 0.207 0.219
Chain 1: 6800 -10006.869 0.163 0.172
Chain 1: 6900 -9917.796 0.142 0.072
Chain 1: 7000 -14698.693 0.143 0.072
Chain 1: 7100 -8181.269 0.218 0.172
Chain 1: 7200 -9009.605 0.221 0.172
Chain 1: 7300 -10382.906 0.232 0.172
Chain 1: 7400 -11217.330 0.208 0.132
Chain 1: 7500 -9036.173 0.193 0.132
Chain 1: 7600 -8352.671 0.194 0.132
Chain 1: 7700 -8701.705 0.196 0.132
Chain 1: 7800 -11738.117 0.205 0.132
Chain 1: 7900 -10472.861 0.216 0.132
Chain 1: 8000 -10339.553 0.185 0.121
Chain 1: 8100 -8202.108 0.131 0.121
Chain 1: 8200 -8110.774 0.123 0.121
Chain 1: 8300 -8333.322 0.113 0.082
Chain 1: 8400 -9695.331 0.119 0.121
Chain 1: 8500 -11115.663 0.108 0.121
Chain 1: 8600 -8064.766 0.138 0.128
Chain 1: 8700 -9601.767 0.150 0.140
Chain 1: 8800 -8221.627 0.141 0.140
Chain 1: 8900 -12941.262 0.165 0.160
Chain 1: 9000 -8464.767 0.217 0.168
Chain 1: 9100 -8141.876 0.195 0.160
Chain 1: 9200 -8526.298 0.198 0.160
Chain 1: 9300 -8909.395 0.200 0.160
Chain 1: 9400 -8891.575 0.186 0.160
Chain 1: 9500 -8449.523 0.178 0.160
Chain 1: 9600 -9235.030 0.149 0.085
Chain 1: 9700 -11065.108 0.149 0.085
Chain 1: 9800 -11402.361 0.136 0.052
Chain 1: 9900 -8331.411 0.136 0.052
Chain 1: 10000 -8127.314 0.086 0.045
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001868 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 18.68 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58113.534 1.000 1.000
Chain 1: 200 -17559.752 1.655 2.309
Chain 1: 300 -8610.699 1.450 1.039
Chain 1: 400 -8204.412 1.100 1.039
Chain 1: 500 -8092.295 0.882 1.000
Chain 1: 600 -8831.073 0.749 1.000
Chain 1: 700 -8135.386 0.654 0.086
Chain 1: 800 -8016.995 0.575 0.086
Chain 1: 900 -7888.296 0.512 0.084
Chain 1: 1000 -7733.720 0.463 0.084
Chain 1: 1100 -7882.122 0.365 0.050
Chain 1: 1200 -7782.074 0.135 0.020
Chain 1: 1300 -7763.149 0.032 0.019
Chain 1: 1400 -7884.894 0.028 0.016
Chain 1: 1500 -7579.056 0.031 0.019
Chain 1: 1600 -7769.293 0.025 0.019
Chain 1: 1700 -7522.652 0.020 0.019
Chain 1: 1800 -7608.690 0.019 0.019
Chain 1: 1900 -7626.440 0.018 0.019
Chain 1: 2000 -7598.585 0.016 0.015
Chain 1: 2100 -7619.368 0.015 0.013
Chain 1: 2200 -7693.793 0.015 0.011
Chain 1: 2300 -7567.491 0.016 0.015
Chain 1: 2400 -7634.106 0.015 0.011
Chain 1: 2500 -7501.534 0.013 0.011
Chain 1: 2600 -7543.392 0.011 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002751 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 27.51 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85832.653 1.000 1.000
Chain 1: 200 -13326.654 3.220 5.441
Chain 1: 300 -9766.338 2.268 1.000
Chain 1: 400 -10679.288 1.723 1.000
Chain 1: 500 -8682.668 1.424 0.365
Chain 1: 600 -8313.877 1.194 0.365
Chain 1: 700 -8381.250 1.025 0.230
Chain 1: 800 -8814.022 0.903 0.230
Chain 1: 900 -8574.176 0.806 0.085
Chain 1: 1000 -8344.856 0.728 0.085
Chain 1: 1100 -8690.947 0.632 0.049 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8208.416 0.094 0.049
Chain 1: 1300 -8362.018 0.059 0.044
Chain 1: 1400 -8439.169 0.051 0.040
Chain 1: 1500 -8368.064 0.029 0.028
Chain 1: 1600 -8367.237 0.025 0.027
Chain 1: 1700 -8292.549 0.025 0.027
Chain 1: 1800 -8181.854 0.021 0.018
Chain 1: 1900 -8299.769 0.020 0.014
Chain 1: 2000 -8260.103 0.018 0.014
Chain 1: 2100 -8388.888 0.015 0.014
Chain 1: 2200 -8179.992 0.012 0.014
Chain 1: 2300 -8322.366 0.012 0.014
Chain 1: 2400 -8336.462 0.011 0.014
Chain 1: 2500 -8304.291 0.011 0.014
Chain 1: 2600 -8303.390 0.011 0.014
Chain 1: 2700 -8211.493 0.011 0.014
Chain 1: 2800 -8186.570 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003412 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.12 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8389641.587 1.000 1.000
Chain 1: 200 -1582577.584 2.651 4.301
Chain 1: 300 -890761.416 2.026 1.000
Chain 1: 400 -458183.408 1.756 1.000
Chain 1: 500 -358583.239 1.460 0.944
Chain 1: 600 -233438.038 1.306 0.944
Chain 1: 700 -119324.877 1.256 0.944
Chain 1: 800 -86459.185 1.147 0.944
Chain 1: 900 -66737.269 1.052 0.777
Chain 1: 1000 -51483.600 0.976 0.777
Chain 1: 1100 -38920.153 0.909 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38085.747 0.481 0.380
Chain 1: 1300 -26012.980 0.450 0.380
Chain 1: 1400 -25726.383 0.356 0.323
Chain 1: 1500 -22307.069 0.344 0.323
Chain 1: 1600 -21520.649 0.294 0.296
Chain 1: 1700 -20391.605 0.204 0.296
Chain 1: 1800 -20334.800 0.166 0.153
Chain 1: 1900 -20660.407 0.138 0.055
Chain 1: 2000 -19170.977 0.116 0.055
Chain 1: 2100 -19409.303 0.085 0.037
Chain 1: 2200 -19635.746 0.084 0.037
Chain 1: 2300 -19253.092 0.040 0.020
Chain 1: 2400 -19025.333 0.040 0.020
Chain 1: 2500 -18827.512 0.025 0.016
Chain 1: 2600 -18458.080 0.024 0.016
Chain 1: 2700 -18415.069 0.018 0.012
Chain 1: 2800 -18132.248 0.020 0.016
Chain 1: 2900 -18413.318 0.020 0.015
Chain 1: 3000 -18399.464 0.012 0.012
Chain 1: 3100 -18484.428 0.011 0.012
Chain 1: 3200 -18175.386 0.012 0.015
Chain 1: 3300 -18379.855 0.011 0.012
Chain 1: 3400 -17855.385 0.013 0.015
Chain 1: 3500 -18466.437 0.015 0.016
Chain 1: 3600 -17774.169 0.017 0.016
Chain 1: 3700 -18160.254 0.019 0.017
Chain 1: 3800 -17121.656 0.023 0.021
Chain 1: 3900 -17117.859 0.022 0.021
Chain 1: 4000 -17235.134 0.022 0.021
Chain 1: 4100 -17149.030 0.022 0.021
Chain 1: 4200 -16965.594 0.022 0.021
Chain 1: 4300 -17103.736 0.021 0.021
Chain 1: 4400 -17060.864 0.019 0.011
Chain 1: 4500 -16963.471 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001541 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.41 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12464.975 1.000 1.000
Chain 1: 200 -9388.913 0.664 1.000
Chain 1: 300 -8031.843 0.499 0.328
Chain 1: 400 -8213.078 0.380 0.328
Chain 1: 500 -8200.559 0.304 0.169
Chain 1: 600 -7981.155 0.258 0.169
Chain 1: 700 -7887.660 0.223 0.027
Chain 1: 800 -7918.553 0.195 0.027
Chain 1: 900 -8034.287 0.175 0.022
Chain 1: 1000 -7924.784 0.159 0.022
Chain 1: 1100 -7969.065 0.060 0.014
Chain 1: 1200 -7909.344 0.028 0.014
Chain 1: 1300 -7850.962 0.012 0.012
Chain 1: 1400 -7876.258 0.010 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001391 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.91 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58164.047 1.000 1.000
Chain 1: 200 -17763.243 1.637 2.274
Chain 1: 300 -8712.081 1.438 1.039
Chain 1: 400 -8148.160 1.096 1.039
Chain 1: 500 -8386.830 0.882 1.000
Chain 1: 600 -8711.074 0.741 1.000
Chain 1: 700 -7713.066 0.654 0.129
Chain 1: 800 -8237.565 0.580 0.129
Chain 1: 900 -8060.333 0.518 0.069
Chain 1: 1000 -7821.672 0.469 0.069
Chain 1: 1100 -7674.362 0.371 0.064
Chain 1: 1200 -7643.415 0.144 0.037
Chain 1: 1300 -7774.620 0.042 0.031
Chain 1: 1400 -7854.444 0.036 0.028
Chain 1: 1500 -7565.413 0.037 0.031
Chain 1: 1600 -7527.859 0.034 0.022
Chain 1: 1700 -7545.039 0.021 0.019
Chain 1: 1800 -7601.540 0.016 0.017
Chain 1: 1900 -7615.313 0.014 0.010
Chain 1: 2000 -7608.162 0.011 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003275 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.75 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87155.765 1.000 1.000
Chain 1: 200 -13578.782 3.209 5.419
Chain 1: 300 -9900.677 2.263 1.000
Chain 1: 400 -10939.439 1.721 1.000
Chain 1: 500 -8886.737 1.423 0.372
Chain 1: 600 -8347.953 1.197 0.372
Chain 1: 700 -8492.363 1.028 0.231
Chain 1: 800 -9021.871 0.907 0.231
Chain 1: 900 -8703.819 0.810 0.095
Chain 1: 1000 -8654.546 0.730 0.095
Chain 1: 1100 -8633.212 0.630 0.065 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8332.403 0.092 0.059
Chain 1: 1300 -8572.587 0.058 0.037
Chain 1: 1400 -8597.404 0.048 0.036
Chain 1: 1500 -8445.157 0.027 0.028
Chain 1: 1600 -8558.332 0.022 0.018
Chain 1: 1700 -8633.018 0.021 0.018
Chain 1: 1800 -8207.221 0.020 0.018
Chain 1: 1900 -8309.561 0.018 0.013
Chain 1: 2000 -8284.239 0.018 0.013
Chain 1: 2100 -8410.904 0.019 0.015
Chain 1: 2200 -8210.695 0.018 0.015
Chain 1: 2300 -8304.603 0.016 0.013
Chain 1: 2400 -8372.827 0.017 0.013
Chain 1: 2500 -8319.035 0.015 0.012
Chain 1: 2600 -8321.166 0.014 0.011
Chain 1: 2700 -8237.543 0.014 0.011
Chain 1: 2800 -8196.426 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003118 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.18 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8431299.533 1.000 1.000
Chain 1: 200 -1590997.367 2.650 4.299
Chain 1: 300 -891506.024 2.028 1.000
Chain 1: 400 -457470.265 1.758 1.000
Chain 1: 500 -357187.524 1.463 0.949
Chain 1: 600 -232322.938 1.308 0.949
Chain 1: 700 -118957.159 1.258 0.949
Chain 1: 800 -86215.279 1.148 0.949
Chain 1: 900 -66648.321 1.053 0.785
Chain 1: 1000 -51505.952 0.977 0.785
Chain 1: 1100 -39035.072 0.909 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38223.159 0.481 0.380
Chain 1: 1300 -26236.834 0.448 0.380
Chain 1: 1400 -25962.737 0.355 0.319
Chain 1: 1500 -22562.981 0.342 0.319
Chain 1: 1600 -21783.500 0.291 0.294
Chain 1: 1700 -20664.197 0.202 0.294
Chain 1: 1800 -20610.062 0.164 0.151
Chain 1: 1900 -20936.368 0.136 0.054
Chain 1: 2000 -19450.580 0.114 0.054
Chain 1: 2100 -19689.133 0.084 0.036
Chain 1: 2200 -19914.808 0.083 0.036
Chain 1: 2300 -19532.623 0.039 0.020
Chain 1: 2400 -19304.708 0.039 0.020
Chain 1: 2500 -19106.269 0.025 0.016
Chain 1: 2600 -18736.798 0.023 0.016
Chain 1: 2700 -18693.908 0.018 0.012
Chain 1: 2800 -18410.396 0.019 0.015
Chain 1: 2900 -18691.723 0.019 0.015
Chain 1: 3000 -18678.067 0.012 0.012
Chain 1: 3100 -18762.984 0.011 0.012
Chain 1: 3200 -18453.707 0.012 0.015
Chain 1: 3300 -18658.438 0.011 0.012
Chain 1: 3400 -18133.165 0.012 0.015
Chain 1: 3500 -18745.123 0.015 0.015
Chain 1: 3600 -18051.758 0.017 0.015
Chain 1: 3700 -18438.491 0.018 0.017
Chain 1: 3800 -17397.934 0.023 0.021
Chain 1: 3900 -17393.991 0.021 0.021
Chain 1: 4000 -17511.396 0.022 0.021
Chain 1: 4100 -17425.019 0.022 0.021
Chain 1: 4200 -17241.261 0.021 0.021
Chain 1: 4300 -17379.739 0.021 0.021
Chain 1: 4400 -17336.556 0.018 0.011
Chain 1: 4500 -17238.987 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00128 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.8 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49013.031 1.000 1.000
Chain 1: 200 -23526.726 1.042 1.083
Chain 1: 300 -14481.643 0.903 1.000
Chain 1: 400 -14803.079 0.682 1.000
Chain 1: 500 -16498.786 0.566 0.625
Chain 1: 600 -12701.758 0.522 0.625
Chain 1: 700 -12705.950 0.447 0.299
Chain 1: 800 -11555.478 0.404 0.299
Chain 1: 900 -12591.113 0.368 0.103
Chain 1: 1000 -30585.638 0.390 0.299
Chain 1: 1100 -10776.521 0.474 0.299
Chain 1: 1200 -14761.580 0.393 0.270
Chain 1: 1300 -12225.784 0.351 0.207
Chain 1: 1400 -10709.208 0.363 0.207
Chain 1: 1500 -12640.286 0.368 0.207
Chain 1: 1600 -10870.784 0.354 0.163
Chain 1: 1700 -11362.618 0.359 0.163
Chain 1: 1800 -13753.919 0.366 0.174
Chain 1: 1900 -10936.187 0.384 0.207
Chain 1: 2000 -10039.887 0.334 0.174
Chain 1: 2100 -10136.992 0.151 0.163
Chain 1: 2200 -11026.273 0.132 0.153
Chain 1: 2300 -10937.741 0.112 0.142
Chain 1: 2400 -11646.515 0.104 0.089
Chain 1: 2500 -9465.262 0.112 0.089
Chain 1: 2600 -9667.208 0.097 0.081
Chain 1: 2700 -17093.559 0.137 0.089
Chain 1: 2800 -9193.194 0.205 0.089
Chain 1: 2900 -14098.908 0.214 0.089
Chain 1: 3000 -11304.758 0.230 0.230
Chain 1: 3100 -10206.820 0.240 0.230
Chain 1: 3200 -9531.763 0.239 0.230
Chain 1: 3300 -9214.786 0.241 0.230
Chain 1: 3400 -20993.209 0.291 0.247
Chain 1: 3500 -9160.315 0.398 0.348
Chain 1: 3600 -9192.970 0.396 0.348
Chain 1: 3700 -20261.572 0.407 0.348
Chain 1: 3800 -10607.739 0.412 0.348
Chain 1: 3900 -9797.446 0.386 0.247
Chain 1: 4000 -8748.640 0.373 0.120
Chain 1: 4100 -12186.133 0.390 0.282
Chain 1: 4200 -9440.580 0.412 0.291
Chain 1: 4300 -12520.996 0.433 0.291
Chain 1: 4400 -8889.169 0.418 0.291
Chain 1: 4500 -10761.973 0.306 0.282
Chain 1: 4600 -9644.826 0.318 0.282
Chain 1: 4700 -10067.599 0.267 0.246
Chain 1: 4800 -9813.896 0.179 0.174
Chain 1: 4900 -9220.130 0.177 0.174
Chain 1: 5000 -15781.566 0.207 0.246
Chain 1: 5100 -8752.217 0.259 0.246
Chain 1: 5200 -16621.223 0.277 0.246
Chain 1: 5300 -8948.994 0.338 0.409
Chain 1: 5400 -8746.421 0.299 0.174
Chain 1: 5500 -11703.044 0.307 0.253
Chain 1: 5600 -8533.817 0.333 0.371
Chain 1: 5700 -9055.821 0.334 0.371
Chain 1: 5800 -12148.249 0.357 0.371
Chain 1: 5900 -14707.716 0.368 0.371
Chain 1: 6000 -11030.713 0.360 0.333
Chain 1: 6100 -10490.809 0.285 0.255
Chain 1: 6200 -12087.677 0.251 0.253
Chain 1: 6300 -8506.028 0.207 0.253
Chain 1: 6400 -12478.946 0.237 0.255
Chain 1: 6500 -8637.085 0.256 0.318
Chain 1: 6600 -8468.193 0.221 0.255
Chain 1: 6700 -8625.918 0.217 0.255
Chain 1: 6800 -8628.850 0.191 0.174
Chain 1: 6900 -8708.477 0.175 0.132
Chain 1: 7000 -12310.476 0.171 0.132
Chain 1: 7100 -8255.314 0.215 0.293
Chain 1: 7200 -9549.722 0.215 0.293
Chain 1: 7300 -8450.754 0.186 0.136
Chain 1: 7400 -8586.457 0.156 0.130
Chain 1: 7500 -10743.660 0.131 0.130
Chain 1: 7600 -8643.482 0.154 0.136
Chain 1: 7700 -8930.203 0.155 0.136
Chain 1: 7800 -9695.658 0.163 0.136
Chain 1: 7900 -8539.417 0.176 0.136
Chain 1: 8000 -8416.313 0.148 0.135
Chain 1: 8100 -9715.576 0.112 0.134
Chain 1: 8200 -9358.493 0.102 0.130
Chain 1: 8300 -8758.499 0.096 0.079
Chain 1: 8400 -8567.008 0.097 0.079
Chain 1: 8500 -8286.528 0.080 0.069
Chain 1: 8600 -12358.517 0.089 0.069
Chain 1: 8700 -8847.707 0.125 0.079
Chain 1: 8800 -8142.096 0.126 0.087
Chain 1: 8900 -9024.185 0.122 0.087
Chain 1: 9000 -11272.801 0.141 0.098
Chain 1: 9100 -8529.381 0.159 0.098
Chain 1: 9200 -8863.306 0.159 0.098
Chain 1: 9300 -8845.861 0.153 0.098
Chain 1: 9400 -10379.602 0.165 0.148
Chain 1: 9500 -8703.147 0.181 0.193
Chain 1: 9600 -8374.004 0.152 0.148
Chain 1: 9700 -8972.712 0.119 0.098
Chain 1: 9800 -11347.583 0.131 0.148
Chain 1: 9900 -9365.534 0.143 0.193
Chain 1: 10000 -10713.970 0.135 0.148
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001376 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.76 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57437.018 1.000 1.000
Chain 1: 200 -17434.385 1.647 2.294
Chain 1: 300 -8759.319 1.428 1.000
Chain 1: 400 -8145.167 1.090 1.000
Chain 1: 500 -8771.414 0.886 0.990
Chain 1: 600 -8351.022 0.747 0.990
Chain 1: 700 -7862.157 0.649 0.075
Chain 1: 800 -8501.968 0.577 0.075
Chain 1: 900 -8030.542 0.520 0.075
Chain 1: 1000 -7947.070 0.469 0.075
Chain 1: 1100 -7716.952 0.372 0.071
Chain 1: 1200 -7853.688 0.144 0.062
Chain 1: 1300 -7661.683 0.048 0.059
Chain 1: 1400 -7903.969 0.043 0.050
Chain 1: 1500 -7582.765 0.040 0.042
Chain 1: 1600 -7657.768 0.036 0.031
Chain 1: 1700 -7541.915 0.031 0.030
Chain 1: 1800 -7573.174 0.024 0.025
Chain 1: 1900 -7614.243 0.019 0.017
Chain 1: 2000 -7623.001 0.018 0.017
Chain 1: 2100 -7587.304 0.016 0.015
Chain 1: 2200 -7729.222 0.016 0.015
Chain 1: 2300 -7534.192 0.016 0.015
Chain 1: 2400 -7674.536 0.015 0.015
Chain 1: 2500 -7440.653 0.013 0.015
Chain 1: 2600 -7559.762 0.014 0.016
Chain 1: 2700 -7497.724 0.013 0.016
Chain 1: 2800 -7591.095 0.014 0.016
Chain 1: 2900 -7407.572 0.016 0.018
Chain 1: 3000 -7538.730 0.018 0.018
Chain 1: 3100 -7536.690 0.017 0.018
Chain 1: 3200 -7725.959 0.018 0.018
Chain 1: 3300 -7475.853 0.019 0.018
Chain 1: 3400 -7672.143 0.019 0.024
Chain 1: 3500 -7451.079 0.019 0.024
Chain 1: 3600 -7512.650 0.018 0.024
Chain 1: 3700 -7463.479 0.018 0.024
Chain 1: 3800 -7471.800 0.017 0.024
Chain 1: 3900 -7448.418 0.015 0.017
Chain 1: 4000 -7429.509 0.014 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003194 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.94 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86066.488 1.000 1.000
Chain 1: 200 -13590.159 3.167 5.333
Chain 1: 300 -9952.485 2.233 1.000
Chain 1: 400 -10881.506 1.696 1.000
Chain 1: 500 -8778.579 1.405 0.366
Chain 1: 600 -8407.186 1.178 0.366
Chain 1: 700 -8645.625 1.014 0.240
Chain 1: 800 -9237.702 0.895 0.240
Chain 1: 900 -8795.295 0.801 0.085
Chain 1: 1000 -8572.005 0.724 0.085
Chain 1: 1100 -8740.046 0.625 0.064 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8421.511 0.096 0.050
Chain 1: 1300 -8660.521 0.062 0.044
Chain 1: 1400 -8666.821 0.054 0.038
Chain 1: 1500 -8515.889 0.032 0.028
Chain 1: 1600 -8629.270 0.028 0.028
Chain 1: 1700 -8711.904 0.027 0.026
Chain 1: 1800 -8297.612 0.025 0.026
Chain 1: 1900 -8394.254 0.021 0.019
Chain 1: 2000 -8367.723 0.019 0.018
Chain 1: 2100 -8490.622 0.019 0.014
Chain 1: 2200 -8310.309 0.017 0.014
Chain 1: 2300 -8389.009 0.015 0.013
Chain 1: 2400 -8458.751 0.016 0.013
Chain 1: 2500 -8404.326 0.015 0.012
Chain 1: 2600 -8403.970 0.013 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003068 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.68 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8392582.343 1.000 1.000
Chain 1: 200 -1584523.440 2.648 4.297
Chain 1: 300 -890970.252 2.025 1.000
Chain 1: 400 -457953.339 1.755 1.000
Chain 1: 500 -358209.430 1.460 0.946
Chain 1: 600 -233206.997 1.306 0.946
Chain 1: 700 -119355.123 1.256 0.946
Chain 1: 800 -86559.605 1.146 0.946
Chain 1: 900 -66893.514 1.051 0.778
Chain 1: 1000 -51686.452 0.976 0.778
Chain 1: 1100 -39160.521 0.908 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38336.855 0.480 0.379
Chain 1: 1300 -26288.080 0.448 0.379
Chain 1: 1400 -26007.157 0.355 0.320
Chain 1: 1500 -22592.778 0.342 0.320
Chain 1: 1600 -21809.122 0.292 0.294
Chain 1: 1700 -20682.050 0.202 0.294
Chain 1: 1800 -20626.138 0.164 0.151
Chain 1: 1900 -20952.251 0.136 0.054
Chain 1: 2000 -19463.033 0.115 0.054
Chain 1: 2100 -19701.461 0.084 0.036
Chain 1: 2200 -19927.969 0.083 0.036
Chain 1: 2300 -19545.092 0.039 0.020
Chain 1: 2400 -19317.161 0.039 0.020
Chain 1: 2500 -19119.251 0.025 0.016
Chain 1: 2600 -18749.409 0.023 0.016
Chain 1: 2700 -18706.388 0.018 0.012
Chain 1: 2800 -18423.258 0.019 0.015
Chain 1: 2900 -18704.513 0.019 0.015
Chain 1: 3000 -18690.694 0.012 0.012
Chain 1: 3100 -18775.686 0.011 0.012
Chain 1: 3200 -18466.385 0.012 0.015
Chain 1: 3300 -18671.106 0.011 0.012
Chain 1: 3400 -18146.067 0.012 0.015
Chain 1: 3500 -18757.936 0.015 0.015
Chain 1: 3600 -18064.610 0.017 0.015
Chain 1: 3700 -18451.392 0.018 0.017
Chain 1: 3800 -17411.160 0.023 0.021
Chain 1: 3900 -17407.312 0.021 0.021
Chain 1: 4000 -17524.597 0.022 0.021
Chain 1: 4100 -17438.356 0.022 0.021
Chain 1: 4200 -17254.630 0.021 0.021
Chain 1: 4300 -17393.007 0.021 0.021
Chain 1: 4400 -17349.830 0.018 0.011
Chain 1: 4500 -17252.368 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001321 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.21 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48946.085 1.000 1.000
Chain 1: 200 -18169.079 1.347 1.694
Chain 1: 300 -16965.898 0.922 1.000
Chain 1: 400 -13200.263 0.763 1.000
Chain 1: 500 -16329.265 0.648 0.285
Chain 1: 600 -19464.644 0.567 0.285
Chain 1: 700 -14969.957 0.529 0.285
Chain 1: 800 -10877.008 0.510 0.300
Chain 1: 900 -12925.369 0.471 0.285
Chain 1: 1000 -17026.837 0.448 0.285
Chain 1: 1100 -14426.978 0.366 0.241
Chain 1: 1200 -12296.139 0.214 0.192
Chain 1: 1300 -12238.722 0.207 0.192
Chain 1: 1400 -11254.919 0.187 0.180
Chain 1: 1500 -11103.771 0.170 0.173
Chain 1: 1600 -11817.153 0.160 0.173
Chain 1: 1700 -12589.061 0.136 0.158
Chain 1: 1800 -9655.416 0.128 0.158
Chain 1: 1900 -14768.616 0.147 0.173
Chain 1: 2000 -10188.695 0.168 0.173
Chain 1: 2100 -10497.756 0.153 0.087
Chain 1: 2200 -11124.514 0.141 0.061
Chain 1: 2300 -9196.082 0.162 0.087
Chain 1: 2400 -14495.984 0.190 0.210
Chain 1: 2500 -10169.297 0.231 0.304
Chain 1: 2600 -9767.449 0.229 0.304
Chain 1: 2700 -9826.960 0.223 0.304
Chain 1: 2800 -9002.997 0.202 0.210
Chain 1: 2900 -9779.857 0.175 0.092
Chain 1: 3000 -9590.101 0.132 0.079
Chain 1: 3100 -9408.547 0.131 0.079
Chain 1: 3200 -13648.065 0.157 0.092
Chain 1: 3300 -15515.468 0.148 0.092
Chain 1: 3400 -14528.011 0.118 0.079
Chain 1: 3500 -9475.883 0.129 0.079
Chain 1: 3600 -9741.543 0.128 0.079
Chain 1: 3700 -9564.058 0.129 0.079
Chain 1: 3800 -9622.285 0.120 0.068
Chain 1: 3900 -10144.516 0.117 0.051
Chain 1: 4000 -14347.139 0.145 0.068
Chain 1: 4100 -8763.560 0.207 0.120
Chain 1: 4200 -11042.622 0.196 0.120
Chain 1: 4300 -9701.513 0.198 0.138
Chain 1: 4400 -8776.542 0.202 0.138
Chain 1: 4500 -8613.318 0.150 0.105
Chain 1: 4600 -12360.297 0.178 0.138
Chain 1: 4700 -9662.379 0.204 0.206
Chain 1: 4800 -13300.255 0.231 0.274
Chain 1: 4900 -9313.793 0.268 0.279
Chain 1: 5000 -11454.715 0.258 0.274
Chain 1: 5100 -8680.145 0.226 0.274
Chain 1: 5200 -10146.806 0.220 0.274
Chain 1: 5300 -12775.975 0.227 0.274
Chain 1: 5400 -11595.101 0.226 0.274
Chain 1: 5500 -9808.039 0.242 0.274
Chain 1: 5600 -8366.871 0.229 0.206
Chain 1: 5700 -13223.475 0.238 0.206
Chain 1: 5800 -11614.610 0.225 0.187
Chain 1: 5900 -8245.833 0.223 0.187
Chain 1: 6000 -12661.104 0.239 0.206
Chain 1: 6100 -8058.665 0.264 0.206
Chain 1: 6200 -9224.635 0.262 0.206
Chain 1: 6300 -8478.636 0.250 0.182
Chain 1: 6400 -12005.184 0.270 0.294
Chain 1: 6500 -11237.577 0.258 0.294
Chain 1: 6600 -12655.075 0.252 0.294
Chain 1: 6700 -10165.838 0.240 0.245
Chain 1: 6800 -8994.056 0.239 0.245
Chain 1: 6900 -11531.367 0.220 0.220
Chain 1: 7000 -8257.859 0.225 0.220
Chain 1: 7100 -8107.020 0.170 0.130
Chain 1: 7200 -8363.781 0.160 0.130
Chain 1: 7300 -8039.140 0.156 0.130
Chain 1: 7400 -8356.364 0.130 0.112
Chain 1: 7500 -8042.785 0.127 0.112
Chain 1: 7600 -8440.983 0.121 0.047
Chain 1: 7700 -8402.450 0.097 0.040
Chain 1: 7800 -8537.951 0.085 0.039
Chain 1: 7900 -7977.130 0.070 0.039
Chain 1: 8000 -8275.538 0.034 0.038
Chain 1: 8100 -9372.448 0.044 0.039
Chain 1: 8200 -8617.341 0.050 0.040
Chain 1: 8300 -8249.113 0.050 0.045
Chain 1: 8400 -8172.909 0.047 0.045
Chain 1: 8500 -8257.326 0.044 0.045
Chain 1: 8600 -8301.039 0.040 0.036
Chain 1: 8700 -8293.419 0.040 0.036
Chain 1: 8800 -8343.341 0.039 0.036
Chain 1: 8900 -8394.746 0.032 0.010
Chain 1: 9000 -8409.370 0.029 0.009 MEDIAN ELBO CONVERGED
Chain 1: Informational Message: The ELBO at a previous iteration is larger than the ELBO upon convergence!
Chain 1: This variational approximation may not have converged to a good optimum.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001458 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.58 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -63094.773 1.000 1.000
Chain 1: 200 -18104.251 1.743 2.485
Chain 1: 300 -8747.474 1.518 1.070
Chain 1: 400 -8560.205 1.144 1.070
Chain 1: 500 -8478.017 0.917 1.000
Chain 1: 600 -8238.914 0.769 1.000
Chain 1: 700 -7929.307 0.665 0.039
Chain 1: 800 -8018.131 0.583 0.039
Chain 1: 900 -7731.371 0.523 0.037
Chain 1: 1000 -7760.323 0.471 0.037
Chain 1: 1100 -7590.292 0.373 0.029
Chain 1: 1200 -7693.541 0.126 0.022
Chain 1: 1300 -7714.036 0.019 0.022
Chain 1: 1400 -7842.380 0.018 0.016
Chain 1: 1500 -7582.461 0.021 0.022
Chain 1: 1600 -7774.640 0.020 0.022
Chain 1: 1700 -7502.368 0.020 0.022
Chain 1: 1800 -7583.755 0.020 0.022
Chain 1: 1900 -7582.326 0.016 0.016
Chain 1: 2000 -7655.589 0.017 0.016
Chain 1: 2100 -7627.674 0.015 0.013
Chain 1: 2200 -7705.852 0.015 0.011
Chain 1: 2300 -7578.347 0.016 0.016
Chain 1: 2400 -7636.577 0.015 0.011
Chain 1: 2500 -7570.448 0.013 0.010
Chain 1: 2600 -7536.255 0.011 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003032 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.32 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86464.771 1.000 1.000
Chain 1: 200 -13402.590 3.226 5.451
Chain 1: 300 -9695.600 2.278 1.000
Chain 1: 400 -10960.472 1.737 1.000
Chain 1: 500 -8668.819 1.443 0.382
Chain 1: 600 -8091.393 1.214 0.382
Chain 1: 700 -8275.373 1.044 0.264
Chain 1: 800 -8455.287 0.916 0.264
Chain 1: 900 -8497.931 0.815 0.115
Chain 1: 1000 -8077.822 0.739 0.115
Chain 1: 1100 -8511.490 0.644 0.071 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8082.826 0.104 0.053
Chain 1: 1300 -8320.007 0.068 0.052
Chain 1: 1400 -8340.914 0.057 0.051
Chain 1: 1500 -8210.134 0.032 0.029
Chain 1: 1600 -8322.433 0.026 0.022
Chain 1: 1700 -8395.706 0.025 0.021
Chain 1: 1800 -7959.659 0.028 0.029
Chain 1: 1900 -8065.183 0.029 0.029
Chain 1: 2000 -8041.125 0.024 0.016
Chain 1: 2100 -8007.756 0.020 0.013
Chain 1: 2200 -7983.724 0.015 0.013
Chain 1: 2300 -8119.130 0.014 0.013
Chain 1: 2400 -7966.628 0.015 0.013
Chain 1: 2500 -8035.801 0.014 0.013
Chain 1: 2600 -7954.128 0.014 0.010
Chain 1: 2700 -7985.443 0.014 0.010
Chain 1: 2800 -7945.551 0.009 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00346 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.6 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8445482.975 1.000 1.000
Chain 1: 200 -1590563.944 2.655 4.310
Chain 1: 300 -891034.861 2.032 1.000
Chain 1: 400 -457267.841 1.761 1.000
Chain 1: 500 -357305.129 1.465 0.949
Chain 1: 600 -232340.959 1.310 0.949
Chain 1: 700 -118843.671 1.259 0.949
Chain 1: 800 -86120.913 1.150 0.949
Chain 1: 900 -66527.468 1.055 0.785
Chain 1: 1000 -51378.884 0.979 0.785
Chain 1: 1100 -38903.752 0.911 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38089.069 0.482 0.380
Chain 1: 1300 -26088.618 0.449 0.380
Chain 1: 1400 -25813.497 0.355 0.321
Chain 1: 1500 -22411.738 0.343 0.321
Chain 1: 1600 -21631.956 0.292 0.295
Chain 1: 1700 -20510.490 0.202 0.295
Chain 1: 1800 -20455.967 0.165 0.152
Chain 1: 1900 -20782.509 0.137 0.055
Chain 1: 2000 -19295.545 0.115 0.055
Chain 1: 2100 -19533.836 0.084 0.036
Chain 1: 2200 -19760.137 0.083 0.036
Chain 1: 2300 -19377.410 0.039 0.020
Chain 1: 2400 -19149.407 0.039 0.020
Chain 1: 2500 -18951.207 0.025 0.016
Chain 1: 2600 -18581.252 0.024 0.016
Chain 1: 2700 -18538.198 0.018 0.012
Chain 1: 2800 -18254.789 0.020 0.016
Chain 1: 2900 -18536.146 0.020 0.015
Chain 1: 3000 -18522.385 0.012 0.012
Chain 1: 3100 -18607.406 0.011 0.012
Chain 1: 3200 -18297.907 0.012 0.015
Chain 1: 3300 -18502.769 0.011 0.012
Chain 1: 3400 -17977.264 0.013 0.015
Chain 1: 3500 -18589.686 0.015 0.016
Chain 1: 3600 -17895.642 0.017 0.016
Chain 1: 3700 -18282.934 0.019 0.017
Chain 1: 3800 -17241.450 0.023 0.021
Chain 1: 3900 -17237.519 0.022 0.021
Chain 1: 4000 -17354.877 0.022 0.021
Chain 1: 4100 -17268.557 0.022 0.021
Chain 1: 4200 -17084.538 0.022 0.021
Chain 1: 4300 -17223.149 0.021 0.021
Chain 1: 4400 -17179.755 0.019 0.011
Chain 1: 4500 -17082.220 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001255 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.55 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13330.717 1.000 1.000
Chain 1: 200 -9693.207 0.688 1.000
Chain 1: 300 -8163.689 0.521 0.375
Chain 1: 400 -8498.784 0.401 0.375
Chain 1: 500 -8034.915 0.332 0.187
Chain 1: 600 -8159.653 0.279 0.187
Chain 1: 700 -8004.328 0.242 0.058
Chain 1: 800 -8035.781 0.212 0.058
Chain 1: 900 -8047.905 0.189 0.039
Chain 1: 1000 -8113.894 0.171 0.039
Chain 1: 1100 -7945.009 0.073 0.021
Chain 1: 1200 -8016.413 0.036 0.019
Chain 1: 1300 -7984.282 0.018 0.015
Chain 1: 1400 -8002.234 0.014 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001471 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.71 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56545.283 1.000 1.000
Chain 1: 200 -17956.041 1.575 2.149
Chain 1: 300 -9112.315 1.373 1.000
Chain 1: 400 -8403.919 1.051 1.000
Chain 1: 500 -8446.164 0.842 0.971
Chain 1: 600 -9080.923 0.713 0.971
Chain 1: 700 -8218.519 0.626 0.105
Chain 1: 800 -8254.682 0.549 0.105
Chain 1: 900 -7424.733 0.500 0.105
Chain 1: 1000 -8124.872 0.459 0.105
Chain 1: 1100 -7777.920 0.363 0.086
Chain 1: 1200 -7775.461 0.148 0.084
Chain 1: 1300 -7856.299 0.052 0.070
Chain 1: 1400 -8081.919 0.047 0.045
Chain 1: 1500 -7567.144 0.053 0.068
Chain 1: 1600 -7607.349 0.046 0.045
Chain 1: 1700 -7491.498 0.037 0.028
Chain 1: 1800 -7694.352 0.040 0.028
Chain 1: 1900 -7584.847 0.030 0.026
Chain 1: 2000 -7678.248 0.022 0.015
Chain 1: 2100 -7455.554 0.021 0.015
Chain 1: 2200 -7896.417 0.027 0.026
Chain 1: 2300 -7590.427 0.030 0.028
Chain 1: 2400 -7552.012 0.027 0.026
Chain 1: 2500 -7576.738 0.021 0.015
Chain 1: 2600 -7514.927 0.021 0.015
Chain 1: 2700 -7415.187 0.021 0.014
Chain 1: 2800 -7599.098 0.021 0.014
Chain 1: 2900 -7347.467 0.023 0.024
Chain 1: 3000 -7504.442 0.024 0.024
Chain 1: 3100 -7492.802 0.021 0.021
Chain 1: 3200 -7762.579 0.019 0.021
Chain 1: 3300 -7370.992 0.020 0.021
Chain 1: 3400 -7502.493 0.021 0.021
Chain 1: 3500 -7395.438 0.022 0.021
Chain 1: 3600 -7436.729 0.022 0.021
Chain 1: 3700 -7359.839 0.022 0.021
Chain 1: 3800 -7427.614 0.020 0.018
Chain 1: 3900 -7365.079 0.018 0.014
Chain 1: 4000 -7362.590 0.016 0.010
Chain 1: 4100 -7369.839 0.015 0.010
Chain 1: 4200 -7494.525 0.014 0.010
Chain 1: 4300 -7348.317 0.010 0.010
Chain 1: 4400 -7400.321 0.009 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002909 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.09 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -88117.654 1.000 1.000
Chain 1: 200 -14372.455 3.066 5.131
Chain 1: 300 -10433.579 2.170 1.000
Chain 1: 400 -13033.289 1.677 1.000
Chain 1: 500 -10093.452 1.400 0.378
Chain 1: 600 -8623.161 1.195 0.378
Chain 1: 700 -8567.248 1.025 0.291
Chain 1: 800 -9369.636 0.908 0.291
Chain 1: 900 -8972.627 0.812 0.199
Chain 1: 1000 -9595.658 0.737 0.199
Chain 1: 1100 -8998.844 0.644 0.171 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8419.301 0.138 0.086
Chain 1: 1300 -8942.402 0.106 0.069
Chain 1: 1400 -8796.190 0.087 0.066
Chain 1: 1500 -8724.332 0.059 0.065
Chain 1: 1600 -8728.621 0.042 0.058
Chain 1: 1700 -8904.032 0.043 0.058
Chain 1: 1800 -8413.625 0.041 0.058
Chain 1: 1900 -8521.808 0.037 0.058
Chain 1: 2000 -8531.522 0.031 0.020
Chain 1: 2100 -8691.572 0.026 0.018
Chain 1: 2200 -8379.659 0.023 0.018
Chain 1: 2300 -8463.938 0.018 0.017
Chain 1: 2400 -8558.956 0.018 0.013
Chain 1: 2500 -8449.838 0.018 0.013
Chain 1: 2600 -8501.743 0.019 0.013
Chain 1: 2700 -8411.061 0.018 0.013
Chain 1: 2800 -8381.772 0.012 0.011
Chain 1: 2900 -8471.529 0.012 0.011
Chain 1: 3000 -8407.766 0.013 0.011
Chain 1: 3100 -8357.256 0.012 0.011
Chain 1: 3200 -8310.961 0.008 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003559 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.59 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8420078.612 1.000 1.000
Chain 1: 200 -1585659.572 2.655 4.310
Chain 1: 300 -890296.121 2.030 1.000
Chain 1: 400 -457620.212 1.759 1.000
Chain 1: 500 -357712.488 1.463 0.945
Chain 1: 600 -232712.000 1.309 0.945
Chain 1: 700 -119564.530 1.257 0.945
Chain 1: 800 -86917.438 1.147 0.945
Chain 1: 900 -67390.913 1.052 0.781
Chain 1: 1000 -52305.023 0.975 0.781
Chain 1: 1100 -39875.635 0.906 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39080.756 0.478 0.376
Chain 1: 1300 -27104.212 0.444 0.376
Chain 1: 1400 -26838.605 0.350 0.312
Chain 1: 1500 -23440.496 0.337 0.312
Chain 1: 1600 -22664.003 0.286 0.290
Chain 1: 1700 -21543.837 0.197 0.288
Chain 1: 1800 -21490.856 0.160 0.145
Chain 1: 1900 -21818.485 0.132 0.052
Chain 1: 2000 -20329.945 0.111 0.052
Chain 1: 2100 -20568.690 0.081 0.034
Chain 1: 2200 -20795.365 0.080 0.034
Chain 1: 2300 -20411.940 0.037 0.019
Chain 1: 2400 -20183.491 0.037 0.019
Chain 1: 2500 -19985.029 0.024 0.015
Chain 1: 2600 -19614.063 0.022 0.015
Chain 1: 2700 -19570.865 0.017 0.012
Chain 1: 2800 -19286.708 0.019 0.015
Chain 1: 2900 -19568.607 0.019 0.014
Chain 1: 3000 -19554.866 0.011 0.012
Chain 1: 3100 -19640.000 0.011 0.011
Chain 1: 3200 -19329.799 0.011 0.014
Chain 1: 3300 -19535.280 0.010 0.011
Chain 1: 3400 -19008.334 0.012 0.014
Chain 1: 3500 -19622.831 0.014 0.015
Chain 1: 3600 -18926.092 0.016 0.015
Chain 1: 3700 -19315.261 0.018 0.016
Chain 1: 3800 -18269.551 0.022 0.020
Chain 1: 3900 -18265.477 0.021 0.020
Chain 1: 4000 -18382.873 0.021 0.020
Chain 1: 4100 -18296.231 0.021 0.020
Chain 1: 4200 -18111.392 0.021 0.020
Chain 1: 4300 -18250.646 0.020 0.020
Chain 1: 4400 -18206.486 0.018 0.010
Chain 1: 4500 -18108.764 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001267 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.67 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12007.673 1.000 1.000
Chain 1: 200 -8893.304 0.675 1.000
Chain 1: 300 -7834.667 0.495 0.350
Chain 1: 400 -7997.590 0.376 0.350
Chain 1: 500 -7856.611 0.305 0.135
Chain 1: 600 -7781.420 0.256 0.135
Chain 1: 700 -7706.890 0.220 0.020
Chain 1: 800 -7730.080 0.193 0.020
Chain 1: 900 -7747.216 0.172 0.018
Chain 1: 1000 -7755.698 0.155 0.018
Chain 1: 1100 -7824.816 0.056 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001491 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.91 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61338.303 1.000 1.000
Chain 1: 200 -17448.244 1.758 2.515
Chain 1: 300 -8675.049 1.509 1.011
Chain 1: 400 -8153.318 1.148 1.011
Chain 1: 500 -8264.830 0.921 1.000
Chain 1: 600 -7841.583 0.776 1.000
Chain 1: 700 -8026.723 0.669 0.064
Chain 1: 800 -8014.990 0.585 0.064
Chain 1: 900 -7832.941 0.523 0.054
Chain 1: 1000 -7645.685 0.473 0.054
Chain 1: 1100 -7580.182 0.374 0.024
Chain 1: 1200 -7730.835 0.124 0.023
Chain 1: 1300 -7542.379 0.026 0.023
Chain 1: 1400 -7773.634 0.022 0.023
Chain 1: 1500 -7563.436 0.024 0.024
Chain 1: 1600 -7467.875 0.020 0.023
Chain 1: 1700 -7459.590 0.017 0.023
Chain 1: 1800 -7490.013 0.018 0.023
Chain 1: 1900 -7546.810 0.016 0.019
Chain 1: 2000 -7530.464 0.014 0.013
Chain 1: 2100 -7606.596 0.014 0.013
Chain 1: 2200 -7642.760 0.012 0.010
Chain 1: 2300 -7533.620 0.011 0.010
Chain 1: 2400 -7553.295 0.009 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003181 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.81 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85984.567 1.000 1.000
Chain 1: 200 -13126.494 3.275 5.550
Chain 1: 300 -9580.630 2.307 1.000
Chain 1: 400 -10505.749 1.752 1.000
Chain 1: 500 -8508.343 1.449 0.370
Chain 1: 600 -8371.695 1.210 0.370
Chain 1: 700 -8209.032 1.040 0.235
Chain 1: 800 -8674.106 0.917 0.235
Chain 1: 900 -8402.551 0.818 0.088
Chain 1: 1000 -8151.534 0.740 0.088
Chain 1: 1100 -8475.591 0.643 0.054 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8148.634 0.092 0.040
Chain 1: 1300 -8318.038 0.057 0.038
Chain 1: 1400 -8304.217 0.049 0.032
Chain 1: 1500 -8212.742 0.026 0.031
Chain 1: 1600 -8309.051 0.026 0.031
Chain 1: 1700 -8397.367 0.025 0.031
Chain 1: 1800 -8012.018 0.024 0.031
Chain 1: 1900 -8114.401 0.023 0.020
Chain 1: 2000 -8084.337 0.020 0.013
Chain 1: 2100 -8217.885 0.018 0.013
Chain 1: 2200 -8002.910 0.016 0.013
Chain 1: 2300 -8144.194 0.016 0.013
Chain 1: 2400 -8155.756 0.016 0.013
Chain 1: 2500 -8123.993 0.015 0.013
Chain 1: 2600 -8122.647 0.014 0.013
Chain 1: 2700 -8031.530 0.014 0.013
Chain 1: 2800 -8008.742 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003021 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.21 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8432789.836 1.000 1.000
Chain 1: 200 -1590386.155 2.651 4.302
Chain 1: 300 -891920.145 2.028 1.000
Chain 1: 400 -457542.748 1.759 1.000
Chain 1: 500 -357109.098 1.463 0.949
Chain 1: 600 -232069.362 1.309 0.949
Chain 1: 700 -118538.484 1.259 0.949
Chain 1: 800 -85806.246 1.149 0.949
Chain 1: 900 -66204.683 1.054 0.783
Chain 1: 1000 -51043.391 0.979 0.783
Chain 1: 1100 -38564.561 0.911 0.539 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37743.497 0.483 0.381
Chain 1: 1300 -25754.171 0.451 0.381
Chain 1: 1400 -25476.907 0.357 0.324
Chain 1: 1500 -22078.371 0.345 0.324
Chain 1: 1600 -21298.637 0.294 0.297
Chain 1: 1700 -20179.340 0.204 0.296
Chain 1: 1800 -20124.933 0.166 0.154
Chain 1: 1900 -20450.489 0.138 0.055
Chain 1: 2000 -18966.481 0.116 0.055
Chain 1: 2100 -19204.554 0.085 0.037
Chain 1: 2200 -19429.994 0.084 0.037
Chain 1: 2300 -19048.282 0.040 0.020
Chain 1: 2400 -18820.636 0.040 0.020
Chain 1: 2500 -18622.453 0.026 0.016
Chain 1: 2600 -18253.369 0.024 0.016
Chain 1: 2700 -18210.647 0.019 0.012
Chain 1: 2800 -17927.578 0.020 0.016
Chain 1: 2900 -18208.569 0.020 0.015
Chain 1: 3000 -18194.870 0.012 0.012
Chain 1: 3100 -18279.710 0.011 0.012
Chain 1: 3200 -17970.817 0.012 0.015
Chain 1: 3300 -18175.262 0.011 0.012
Chain 1: 3400 -17650.809 0.013 0.015
Chain 1: 3500 -18261.577 0.015 0.016
Chain 1: 3600 -17569.778 0.017 0.016
Chain 1: 3700 -17955.348 0.019 0.017
Chain 1: 3800 -16917.300 0.023 0.021
Chain 1: 3900 -16913.486 0.022 0.021
Chain 1: 4000 -17030.828 0.023 0.021
Chain 1: 4100 -16944.602 0.023 0.021
Chain 1: 4200 -16761.421 0.022 0.021
Chain 1: 4300 -16899.443 0.022 0.021
Chain 1: 4400 -16856.663 0.019 0.011
Chain 1: 4500 -16759.265 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001301 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.01 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48472.996 1.000 1.000
Chain 1: 200 -17155.307 1.413 1.826
Chain 1: 300 -17129.485 0.942 1.000
Chain 1: 400 -12865.358 0.790 1.000
Chain 1: 500 -19492.689 0.700 0.340
Chain 1: 600 -13878.990 0.650 0.404
Chain 1: 700 -13872.527 0.558 0.340
Chain 1: 800 -12423.384 0.503 0.340
Chain 1: 900 -19535.120 0.487 0.340
Chain 1: 1000 -12151.054 0.499 0.364
Chain 1: 1100 -10972.797 0.410 0.340
Chain 1: 1200 -11902.951 0.235 0.331
Chain 1: 1300 -11263.642 0.241 0.331
Chain 1: 1400 -11565.820 0.210 0.117
Chain 1: 1500 -10231.182 0.189 0.117
Chain 1: 1600 -15753.149 0.184 0.117
Chain 1: 1700 -13118.126 0.204 0.130
Chain 1: 1800 -21661.769 0.232 0.201
Chain 1: 1900 -9593.884 0.321 0.201
Chain 1: 2000 -11080.668 0.274 0.134
Chain 1: 2100 -10625.572 0.267 0.134
Chain 1: 2200 -10743.290 0.260 0.134
Chain 1: 2300 -10016.116 0.262 0.134
Chain 1: 2400 -10994.075 0.268 0.134
Chain 1: 2500 -14445.342 0.279 0.201
Chain 1: 2600 -9262.016 0.300 0.201
Chain 1: 2700 -10298.932 0.290 0.134
Chain 1: 2800 -10016.319 0.253 0.101
Chain 1: 2900 -9375.410 0.135 0.089
Chain 1: 3000 -9113.867 0.124 0.073
Chain 1: 3100 -14598.132 0.157 0.089
Chain 1: 3200 -9748.488 0.206 0.101
Chain 1: 3300 -9844.000 0.200 0.101
Chain 1: 3400 -8912.770 0.201 0.104
Chain 1: 3500 -9401.384 0.182 0.101
Chain 1: 3600 -10478.209 0.137 0.101
Chain 1: 3700 -8672.198 0.148 0.103
Chain 1: 3800 -15338.196 0.188 0.104
Chain 1: 3900 -8537.469 0.261 0.208
Chain 1: 4000 -9426.294 0.268 0.208
Chain 1: 4100 -8517.998 0.241 0.107
Chain 1: 4200 -10066.477 0.206 0.107
Chain 1: 4300 -11858.218 0.220 0.151
Chain 1: 4400 -8760.494 0.245 0.154
Chain 1: 4500 -9995.180 0.253 0.154
Chain 1: 4600 -9949.183 0.243 0.154
Chain 1: 4700 -8366.447 0.241 0.154
Chain 1: 4800 -8706.220 0.201 0.151
Chain 1: 4900 -8709.191 0.122 0.124
Chain 1: 5000 -13949.972 0.150 0.151
Chain 1: 5100 -8718.609 0.199 0.154
Chain 1: 5200 -10069.811 0.197 0.151
Chain 1: 5300 -14405.675 0.212 0.189
Chain 1: 5400 -8460.266 0.247 0.189
Chain 1: 5500 -8402.292 0.235 0.189
Chain 1: 5600 -8699.926 0.238 0.189
Chain 1: 5700 -14628.287 0.260 0.301
Chain 1: 5800 -8943.466 0.320 0.376
Chain 1: 5900 -8871.215 0.320 0.376
Chain 1: 6000 -10305.605 0.297 0.301
Chain 1: 6100 -9319.440 0.247 0.139
Chain 1: 6200 -8298.501 0.246 0.139
Chain 1: 6300 -12471.990 0.250 0.139
Chain 1: 6400 -12337.227 0.180 0.123
Chain 1: 6500 -10739.496 0.195 0.139
Chain 1: 6600 -8515.895 0.217 0.149
Chain 1: 6700 -8437.250 0.178 0.139
Chain 1: 6800 -8148.385 0.118 0.123
Chain 1: 6900 -11587.776 0.147 0.139
Chain 1: 7000 -8765.565 0.165 0.149
Chain 1: 7100 -8073.529 0.163 0.149
Chain 1: 7200 -8205.390 0.152 0.149
Chain 1: 7300 -11861.503 0.149 0.149
Chain 1: 7400 -10604.533 0.160 0.149
Chain 1: 7500 -11768.675 0.155 0.119
Chain 1: 7600 -9134.742 0.158 0.119
Chain 1: 7700 -8256.115 0.168 0.119
Chain 1: 7800 -8275.478 0.164 0.119
Chain 1: 7900 -8648.039 0.139 0.106
Chain 1: 8000 -8139.422 0.113 0.099
Chain 1: 8100 -8360.722 0.107 0.099
Chain 1: 8200 -8521.718 0.107 0.099
Chain 1: 8300 -8158.111 0.081 0.062
Chain 1: 8400 -11011.295 0.095 0.062
Chain 1: 8500 -10972.164 0.086 0.045
Chain 1: 8600 -8756.380 0.082 0.045
Chain 1: 8700 -8145.296 0.079 0.045
Chain 1: 8800 -8207.314 0.079 0.045
Chain 1: 8900 -8826.939 0.082 0.062
Chain 1: 9000 -8649.394 0.078 0.045
Chain 1: 9100 -10749.723 0.095 0.070
Chain 1: 9200 -8690.030 0.117 0.075
Chain 1: 9300 -9776.873 0.123 0.111
Chain 1: 9400 -9507.288 0.100 0.075
Chain 1: 9500 -8885.447 0.107 0.075
Chain 1: 9600 -8640.411 0.084 0.070
Chain 1: 9700 -8223.459 0.082 0.070
Chain 1: 9800 -10237.314 0.101 0.070
Chain 1: 9900 -10868.059 0.100 0.070
Chain 1: 10000 -8070.750 0.132 0.111
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001399 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.99 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56857.627 1.000 1.000
Chain 1: 200 -17232.992 1.650 2.299
Chain 1: 300 -8665.018 1.429 1.000
Chain 1: 400 -8032.629 1.092 1.000
Chain 1: 500 -8034.815 0.873 0.989
Chain 1: 600 -7968.748 0.729 0.989
Chain 1: 700 -7756.577 0.629 0.079
Chain 1: 800 -8251.043 0.558 0.079
Chain 1: 900 -8011.713 0.499 0.060
Chain 1: 1000 -7902.611 0.451 0.060
Chain 1: 1100 -7729.836 0.353 0.030
Chain 1: 1200 -7651.128 0.124 0.027
Chain 1: 1300 -7776.939 0.027 0.022
Chain 1: 1400 -7696.011 0.020 0.016
Chain 1: 1500 -7640.780 0.021 0.016
Chain 1: 1600 -7571.579 0.021 0.016
Chain 1: 1700 -7530.868 0.018 0.014
Chain 1: 1800 -7600.089 0.013 0.011
Chain 1: 1900 -7618.079 0.011 0.010
Chain 1: 2000 -7615.263 0.009 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002536 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.36 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85890.699 1.000 1.000
Chain 1: 200 -13319.120 3.224 5.449
Chain 1: 300 -9772.573 2.271 1.000
Chain 1: 400 -10549.222 1.721 1.000
Chain 1: 500 -8708.968 1.419 0.363
Chain 1: 600 -8324.789 1.190 0.363
Chain 1: 700 -8425.519 1.022 0.211
Chain 1: 800 -8613.771 0.897 0.211
Chain 1: 900 -8609.662 0.797 0.074
Chain 1: 1000 -8419.926 0.720 0.074
Chain 1: 1100 -8599.125 0.622 0.046 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8279.740 0.081 0.039
Chain 1: 1300 -8510.810 0.047 0.027
Chain 1: 1400 -8509.259 0.040 0.023
Chain 1: 1500 -8399.073 0.020 0.022
Chain 1: 1600 -8494.170 0.017 0.021
Chain 1: 1700 -8584.381 0.017 0.021
Chain 1: 1800 -8194.976 0.019 0.021
Chain 1: 1900 -8297.436 0.020 0.021
Chain 1: 2000 -8267.337 0.019 0.013
Chain 1: 2100 -8398.199 0.018 0.013
Chain 1: 2200 -8184.381 0.017 0.013
Chain 1: 2300 -8326.504 0.016 0.013
Chain 1: 2400 -8339.292 0.016 0.013
Chain 1: 2500 -8307.180 0.015 0.012
Chain 1: 2600 -8307.391 0.014 0.012
Chain 1: 2700 -8215.359 0.014 0.012
Chain 1: 2800 -8190.934 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002713 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 27.13 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8392942.489 1.000 1.000
Chain 1: 200 -1579508.349 2.657 4.314
Chain 1: 300 -889327.154 2.030 1.000
Chain 1: 400 -456697.483 1.759 1.000
Chain 1: 500 -357333.652 1.463 0.947
Chain 1: 600 -232525.934 1.309 0.947
Chain 1: 700 -118938.255 1.258 0.947
Chain 1: 800 -86190.646 1.148 0.947
Chain 1: 900 -66552.438 1.054 0.776
Chain 1: 1000 -51354.912 0.978 0.776
Chain 1: 1100 -38842.380 0.910 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38017.326 0.481 0.380
Chain 1: 1300 -25984.778 0.449 0.380
Chain 1: 1400 -25703.311 0.356 0.322
Chain 1: 1500 -22293.654 0.343 0.322
Chain 1: 1600 -21511.030 0.293 0.296
Chain 1: 1700 -20386.044 0.203 0.295
Chain 1: 1800 -20330.405 0.166 0.153
Chain 1: 1900 -20656.136 0.138 0.055
Chain 1: 2000 -19169.039 0.116 0.055
Chain 1: 2100 -19407.177 0.085 0.036
Chain 1: 2200 -19633.301 0.084 0.036
Chain 1: 2300 -19250.940 0.040 0.020
Chain 1: 2400 -19023.222 0.040 0.020
Chain 1: 2500 -18825.304 0.025 0.016
Chain 1: 2600 -18455.907 0.024 0.016
Chain 1: 2700 -18413.056 0.018 0.012
Chain 1: 2800 -18130.171 0.020 0.016
Chain 1: 2900 -18411.165 0.020 0.015
Chain 1: 3000 -18397.361 0.012 0.012
Chain 1: 3100 -18482.284 0.011 0.012
Chain 1: 3200 -18173.277 0.012 0.015
Chain 1: 3300 -18377.787 0.011 0.012
Chain 1: 3400 -17853.304 0.013 0.015
Chain 1: 3500 -18464.302 0.015 0.016
Chain 1: 3600 -17772.133 0.017 0.016
Chain 1: 3700 -18158.080 0.019 0.017
Chain 1: 3800 -17119.611 0.023 0.021
Chain 1: 3900 -17115.832 0.022 0.021
Chain 1: 4000 -17233.096 0.022 0.021
Chain 1: 4100 -17146.951 0.022 0.021
Chain 1: 4200 -16963.640 0.022 0.021
Chain 1: 4300 -17101.727 0.021 0.021
Chain 1: 4400 -17058.877 0.019 0.011
Chain 1: 4500 -16961.488 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001301 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.01 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12201.584 1.000 1.000
Chain 1: 200 -9144.723 0.667 1.000
Chain 1: 300 -7906.211 0.497 0.334
Chain 1: 400 -8116.075 0.379 0.334
Chain 1: 500 -7988.586 0.307 0.157
Chain 1: 600 -7859.214 0.258 0.157
Chain 1: 700 -7770.397 0.223 0.026
Chain 1: 800 -7779.265 0.195 0.026
Chain 1: 900 -7700.412 0.175 0.016
Chain 1: 1000 -7877.832 0.159 0.023
Chain 1: 1100 -7906.217 0.060 0.016
Chain 1: 1200 -7802.195 0.028 0.016
Chain 1: 1300 -7748.374 0.013 0.013
Chain 1: 1400 -7765.785 0.010 0.011
Chain 1: 1500 -7852.371 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001443 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.43 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61653.658 1.000 1.000
Chain 1: 200 -17725.412 1.739 2.478
Chain 1: 300 -8749.680 1.501 1.026
Chain 1: 400 -8141.766 1.145 1.026
Chain 1: 500 -8265.602 0.919 1.000
Chain 1: 600 -8501.200 0.770 1.000
Chain 1: 700 -7698.113 0.675 0.104
Chain 1: 800 -8017.605 0.596 0.104
Chain 1: 900 -7794.445 0.533 0.075
Chain 1: 1000 -7658.166 0.481 0.075
Chain 1: 1100 -7667.785 0.381 0.040
Chain 1: 1200 -7489.804 0.136 0.029
Chain 1: 1300 -7709.114 0.036 0.028
Chain 1: 1400 -7787.172 0.030 0.028
Chain 1: 1500 -7487.476 0.032 0.028
Chain 1: 1600 -7573.499 0.031 0.028
Chain 1: 1700 -7459.568 0.022 0.024
Chain 1: 1800 -7473.532 0.018 0.018
Chain 1: 1900 -7502.495 0.015 0.015
Chain 1: 2000 -7509.329 0.014 0.011
Chain 1: 2100 -7503.354 0.014 0.011
Chain 1: 2200 -7612.311 0.013 0.011
Chain 1: 2300 -7504.502 0.011 0.011
Chain 1: 2400 -7565.880 0.011 0.011
Chain 1: 2500 -7508.082 0.008 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002514 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.14 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85993.937 1.000 1.000
Chain 1: 200 -13336.300 3.224 5.448
Chain 1: 300 -9722.441 2.273 1.000
Chain 1: 400 -10562.442 1.725 1.000
Chain 1: 500 -8689.151 1.423 0.372
Chain 1: 600 -8211.513 1.196 0.372
Chain 1: 700 -8269.812 1.026 0.216
Chain 1: 800 -8860.577 0.906 0.216
Chain 1: 900 -8571.858 0.809 0.080
Chain 1: 1000 -8321.752 0.731 0.080
Chain 1: 1100 -8536.017 0.634 0.067 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8279.220 0.092 0.058
Chain 1: 1300 -8435.777 0.057 0.034
Chain 1: 1400 -8440.110 0.049 0.031
Chain 1: 1500 -8298.282 0.029 0.030
Chain 1: 1600 -8411.220 0.024 0.025
Chain 1: 1700 -8497.772 0.025 0.025
Chain 1: 1800 -8091.824 0.023 0.025
Chain 1: 1900 -8188.493 0.021 0.019
Chain 1: 2000 -8160.584 0.018 0.017
Chain 1: 2100 -8281.072 0.017 0.015
Chain 1: 2200 -8091.227 0.016 0.015
Chain 1: 2300 -8228.142 0.016 0.015
Chain 1: 2400 -8235.151 0.016 0.015
Chain 1: 2500 -8201.797 0.015 0.013
Chain 1: 2600 -8199.781 0.014 0.012
Chain 1: 2700 -8113.781 0.014 0.012
Chain 1: 2800 -8079.009 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003228 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.28 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8397495.079 1.000 1.000
Chain 1: 200 -1585413.973 2.648 4.297
Chain 1: 300 -891681.689 2.025 1.000
Chain 1: 400 -457831.095 1.756 1.000
Chain 1: 500 -358104.317 1.460 0.948
Chain 1: 600 -233034.706 1.306 0.948
Chain 1: 700 -119157.405 1.256 0.948
Chain 1: 800 -86320.217 1.147 0.948
Chain 1: 900 -66644.994 1.052 0.778
Chain 1: 1000 -51431.399 0.976 0.778
Chain 1: 1100 -38894.800 0.909 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38071.111 0.481 0.380
Chain 1: 1300 -26020.472 0.450 0.380
Chain 1: 1400 -25738.252 0.356 0.322
Chain 1: 1500 -22323.455 0.343 0.322
Chain 1: 1600 -21538.984 0.293 0.296
Chain 1: 1700 -20412.271 0.203 0.295
Chain 1: 1800 -20356.308 0.166 0.153
Chain 1: 1900 -20682.235 0.138 0.055
Chain 1: 2000 -19193.524 0.116 0.055
Chain 1: 2100 -19431.882 0.085 0.036
Chain 1: 2200 -19658.155 0.084 0.036
Chain 1: 2300 -19275.626 0.040 0.020
Chain 1: 2400 -19047.798 0.040 0.020
Chain 1: 2500 -18849.762 0.025 0.016
Chain 1: 2600 -18480.115 0.024 0.016
Chain 1: 2700 -18437.211 0.018 0.012
Chain 1: 2800 -18154.051 0.020 0.016
Chain 1: 2900 -18435.317 0.020 0.015
Chain 1: 3000 -18421.535 0.012 0.012
Chain 1: 3100 -18506.439 0.011 0.012
Chain 1: 3200 -18197.249 0.012 0.015
Chain 1: 3300 -18401.919 0.011 0.012
Chain 1: 3400 -17877.001 0.013 0.015
Chain 1: 3500 -18488.544 0.015 0.016
Chain 1: 3600 -17795.786 0.017 0.016
Chain 1: 3700 -18182.130 0.019 0.017
Chain 1: 3800 -17142.582 0.023 0.021
Chain 1: 3900 -17138.765 0.022 0.021
Chain 1: 4000 -17256.085 0.022 0.021
Chain 1: 4100 -17169.795 0.022 0.021
Chain 1: 4200 -16986.292 0.022 0.021
Chain 1: 4300 -17124.521 0.021 0.021
Chain 1: 4400 -17081.482 0.019 0.011
Chain 1: 4500 -16984.077 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001625 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 16.25 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48251.694 1.000 1.000
Chain 1: 200 -15225.281 1.585 2.169
Chain 1: 300 -12013.418 1.146 1.000
Chain 1: 400 -21161.159 0.967 1.000
Chain 1: 500 -12171.036 0.921 0.739
Chain 1: 600 -20598.828 0.836 0.739
Chain 1: 700 -14340.083 0.779 0.436
Chain 1: 800 -11464.046 0.713 0.436
Chain 1: 900 -23660.172 0.691 0.436
Chain 1: 1000 -24487.165 0.625 0.436
Chain 1: 1100 -10499.720 0.659 0.436 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -12221.519 0.456 0.432
Chain 1: 1300 -11608.716 0.434 0.432
Chain 1: 1400 -10624.763 0.400 0.409
Chain 1: 1500 -12696.245 0.343 0.251
Chain 1: 1600 -10400.637 0.324 0.221
Chain 1: 1700 -10929.998 0.285 0.163
Chain 1: 1800 -9821.027 0.271 0.141
Chain 1: 1900 -9576.513 0.222 0.113
Chain 1: 2000 -11014.196 0.232 0.131
Chain 1: 2100 -9338.089 0.117 0.131
Chain 1: 2200 -9686.718 0.106 0.113
Chain 1: 2300 -9740.718 0.101 0.113
Chain 1: 2400 -9737.051 0.092 0.113
Chain 1: 2500 -11185.349 0.089 0.113
Chain 1: 2600 -9031.387 0.091 0.113
Chain 1: 2700 -11906.527 0.110 0.129
Chain 1: 2800 -9132.456 0.129 0.131
Chain 1: 2900 -9094.799 0.127 0.131
Chain 1: 3000 -16178.272 0.158 0.179
Chain 1: 3100 -9561.755 0.209 0.238
Chain 1: 3200 -8622.161 0.216 0.238
Chain 1: 3300 -9147.606 0.221 0.238
Chain 1: 3400 -12471.453 0.248 0.241
Chain 1: 3500 -8809.815 0.277 0.267
Chain 1: 3600 -10214.718 0.267 0.267
Chain 1: 3700 -9056.150 0.255 0.267
Chain 1: 3800 -12348.921 0.251 0.267
Chain 1: 3900 -8832.506 0.291 0.267
Chain 1: 4000 -8705.197 0.249 0.267
Chain 1: 4100 -9172.084 0.184 0.138
Chain 1: 4200 -11347.735 0.193 0.192
Chain 1: 4300 -15075.649 0.212 0.247
Chain 1: 4400 -12319.583 0.207 0.224
Chain 1: 4500 -8402.402 0.212 0.224
Chain 1: 4600 -12862.716 0.233 0.247
Chain 1: 4700 -11469.076 0.233 0.247
Chain 1: 4800 -8302.522 0.244 0.247
Chain 1: 4900 -8991.588 0.212 0.224
Chain 1: 5000 -9002.454 0.211 0.224
Chain 1: 5100 -11169.800 0.225 0.224
Chain 1: 5200 -8566.521 0.236 0.247
Chain 1: 5300 -13207.418 0.247 0.304
Chain 1: 5400 -8767.420 0.275 0.347
Chain 1: 5500 -8720.051 0.229 0.304
Chain 1: 5600 -15309.780 0.237 0.304
Chain 1: 5700 -8940.067 0.296 0.351
Chain 1: 5800 -12761.737 0.288 0.304
Chain 1: 5900 -8248.275 0.335 0.351
Chain 1: 6000 -8548.959 0.339 0.351
Chain 1: 6100 -8655.214 0.320 0.351
Chain 1: 6200 -7905.218 0.300 0.351
Chain 1: 6300 -9082.650 0.277 0.299
Chain 1: 6400 -8690.009 0.231 0.130
Chain 1: 6500 -9219.357 0.236 0.130
Chain 1: 6600 -10242.267 0.203 0.100
Chain 1: 6700 -8054.975 0.159 0.100
Chain 1: 6800 -8853.513 0.138 0.095
Chain 1: 6900 -12832.545 0.115 0.095
Chain 1: 7000 -8287.952 0.166 0.100
Chain 1: 7100 -8236.159 0.165 0.100
Chain 1: 7200 -10264.897 0.176 0.130
Chain 1: 7300 -10935.203 0.169 0.100
Chain 1: 7400 -8373.778 0.195 0.198
Chain 1: 7500 -7996.986 0.194 0.198
Chain 1: 7600 -11231.447 0.213 0.272
Chain 1: 7700 -8222.862 0.222 0.288
Chain 1: 7800 -7930.640 0.217 0.288
Chain 1: 7900 -7962.410 0.186 0.198
Chain 1: 8000 -10607.997 0.156 0.198
Chain 1: 8100 -8134.342 0.186 0.249
Chain 1: 8200 -7975.224 0.168 0.249
Chain 1: 8300 -8116.493 0.164 0.249
Chain 1: 8400 -8223.522 0.135 0.047
Chain 1: 8500 -9888.857 0.147 0.168
Chain 1: 8600 -8170.675 0.139 0.168
Chain 1: 8700 -7916.530 0.106 0.037
Chain 1: 8800 -10001.067 0.123 0.168
Chain 1: 8900 -8190.617 0.144 0.208
Chain 1: 9000 -10067.043 0.138 0.186
Chain 1: 9100 -8149.148 0.131 0.186
Chain 1: 9200 -7973.615 0.131 0.186
Chain 1: 9300 -9038.628 0.141 0.186
Chain 1: 9400 -11393.283 0.161 0.207
Chain 1: 9500 -8975.225 0.171 0.208
Chain 1: 9600 -9641.952 0.157 0.207
Chain 1: 9700 -8672.969 0.165 0.207
Chain 1: 9800 -10348.932 0.160 0.186
Chain 1: 9900 -8479.205 0.160 0.186
Chain 1: 10000 -8468.757 0.142 0.162
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001736 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 17.36 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56335.317 1.000 1.000
Chain 1: 200 -16981.199 1.659 2.318
Chain 1: 300 -8505.273 1.438 1.000
Chain 1: 400 -8862.564 1.089 1.000
Chain 1: 500 -8472.629 0.880 0.997
Chain 1: 600 -8998.724 0.743 0.997
Chain 1: 700 -7663.478 0.662 0.174
Chain 1: 800 -7962.381 0.584 0.174
Chain 1: 900 -7812.624 0.521 0.058
Chain 1: 1000 -7666.070 0.471 0.058
Chain 1: 1100 -7555.247 0.372 0.046
Chain 1: 1200 -7541.092 0.141 0.040
Chain 1: 1300 -7639.764 0.042 0.038
Chain 1: 1400 -7713.238 0.039 0.019
Chain 1: 1500 -7543.689 0.037 0.019
Chain 1: 1600 -7477.571 0.032 0.019
Chain 1: 1700 -7427.306 0.015 0.015
Chain 1: 1800 -7474.428 0.012 0.013
Chain 1: 1900 -7532.851 0.011 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003518 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.18 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85733.576 1.000 1.000
Chain 1: 200 -13073.254 3.279 5.558
Chain 1: 300 -9515.692 2.311 1.000
Chain 1: 400 -10253.156 1.751 1.000
Chain 1: 500 -8425.976 1.444 0.374
Chain 1: 600 -8183.425 1.208 0.374
Chain 1: 700 -8257.948 1.037 0.217
Chain 1: 800 -8519.843 0.911 0.217
Chain 1: 900 -8397.748 0.812 0.072
Chain 1: 1000 -8137.696 0.734 0.072
Chain 1: 1100 -8383.740 0.637 0.032 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8125.426 0.084 0.032
Chain 1: 1300 -8239.147 0.048 0.031
Chain 1: 1400 -8260.094 0.041 0.030
Chain 1: 1500 -8150.243 0.021 0.029
Chain 1: 1600 -8248.211 0.019 0.015
Chain 1: 1700 -8336.620 0.019 0.015
Chain 1: 1800 -7948.745 0.021 0.015
Chain 1: 1900 -8051.335 0.021 0.014
Chain 1: 2000 -8021.093 0.018 0.013
Chain 1: 2100 -8152.949 0.017 0.013
Chain 1: 2200 -7938.795 0.016 0.013
Chain 1: 2300 -8080.527 0.016 0.013
Chain 1: 2400 -8092.821 0.016 0.013
Chain 1: 2500 -8060.894 0.015 0.013
Chain 1: 2600 -8060.602 0.014 0.013
Chain 1: 2700 -7968.874 0.014 0.013
Chain 1: 2800 -7944.942 0.010 0.012 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003116 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.16 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8372009.864 1.000 1.000
Chain 1: 200 -1580827.554 2.648 4.296
Chain 1: 300 -890628.001 2.024 1.000
Chain 1: 400 -457253.579 1.755 1.000
Chain 1: 500 -357680.829 1.459 0.948
Chain 1: 600 -232858.256 1.306 0.948
Chain 1: 700 -118991.684 1.256 0.948
Chain 1: 800 -86109.228 1.146 0.948
Chain 1: 900 -66434.181 1.052 0.775
Chain 1: 1000 -51197.918 0.977 0.775
Chain 1: 1100 -38642.448 0.909 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37814.069 0.482 0.382
Chain 1: 1300 -25751.936 0.451 0.382
Chain 1: 1400 -25467.195 0.357 0.325
Chain 1: 1500 -22048.043 0.345 0.325
Chain 1: 1600 -21261.557 0.295 0.298
Chain 1: 1700 -20134.014 0.205 0.296
Chain 1: 1800 -20077.463 0.167 0.155
Chain 1: 1900 -20403.067 0.139 0.056
Chain 1: 2000 -18914.026 0.117 0.056
Chain 1: 2100 -19152.682 0.086 0.037
Chain 1: 2200 -19378.671 0.085 0.037
Chain 1: 2300 -18996.372 0.040 0.020
Chain 1: 2400 -18768.619 0.040 0.020
Chain 1: 2500 -18570.511 0.026 0.016
Chain 1: 2600 -18201.409 0.024 0.016
Chain 1: 2700 -18158.583 0.019 0.012
Chain 1: 2800 -17875.566 0.020 0.016
Chain 1: 2900 -18156.634 0.020 0.015
Chain 1: 3000 -18142.980 0.012 0.012
Chain 1: 3100 -18227.813 0.011 0.012
Chain 1: 3200 -17918.906 0.012 0.015
Chain 1: 3300 -18123.308 0.011 0.012
Chain 1: 3400 -17598.858 0.013 0.015
Chain 1: 3500 -18209.782 0.015 0.016
Chain 1: 3600 -17517.794 0.017 0.016
Chain 1: 3700 -17903.586 0.019 0.017
Chain 1: 3800 -16865.281 0.024 0.022
Chain 1: 3900 -16861.462 0.022 0.022
Chain 1: 4000 -16978.789 0.023 0.022
Chain 1: 4100 -16892.586 0.023 0.022
Chain 1: 4200 -16709.306 0.022 0.022
Chain 1: 4300 -16847.408 0.022 0.022
Chain 1: 4400 -16804.619 0.019 0.011
Chain 1: 4500 -16707.185 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001533 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.33 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12420.474 1.000 1.000
Chain 1: 200 -9345.352 0.665 1.000
Chain 1: 300 -8084.135 0.495 0.329
Chain 1: 400 -8180.746 0.374 0.329
Chain 1: 500 -8119.937 0.301 0.156
Chain 1: 600 -7979.484 0.254 0.156
Chain 1: 700 -7895.722 0.219 0.018
Chain 1: 800 -7906.887 0.192 0.018
Chain 1: 900 -7859.963 0.171 0.012
Chain 1: 1000 -7962.333 0.155 0.013
Chain 1: 1100 -8040.648 0.056 0.012
Chain 1: 1200 -7905.574 0.025 0.012
Chain 1: 1300 -7863.728 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001742 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 17.42 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58187.918 1.000 1.000
Chain 1: 200 -17723.605 1.642 2.283
Chain 1: 300 -8704.302 1.440 1.036
Chain 1: 400 -8217.457 1.095 1.036
Chain 1: 500 -8111.295 0.878 1.000
Chain 1: 600 -8660.023 0.742 1.000
Chain 1: 700 -7962.457 0.649 0.088
Chain 1: 800 -8085.671 0.570 0.088
Chain 1: 900 -7920.630 0.509 0.063
Chain 1: 1000 -8132.212 0.460 0.063
Chain 1: 1100 -7719.200 0.366 0.059
Chain 1: 1200 -7600.615 0.139 0.054
Chain 1: 1300 -7787.307 0.038 0.026
Chain 1: 1400 -7911.774 0.033 0.024
Chain 1: 1500 -7629.481 0.036 0.026
Chain 1: 1600 -7703.983 0.031 0.024
Chain 1: 1700 -7560.456 0.024 0.021
Chain 1: 1800 -7579.899 0.022 0.021
Chain 1: 1900 -7611.791 0.021 0.019
Chain 1: 2000 -7652.836 0.019 0.016
Chain 1: 2100 -7644.196 0.013 0.016
Chain 1: 2200 -7707.105 0.013 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004597 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 45.97 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85461.493 1.000 1.000
Chain 1: 200 -13499.400 3.165 5.331
Chain 1: 300 -9874.976 2.233 1.000
Chain 1: 400 -10695.604 1.694 1.000
Chain 1: 500 -8859.692 1.396 0.367
Chain 1: 600 -8459.931 1.171 0.367
Chain 1: 700 -8426.801 1.005 0.207
Chain 1: 800 -8954.799 0.886 0.207
Chain 1: 900 -8652.822 0.792 0.077
Chain 1: 1000 -8488.595 0.715 0.077
Chain 1: 1100 -8725.755 0.617 0.059 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8240.970 0.090 0.059
Chain 1: 1300 -8571.710 0.057 0.047
Chain 1: 1400 -8571.433 0.050 0.039
Chain 1: 1500 -8445.002 0.030 0.035
Chain 1: 1600 -8552.462 0.027 0.027
Chain 1: 1700 -8635.835 0.028 0.027
Chain 1: 1800 -8222.007 0.027 0.027
Chain 1: 1900 -8318.286 0.024 0.019
Chain 1: 2000 -8291.568 0.023 0.015
Chain 1: 2100 -8414.337 0.021 0.015
Chain 1: 2200 -8234.399 0.018 0.015
Chain 1: 2300 -8313.085 0.015 0.013
Chain 1: 2400 -8382.803 0.016 0.013
Chain 1: 2500 -8328.284 0.015 0.012
Chain 1: 2600 -8327.805 0.014 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002607 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 26.07 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8388868.194 1.000 1.000
Chain 1: 200 -1582636.244 2.650 4.301
Chain 1: 300 -890937.536 2.026 1.000
Chain 1: 400 -457702.155 1.756 1.000
Chain 1: 500 -358203.460 1.460 0.947
Chain 1: 600 -233290.950 1.306 0.947
Chain 1: 700 -119390.578 1.256 0.947
Chain 1: 800 -86553.576 1.146 0.947
Chain 1: 900 -66873.305 1.052 0.776
Chain 1: 1000 -51648.857 0.976 0.776
Chain 1: 1100 -39103.550 0.908 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38278.081 0.480 0.379
Chain 1: 1300 -26215.913 0.448 0.379
Chain 1: 1400 -25932.967 0.355 0.321
Chain 1: 1500 -22515.165 0.342 0.321
Chain 1: 1600 -21729.925 0.292 0.295
Chain 1: 1700 -20601.681 0.202 0.294
Chain 1: 1800 -20545.373 0.165 0.152
Chain 1: 1900 -20871.447 0.137 0.055
Chain 1: 2000 -19381.498 0.115 0.055
Chain 1: 2100 -19619.993 0.084 0.036
Chain 1: 2200 -19846.563 0.083 0.036
Chain 1: 2300 -19463.692 0.039 0.020
Chain 1: 2400 -19235.786 0.039 0.020
Chain 1: 2500 -19037.809 0.025 0.016
Chain 1: 2600 -18668.049 0.024 0.016
Chain 1: 2700 -18624.987 0.018 0.012
Chain 1: 2800 -18341.868 0.020 0.015
Chain 1: 2900 -18623.140 0.020 0.015
Chain 1: 3000 -18609.352 0.012 0.012
Chain 1: 3100 -18694.331 0.011 0.012
Chain 1: 3200 -18385.014 0.012 0.015
Chain 1: 3300 -18589.726 0.011 0.012
Chain 1: 3400 -18064.678 0.013 0.015
Chain 1: 3500 -18676.500 0.015 0.015
Chain 1: 3600 -17983.284 0.017 0.015
Chain 1: 3700 -18370.038 0.019 0.017
Chain 1: 3800 -17329.866 0.023 0.021
Chain 1: 3900 -17326.018 0.021 0.021
Chain 1: 4000 -17443.334 0.022 0.021
Chain 1: 4100 -17357.090 0.022 0.021
Chain 1: 4200 -17173.369 0.022 0.021
Chain 1: 4300 -17311.746 0.021 0.021
Chain 1: 4400 -17268.602 0.019 0.011
Chain 1: 4500 -17171.147 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001591 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.91 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13802.217 1.000 1.000
Chain 1: 200 -10258.953 0.673 1.000
Chain 1: 300 -8857.284 0.501 0.345
Chain 1: 400 -8623.238 0.383 0.345
Chain 1: 500 -8362.705 0.312 0.158
Chain 1: 600 -8393.547 0.261 0.158
Chain 1: 700 -8237.084 0.226 0.031
Chain 1: 800 -8241.079 0.198 0.031
Chain 1: 900 -8346.724 0.178 0.027
Chain 1: 1000 -8326.008 0.160 0.027
Chain 1: 1100 -8398.576 0.061 0.019
Chain 1: 1200 -8283.183 0.028 0.014
Chain 1: 1300 -8195.842 0.013 0.013
Chain 1: 1400 -8226.557 0.011 0.011
Chain 1: 1500 -8324.629 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001807 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 18.07 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -59113.655 1.000 1.000
Chain 1: 200 -18575.603 1.591 2.182
Chain 1: 300 -9084.113 1.409 1.045
Chain 1: 400 -8067.701 1.088 1.045
Chain 1: 500 -8715.471 0.885 1.000
Chain 1: 600 -9905.416 0.758 1.000
Chain 1: 700 -7893.570 0.686 0.255
Chain 1: 800 -8398.901 0.608 0.255
Chain 1: 900 -7756.788 0.549 0.126
Chain 1: 1000 -7923.002 0.497 0.126
Chain 1: 1100 -7759.897 0.399 0.120
Chain 1: 1200 -7660.376 0.182 0.083
Chain 1: 1300 -7722.315 0.078 0.074
Chain 1: 1400 -7642.098 0.067 0.060
Chain 1: 1500 -7564.492 0.060 0.021
Chain 1: 1600 -7666.475 0.049 0.021
Chain 1: 1700 -7695.309 0.024 0.013
Chain 1: 1800 -7562.314 0.020 0.013
Chain 1: 1900 -7763.592 0.014 0.013
Chain 1: 2000 -7752.272 0.012 0.013
Chain 1: 2100 -7644.035 0.012 0.013
Chain 1: 2200 -7760.552 0.012 0.013
Chain 1: 2300 -7611.050 0.013 0.014
Chain 1: 2400 -7762.383 0.014 0.015
Chain 1: 2500 -7675.044 0.014 0.015
Chain 1: 2600 -7536.359 0.015 0.018
Chain 1: 2700 -7530.883 0.014 0.018
Chain 1: 2800 -7638.584 0.014 0.015
Chain 1: 2900 -7394.157 0.015 0.015
Chain 1: 3000 -7527.878 0.016 0.018
Chain 1: 3100 -7528.338 0.015 0.018
Chain 1: 3200 -7654.109 0.015 0.018
Chain 1: 3300 -7397.177 0.017 0.018
Chain 1: 3400 -7687.608 0.018 0.018
Chain 1: 3500 -7441.352 0.021 0.018
Chain 1: 3600 -7509.711 0.020 0.018
Chain 1: 3700 -7465.300 0.020 0.018
Chain 1: 3800 -7436.258 0.019 0.018
Chain 1: 3900 -7463.834 0.016 0.016
Chain 1: 4000 -7396.549 0.015 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003341 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.41 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87494.985 1.000 1.000
Chain 1: 200 -14283.041 3.063 5.126
Chain 1: 300 -10482.976 2.163 1.000
Chain 1: 400 -12421.733 1.661 1.000
Chain 1: 500 -8882.958 1.409 0.398
Chain 1: 600 -9380.399 1.183 0.398
Chain 1: 700 -8957.531 1.020 0.362
Chain 1: 800 -8815.825 0.895 0.362
Chain 1: 900 -8757.142 0.796 0.156
Chain 1: 1000 -9143.030 0.721 0.156
Chain 1: 1100 -9219.845 0.622 0.053 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8697.951 0.115 0.053
Chain 1: 1300 -9061.070 0.083 0.047
Chain 1: 1400 -9008.564 0.068 0.042
Chain 1: 1500 -8912.353 0.029 0.040
Chain 1: 1600 -8976.652 0.024 0.016
Chain 1: 1700 -9052.343 0.021 0.011
Chain 1: 1800 -8607.455 0.024 0.011
Chain 1: 1900 -8703.170 0.025 0.011
Chain 1: 2000 -8725.109 0.021 0.011
Chain 1: 2100 -8815.339 0.021 0.011
Chain 1: 2200 -8585.025 0.017 0.011
Chain 1: 2300 -8783.286 0.016 0.011
Chain 1: 2400 -8606.762 0.017 0.011
Chain 1: 2500 -8674.479 0.017 0.011
Chain 1: 2600 -8582.717 0.017 0.011
Chain 1: 2700 -8617.150 0.017 0.011
Chain 1: 2800 -8570.055 0.012 0.011
Chain 1: 2900 -8683.291 0.012 0.011
Chain 1: 3000 -8591.666 0.013 0.011
Chain 1: 3100 -8559.606 0.013 0.011
Chain 1: 3200 -8529.682 0.010 0.011
Chain 1: 3300 -8797.625 0.011 0.011
Chain 1: 3400 -8847.929 0.010 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.005061 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 50.61 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8444102.022 1.000 1.000
Chain 1: 200 -1588735.607 2.657 4.315
Chain 1: 300 -890316.692 2.033 1.000
Chain 1: 400 -458242.477 1.761 1.000
Chain 1: 500 -357925.686 1.465 0.943
Chain 1: 600 -233052.888 1.310 0.943
Chain 1: 700 -119603.129 1.258 0.943
Chain 1: 800 -86941.018 1.148 0.943
Chain 1: 900 -67361.904 1.053 0.784
Chain 1: 1000 -52247.408 0.976 0.784
Chain 1: 1100 -39799.115 0.908 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38991.272 0.478 0.376
Chain 1: 1300 -27001.465 0.444 0.376
Chain 1: 1400 -26730.740 0.351 0.313
Chain 1: 1500 -23331.521 0.337 0.313
Chain 1: 1600 -22553.939 0.287 0.291
Chain 1: 1700 -21432.410 0.198 0.289
Chain 1: 1800 -21378.515 0.160 0.146
Chain 1: 1900 -21705.640 0.133 0.052
Chain 1: 2000 -20217.692 0.111 0.052
Chain 1: 2100 -20456.032 0.081 0.034
Chain 1: 2200 -20682.776 0.080 0.034
Chain 1: 2300 -20299.477 0.038 0.019
Chain 1: 2400 -20071.226 0.038 0.019
Chain 1: 2500 -19873.113 0.024 0.015
Chain 1: 2600 -19502.420 0.023 0.015
Chain 1: 2700 -19459.203 0.018 0.012
Chain 1: 2800 -19175.561 0.019 0.015
Chain 1: 2900 -19457.151 0.019 0.014
Chain 1: 3000 -19443.330 0.011 0.012
Chain 1: 3100 -19528.470 0.011 0.011
Chain 1: 3200 -19218.574 0.011 0.014
Chain 1: 3300 -19423.750 0.010 0.011
Chain 1: 3400 -18897.559 0.012 0.014
Chain 1: 3500 -19511.087 0.014 0.015
Chain 1: 3600 -18815.532 0.016 0.015
Chain 1: 3700 -19203.915 0.018 0.016
Chain 1: 3800 -18160.211 0.022 0.020
Chain 1: 3900 -18156.229 0.021 0.020
Chain 1: 4000 -18273.576 0.021 0.020
Chain 1: 4100 -18187.161 0.021 0.020
Chain 1: 4200 -18002.648 0.021 0.020
Chain 1: 4300 -18141.601 0.020 0.020
Chain 1: 4400 -18097.781 0.018 0.010
Chain 1: 4500 -18000.178 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001327 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.27 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12611.031 1.000 1.000
Chain 1: 200 -9525.943 0.662 1.000
Chain 1: 300 -8179.427 0.496 0.324
Chain 1: 400 -8362.527 0.378 0.324
Chain 1: 500 -8341.939 0.303 0.165
Chain 1: 600 -8133.286 0.256 0.165
Chain 1: 700 -8058.971 0.221 0.026
Chain 1: 800 -8071.491 0.194 0.026
Chain 1: 900 -8005.475 0.173 0.022
Chain 1: 1000 -8062.569 0.156 0.022
Chain 1: 1100 -8137.297 0.057 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001391 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.91 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58618.344 1.000 1.000
Chain 1: 200 -17944.109 1.633 2.267
Chain 1: 300 -8819.764 1.434 1.035
Chain 1: 400 -8233.967 1.093 1.035
Chain 1: 500 -8521.621 0.881 1.000
Chain 1: 600 -8789.167 0.739 1.000
Chain 1: 700 -8043.439 0.647 0.093
Chain 1: 800 -8442.150 0.572 0.093
Chain 1: 900 -7732.370 0.519 0.092
Chain 1: 1000 -7731.540 0.467 0.092
Chain 1: 1100 -7916.957 0.369 0.071
Chain 1: 1200 -7938.175 0.143 0.047
Chain 1: 1300 -7827.388 0.041 0.034
Chain 1: 1400 -7733.095 0.035 0.030
Chain 1: 1500 -7605.072 0.033 0.023
Chain 1: 1600 -7793.068 0.033 0.023
Chain 1: 1700 -7576.493 0.026 0.023
Chain 1: 1800 -7683.683 0.023 0.017
Chain 1: 1900 -7718.268 0.014 0.014
Chain 1: 2000 -7711.612 0.014 0.014
Chain 1: 2100 -7666.384 0.012 0.014
Chain 1: 2200 -7769.670 0.013 0.014
Chain 1: 2300 -7609.804 0.014 0.014
Chain 1: 2400 -7723.571 0.014 0.015
Chain 1: 2500 -7696.812 0.013 0.014
Chain 1: 2600 -7585.265 0.012 0.014
Chain 1: 2700 -7596.341 0.009 0.013 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003941 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 39.41 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86502.645 1.000 1.000
Chain 1: 200 -13714.557 3.154 5.307
Chain 1: 300 -10053.118 2.224 1.000
Chain 1: 400 -11105.381 1.692 1.000
Chain 1: 500 -9034.375 1.399 0.364
Chain 1: 600 -8512.556 1.176 0.364
Chain 1: 700 -8476.107 1.009 0.229
Chain 1: 800 -9313.846 0.894 0.229
Chain 1: 900 -8914.088 0.800 0.095
Chain 1: 1000 -8572.393 0.724 0.095
Chain 1: 1100 -8871.654 0.627 0.090 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8431.415 0.101 0.061
Chain 1: 1300 -8767.839 0.069 0.052
Chain 1: 1400 -8740.867 0.060 0.045
Chain 1: 1500 -8587.765 0.039 0.040
Chain 1: 1600 -8703.788 0.034 0.038
Chain 1: 1700 -8779.644 0.034 0.038
Chain 1: 1800 -8353.043 0.030 0.038
Chain 1: 1900 -8455.610 0.027 0.034
Chain 1: 2000 -8430.391 0.023 0.018
Chain 1: 2100 -8557.032 0.021 0.015
Chain 1: 2200 -8356.639 0.019 0.015
Chain 1: 2300 -8450.704 0.016 0.013
Chain 1: 2400 -8518.849 0.016 0.013
Chain 1: 2500 -8465.094 0.015 0.012
Chain 1: 2600 -8467.300 0.014 0.011
Chain 1: 2700 -8383.619 0.014 0.011
Chain 1: 2800 -8342.427 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003571 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.71 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8419173.351 1.000 1.000
Chain 1: 200 -1587284.288 2.652 4.304
Chain 1: 300 -889890.127 2.029 1.000
Chain 1: 400 -457241.353 1.759 1.000
Chain 1: 500 -357357.266 1.463 0.946
Chain 1: 600 -232579.286 1.308 0.946
Chain 1: 700 -119116.303 1.258 0.946
Chain 1: 800 -86422.881 1.148 0.946
Chain 1: 900 -66832.927 1.053 0.784
Chain 1: 1000 -51686.735 0.977 0.784
Chain 1: 1100 -39210.569 0.909 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38394.312 0.480 0.378
Chain 1: 1300 -26393.842 0.447 0.378
Chain 1: 1400 -26117.935 0.354 0.318
Chain 1: 1500 -22716.701 0.341 0.318
Chain 1: 1600 -21937.043 0.291 0.293
Chain 1: 1700 -20815.514 0.201 0.293
Chain 1: 1800 -20760.989 0.163 0.150
Chain 1: 1900 -21087.300 0.135 0.054
Chain 1: 2000 -19600.632 0.114 0.054
Chain 1: 2100 -19838.958 0.083 0.036
Chain 1: 2200 -20065.166 0.082 0.036
Chain 1: 2300 -19682.531 0.039 0.019
Chain 1: 2400 -19454.595 0.039 0.019
Chain 1: 2500 -19256.489 0.025 0.015
Chain 1: 2600 -18886.734 0.023 0.015
Chain 1: 2700 -18843.652 0.018 0.012
Chain 1: 2800 -18560.408 0.019 0.015
Chain 1: 2900 -18841.647 0.019 0.015
Chain 1: 3000 -18827.844 0.012 0.012
Chain 1: 3100 -18912.906 0.011 0.012
Chain 1: 3200 -18603.504 0.012 0.015
Chain 1: 3300 -18808.261 0.011 0.012
Chain 1: 3400 -18283.065 0.012 0.015
Chain 1: 3500 -18895.110 0.015 0.015
Chain 1: 3600 -18201.462 0.016 0.015
Chain 1: 3700 -18588.507 0.018 0.017
Chain 1: 3800 -17547.755 0.023 0.021
Chain 1: 3900 -17543.843 0.021 0.021
Chain 1: 4000 -17661.174 0.022 0.021
Chain 1: 4100 -17574.947 0.022 0.021
Chain 1: 4200 -17391.039 0.021 0.021
Chain 1: 4300 -17529.556 0.021 0.021
Chain 1: 4400 -17486.295 0.018 0.011
Chain 1: 4500 -17388.768 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001329 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.29 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49820.029 1.000 1.000
Chain 1: 200 -13619.948 1.829 2.658
Chain 1: 300 -18441.748 1.306 1.000
Chain 1: 400 -15895.639 1.020 1.000
Chain 1: 500 -13314.517 0.855 0.261
Chain 1: 600 -15949.330 0.740 0.261
Chain 1: 700 -15127.167 0.642 0.194
Chain 1: 800 -14218.342 0.570 0.194
Chain 1: 900 -13101.453 0.516 0.165
Chain 1: 1000 -11843.583 0.475 0.165
Chain 1: 1100 -11159.835 0.381 0.160
Chain 1: 1200 -21730.248 0.164 0.160
Chain 1: 1300 -17811.967 0.160 0.160
Chain 1: 1400 -12966.282 0.181 0.165
Chain 1: 1500 -10878.355 0.181 0.165
Chain 1: 1600 -10406.459 0.169 0.106
Chain 1: 1700 -11459.751 0.173 0.106
Chain 1: 1800 -12518.540 0.175 0.106
Chain 1: 1900 -10603.824 0.184 0.181
Chain 1: 2000 -12121.518 0.186 0.181
Chain 1: 2100 -11008.442 0.190 0.181
Chain 1: 2200 -11387.601 0.145 0.125
Chain 1: 2300 -10212.763 0.134 0.115
Chain 1: 2400 -10475.220 0.099 0.101
Chain 1: 2500 -10122.665 0.084 0.092
Chain 1: 2600 -10590.896 0.084 0.092
Chain 1: 2700 -11900.352 0.085 0.101
Chain 1: 2800 -18476.650 0.113 0.110
Chain 1: 2900 -15520.892 0.114 0.110
Chain 1: 3000 -10239.003 0.153 0.110
Chain 1: 3100 -10273.221 0.143 0.110
Chain 1: 3200 -9757.823 0.145 0.110
Chain 1: 3300 -10398.793 0.139 0.062
Chain 1: 3400 -9887.085 0.142 0.062
Chain 1: 3500 -15258.229 0.174 0.110
Chain 1: 3600 -11719.260 0.200 0.190
Chain 1: 3700 -14138.599 0.206 0.190
Chain 1: 3800 -9740.392 0.215 0.190
Chain 1: 3900 -9891.375 0.198 0.171
Chain 1: 4000 -11480.978 0.160 0.138
Chain 1: 4100 -10380.060 0.170 0.138
Chain 1: 4200 -16988.722 0.204 0.171
Chain 1: 4300 -9604.189 0.275 0.302
Chain 1: 4400 -14320.409 0.302 0.329
Chain 1: 4500 -11634.692 0.290 0.302
Chain 1: 4600 -9349.583 0.284 0.244
Chain 1: 4700 -10789.403 0.281 0.244
Chain 1: 4800 -9427.037 0.250 0.231
Chain 1: 4900 -9333.051 0.250 0.231
Chain 1: 5000 -10493.774 0.247 0.231
Chain 1: 5100 -10439.722 0.237 0.231
Chain 1: 5200 -8961.806 0.214 0.165
Chain 1: 5300 -10920.126 0.155 0.165
Chain 1: 5400 -9116.556 0.142 0.165
Chain 1: 5500 -14851.561 0.158 0.165
Chain 1: 5600 -14586.271 0.135 0.145
Chain 1: 5700 -10039.615 0.167 0.165
Chain 1: 5800 -10625.231 0.158 0.165
Chain 1: 5900 -9266.728 0.172 0.165
Chain 1: 6000 -14972.883 0.199 0.179
Chain 1: 6100 -14213.172 0.204 0.179
Chain 1: 6200 -8933.529 0.246 0.198
Chain 1: 6300 -10100.481 0.240 0.198
Chain 1: 6400 -8889.892 0.234 0.147
Chain 1: 6500 -9795.557 0.204 0.136
Chain 1: 6600 -12212.816 0.222 0.147
Chain 1: 6700 -13270.666 0.185 0.136
Chain 1: 6800 -11476.340 0.195 0.147
Chain 1: 6900 -9291.680 0.204 0.156
Chain 1: 7000 -14433.420 0.201 0.156
Chain 1: 7100 -9607.194 0.246 0.198
Chain 1: 7200 -9092.189 0.193 0.156
Chain 1: 7300 -9555.047 0.186 0.156
Chain 1: 7400 -9528.577 0.173 0.156
Chain 1: 7500 -9191.382 0.167 0.156
Chain 1: 7600 -9289.101 0.148 0.080
Chain 1: 7700 -9047.917 0.143 0.057
Chain 1: 7800 -13455.957 0.160 0.057
Chain 1: 7900 -9085.089 0.185 0.057
Chain 1: 8000 -14139.780 0.185 0.057
Chain 1: 8100 -11188.222 0.161 0.057
Chain 1: 8200 -9070.761 0.179 0.233
Chain 1: 8300 -8864.871 0.176 0.233
Chain 1: 8400 -9942.922 0.187 0.233
Chain 1: 8500 -9150.258 0.192 0.233
Chain 1: 8600 -9006.486 0.192 0.233
Chain 1: 8700 -10842.856 0.207 0.233
Chain 1: 8800 -8850.549 0.196 0.225
Chain 1: 8900 -10869.743 0.167 0.186
Chain 1: 9000 -13043.595 0.148 0.169
Chain 1: 9100 -12065.618 0.130 0.167
Chain 1: 9200 -9030.760 0.140 0.167
Chain 1: 9300 -8718.373 0.141 0.167
Chain 1: 9400 -13065.070 0.164 0.169
Chain 1: 9500 -8991.390 0.200 0.186
Chain 1: 9600 -9339.111 0.202 0.186
Chain 1: 9700 -11168.624 0.202 0.186
Chain 1: 9800 -10616.322 0.184 0.167
Chain 1: 9900 -11504.282 0.174 0.164
Chain 1: 10000 -9918.705 0.173 0.160
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001707 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 17.07 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58224.211 1.000 1.000
Chain 1: 200 -18158.020 1.603 2.207
Chain 1: 300 -9151.872 1.397 1.000
Chain 1: 400 -8348.872 1.072 1.000
Chain 1: 500 -8586.259 0.863 0.984
Chain 1: 600 -8390.825 0.723 0.984
Chain 1: 700 -7807.365 0.630 0.096
Chain 1: 800 -8564.081 0.563 0.096
Chain 1: 900 -7831.904 0.510 0.093
Chain 1: 1000 -8001.297 0.462 0.093
Chain 1: 1100 -7947.822 0.362 0.088
Chain 1: 1200 -7747.157 0.144 0.075
Chain 1: 1300 -7654.642 0.047 0.028
Chain 1: 1400 -7895.140 0.040 0.028
Chain 1: 1500 -7561.861 0.042 0.030
Chain 1: 1600 -7752.646 0.042 0.030
Chain 1: 1700 -7692.690 0.035 0.026
Chain 1: 1800 -7794.912 0.028 0.025
Chain 1: 1900 -7652.613 0.020 0.021
Chain 1: 2000 -7740.893 0.019 0.019
Chain 1: 2100 -7595.507 0.021 0.019
Chain 1: 2200 -8017.557 0.023 0.019
Chain 1: 2300 -7767.747 0.025 0.025
Chain 1: 2400 -7659.849 0.024 0.019
Chain 1: 2500 -7572.802 0.021 0.019
Chain 1: 2600 -7578.851 0.018 0.014
Chain 1: 2700 -7595.881 0.018 0.014
Chain 1: 2800 -7702.013 0.018 0.014
Chain 1: 2900 -7416.226 0.020 0.014
Chain 1: 3000 -7566.826 0.020 0.019
Chain 1: 3100 -7567.373 0.019 0.014
Chain 1: 3200 -7812.853 0.016 0.014
Chain 1: 3300 -7446.005 0.018 0.014
Chain 1: 3400 -7736.299 0.021 0.020
Chain 1: 3500 -7496.404 0.023 0.031
Chain 1: 3600 -7544.482 0.023 0.031
Chain 1: 3700 -7491.117 0.024 0.031
Chain 1: 3800 -7476.562 0.022 0.031
Chain 1: 3900 -7460.474 0.019 0.020
Chain 1: 4000 -7440.496 0.017 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003887 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 38.87 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87221.501 1.000 1.000
Chain 1: 200 -14342.026 3.041 5.082
Chain 1: 300 -10584.992 2.145 1.000
Chain 1: 400 -12228.908 1.643 1.000
Chain 1: 500 -9206.460 1.380 0.355
Chain 1: 600 -9138.146 1.151 0.355
Chain 1: 700 -9379.792 0.990 0.328
Chain 1: 800 -9837.712 0.872 0.328
Chain 1: 900 -9374.603 0.781 0.134
Chain 1: 1000 -9358.398 0.703 0.134
Chain 1: 1100 -9367.535 0.603 0.049 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8963.834 0.099 0.047
Chain 1: 1300 -9227.143 0.067 0.045
Chain 1: 1400 -9043.554 0.055 0.029
Chain 1: 1500 -9074.919 0.023 0.026
Chain 1: 1600 -9183.881 0.023 0.026
Chain 1: 1700 -9240.040 0.021 0.020
Chain 1: 1800 -8793.215 0.022 0.020
Chain 1: 1900 -8900.043 0.018 0.012
Chain 1: 2000 -8883.863 0.018 0.012
Chain 1: 2100 -9022.421 0.020 0.015
Chain 1: 2200 -8794.689 0.018 0.015
Chain 1: 2300 -8890.970 0.016 0.012
Chain 1: 2400 -8965.602 0.015 0.012
Chain 1: 2500 -8907.641 0.015 0.012
Chain 1: 2600 -8924.689 0.014 0.011
Chain 1: 2700 -8830.697 0.014 0.011
Chain 1: 2800 -8775.276 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003796 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 37.96 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8389870.294 1.000 1.000
Chain 1: 200 -1580712.621 2.654 4.308
Chain 1: 300 -890803.329 2.027 1.000
Chain 1: 400 -458222.303 1.757 1.000
Chain 1: 500 -358909.565 1.461 0.944
Chain 1: 600 -234009.822 1.306 0.944
Chain 1: 700 -120197.987 1.255 0.944
Chain 1: 800 -87395.193 1.145 0.944
Chain 1: 900 -67723.918 1.050 0.774
Chain 1: 1000 -52510.840 0.974 0.774
Chain 1: 1100 -39972.961 0.905 0.534 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39155.806 0.477 0.375
Chain 1: 1300 -27082.659 0.444 0.375
Chain 1: 1400 -26803.227 0.350 0.314
Chain 1: 1500 -23381.407 0.337 0.314
Chain 1: 1600 -22596.708 0.287 0.290
Chain 1: 1700 -21465.750 0.198 0.290
Chain 1: 1800 -21409.495 0.161 0.146
Chain 1: 1900 -21736.225 0.133 0.053
Chain 1: 2000 -20243.845 0.112 0.053
Chain 1: 2100 -20482.608 0.081 0.035
Chain 1: 2200 -20709.751 0.080 0.035
Chain 1: 2300 -20326.129 0.038 0.019
Chain 1: 2400 -20097.924 0.038 0.019
Chain 1: 2500 -19900.063 0.024 0.015
Chain 1: 2600 -19529.577 0.023 0.015
Chain 1: 2700 -19486.333 0.018 0.012
Chain 1: 2800 -19202.951 0.019 0.015
Chain 1: 2900 -19484.511 0.019 0.014
Chain 1: 3000 -19470.662 0.011 0.012
Chain 1: 3100 -19555.770 0.011 0.011
Chain 1: 3200 -19246.005 0.011 0.014
Chain 1: 3300 -19451.078 0.010 0.011
Chain 1: 3400 -18925.300 0.012 0.014
Chain 1: 3500 -19538.297 0.014 0.015
Chain 1: 3600 -18843.461 0.016 0.015
Chain 1: 3700 -19231.412 0.018 0.016
Chain 1: 3800 -18188.866 0.022 0.020
Chain 1: 3900 -18184.947 0.021 0.020
Chain 1: 4000 -18302.234 0.021 0.020
Chain 1: 4100 -18215.888 0.021 0.020
Chain 1: 4200 -18031.652 0.021 0.020
Chain 1: 4300 -18170.396 0.020 0.020
Chain 1: 4400 -18126.819 0.018 0.010
Chain 1: 4500 -18029.264 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001277 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.77 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49176.813 1.000 1.000
Chain 1: 200 -16825.497 1.461 1.923
Chain 1: 300 -16178.961 0.988 1.000
Chain 1: 400 -19861.285 0.787 1.000
Chain 1: 500 -15503.702 0.686 0.281
Chain 1: 600 -17000.497 0.586 0.281
Chain 1: 700 -16464.107 0.507 0.185
Chain 1: 800 -15000.608 0.456 0.185
Chain 1: 900 -10848.956 0.448 0.185
Chain 1: 1000 -10909.001 0.404 0.185
Chain 1: 1100 -10583.856 0.307 0.098
Chain 1: 1200 -10455.407 0.116 0.088
Chain 1: 1300 -12754.377 0.130 0.098
Chain 1: 1400 -11209.540 0.125 0.098
Chain 1: 1500 -13122.424 0.111 0.098
Chain 1: 1600 -10258.598 0.130 0.138
Chain 1: 1700 -10302.730 0.128 0.138
Chain 1: 1800 -10655.403 0.121 0.138
Chain 1: 1900 -10860.527 0.085 0.033
Chain 1: 2000 -9568.184 0.098 0.135
Chain 1: 2100 -9625.912 0.095 0.135
Chain 1: 2200 -16602.706 0.136 0.138
Chain 1: 2300 -9310.163 0.196 0.138
Chain 1: 2400 -9781.856 0.187 0.135
Chain 1: 2500 -10778.090 0.182 0.092
Chain 1: 2600 -9565.768 0.167 0.092
Chain 1: 2700 -9368.549 0.168 0.092
Chain 1: 2800 -12740.321 0.192 0.127
Chain 1: 2900 -12425.775 0.192 0.127
Chain 1: 3000 -8835.160 0.219 0.127
Chain 1: 3100 -9337.925 0.224 0.127
Chain 1: 3200 -13208.955 0.211 0.127
Chain 1: 3300 -9506.520 0.172 0.127
Chain 1: 3400 -9279.255 0.170 0.127
Chain 1: 3500 -9607.412 0.164 0.127
Chain 1: 3600 -9468.395 0.153 0.054
Chain 1: 3700 -9847.340 0.154 0.054
Chain 1: 3800 -8796.666 0.140 0.054
Chain 1: 3900 -13292.043 0.171 0.119
Chain 1: 4000 -9597.189 0.169 0.119
Chain 1: 4100 -9059.815 0.170 0.119
Chain 1: 4200 -9217.915 0.142 0.059
Chain 1: 4300 -9945.022 0.110 0.059
Chain 1: 4400 -11013.960 0.118 0.073
Chain 1: 4500 -10429.415 0.120 0.073
Chain 1: 4600 -10013.215 0.123 0.073
Chain 1: 4700 -10210.334 0.121 0.073
Chain 1: 4800 -12339.619 0.126 0.073
Chain 1: 4900 -9127.093 0.127 0.073
Chain 1: 5000 -9722.056 0.095 0.061
Chain 1: 5100 -10612.686 0.097 0.073
Chain 1: 5200 -11930.263 0.107 0.084
Chain 1: 5300 -12465.815 0.104 0.084
Chain 1: 5400 -8558.576 0.140 0.084
Chain 1: 5500 -8488.192 0.135 0.084
Chain 1: 5600 -9289.842 0.139 0.086
Chain 1: 5700 -9945.805 0.144 0.086
Chain 1: 5800 -9454.674 0.132 0.084
Chain 1: 5900 -8513.246 0.108 0.084
Chain 1: 6000 -11743.354 0.129 0.086
Chain 1: 6100 -8747.418 0.155 0.110
Chain 1: 6200 -8749.542 0.144 0.086
Chain 1: 6300 -11846.758 0.166 0.111
Chain 1: 6400 -9203.655 0.149 0.111
Chain 1: 6500 -9087.555 0.149 0.111
Chain 1: 6600 -11379.898 0.161 0.201
Chain 1: 6700 -13847.165 0.172 0.201
Chain 1: 6800 -9929.691 0.206 0.261
Chain 1: 6900 -10262.594 0.199 0.261
Chain 1: 7000 -9610.502 0.178 0.201
Chain 1: 7100 -8548.248 0.156 0.178
Chain 1: 7200 -8568.627 0.156 0.178
Chain 1: 7300 -9222.335 0.137 0.124
Chain 1: 7400 -8716.772 0.114 0.071
Chain 1: 7500 -10804.020 0.132 0.124
Chain 1: 7600 -8750.570 0.136 0.124
Chain 1: 7700 -8782.204 0.118 0.071
Chain 1: 7800 -8547.390 0.081 0.068
Chain 1: 7900 -9211.560 0.085 0.071
Chain 1: 8000 -8585.223 0.086 0.072
Chain 1: 8100 -8342.497 0.076 0.071
Chain 1: 8200 -10740.914 0.099 0.072
Chain 1: 8300 -13151.570 0.110 0.073
Chain 1: 8400 -12220.011 0.112 0.076
Chain 1: 8500 -8425.597 0.137 0.076
Chain 1: 8600 -8839.893 0.119 0.073
Chain 1: 8700 -8220.553 0.126 0.075
Chain 1: 8800 -8946.253 0.131 0.076
Chain 1: 8900 -12314.809 0.151 0.081
Chain 1: 9000 -11293.133 0.153 0.090
Chain 1: 9100 -8871.997 0.177 0.183
Chain 1: 9200 -9745.435 0.164 0.090
Chain 1: 9300 -8363.099 0.162 0.090
Chain 1: 9400 -8685.740 0.158 0.090
Chain 1: 9500 -8237.656 0.119 0.090
Chain 1: 9600 -9911.904 0.131 0.090
Chain 1: 9700 -8773.467 0.136 0.130
Chain 1: 9800 -8461.420 0.132 0.130
Chain 1: 9900 -10624.741 0.125 0.130
Chain 1: 10000 -8256.753 0.145 0.165
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001521 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.21 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57043.469 1.000 1.000
Chain 1: 200 -17498.829 1.630 2.260
Chain 1: 300 -8664.610 1.426 1.020
Chain 1: 400 -7879.860 1.095 1.020
Chain 1: 500 -8648.853 0.894 1.000
Chain 1: 600 -8780.647 0.747 1.000
Chain 1: 700 -7793.165 0.659 0.127
Chain 1: 800 -8103.121 0.581 0.127
Chain 1: 900 -7812.982 0.521 0.100
Chain 1: 1000 -7740.306 0.469 0.100
Chain 1: 1100 -7521.507 0.372 0.089
Chain 1: 1200 -7918.713 0.151 0.050
Chain 1: 1300 -7721.232 0.052 0.038
Chain 1: 1400 -7765.220 0.043 0.037
Chain 1: 1500 -7569.877 0.036 0.029
Chain 1: 1600 -7731.133 0.037 0.029
Chain 1: 1700 -7513.855 0.027 0.029
Chain 1: 1800 -7554.271 0.024 0.026
Chain 1: 1900 -7572.939 0.020 0.026
Chain 1: 2000 -7636.844 0.020 0.026
Chain 1: 2100 -7564.066 0.018 0.021
Chain 1: 2200 -7694.743 0.015 0.017
Chain 1: 2300 -7508.367 0.015 0.017
Chain 1: 2400 -7636.352 0.016 0.017
Chain 1: 2500 -7616.676 0.014 0.017
Chain 1: 2600 -7518.640 0.013 0.013
Chain 1: 2700 -7440.728 0.011 0.010
Chain 1: 2800 -7562.060 0.012 0.013
Chain 1: 2900 -7394.522 0.014 0.016
Chain 1: 3000 -7523.304 0.015 0.017
Chain 1: 3100 -7516.147 0.014 0.017
Chain 1: 3200 -7704.214 0.015 0.017
Chain 1: 3300 -7452.884 0.016 0.017
Chain 1: 3400 -7652.663 0.017 0.017
Chain 1: 3500 -7428.745 0.019 0.023
Chain 1: 3600 -7491.080 0.019 0.023
Chain 1: 3700 -7441.366 0.019 0.023
Chain 1: 3800 -7448.559 0.017 0.023
Chain 1: 3900 -7413.817 0.015 0.017
Chain 1: 4000 -7407.307 0.014 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002502 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.02 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86417.531 1.000 1.000
Chain 1: 200 -13602.639 3.176 5.353
Chain 1: 300 -9992.222 2.238 1.000
Chain 1: 400 -10769.215 1.697 1.000
Chain 1: 500 -8967.621 1.397 0.361
Chain 1: 600 -8483.898 1.174 0.361
Chain 1: 700 -8559.634 1.008 0.201
Chain 1: 800 -9167.397 0.890 0.201
Chain 1: 900 -8819.454 0.795 0.072
Chain 1: 1000 -8647.782 0.718 0.072
Chain 1: 1100 -8790.388 0.620 0.066 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8370.712 0.089 0.057
Chain 1: 1300 -8664.605 0.056 0.050
Chain 1: 1400 -8701.474 0.050 0.039
Chain 1: 1500 -8578.221 0.031 0.034
Chain 1: 1600 -8690.137 0.027 0.020
Chain 1: 1700 -8769.960 0.027 0.020
Chain 1: 1800 -8363.311 0.025 0.020
Chain 1: 1900 -8460.778 0.022 0.016
Chain 1: 2000 -8432.891 0.020 0.014
Chain 1: 2100 -8553.350 0.020 0.014
Chain 1: 2200 -8362.462 0.017 0.014
Chain 1: 2300 -8500.248 0.016 0.014
Chain 1: 2400 -8507.300 0.015 0.014
Chain 1: 2500 -8474.172 0.014 0.013
Chain 1: 2600 -8472.232 0.013 0.012
Chain 1: 2700 -8386.052 0.013 0.012
Chain 1: 2800 -8351.387 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003139 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.39 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8419626.226 1.000 1.000
Chain 1: 200 -1586512.597 2.654 4.307
Chain 1: 300 -890755.754 2.029 1.000
Chain 1: 400 -457979.366 1.758 1.000
Chain 1: 500 -358199.196 1.462 0.945
Chain 1: 600 -233000.892 1.308 0.945
Chain 1: 700 -119239.113 1.258 0.945
Chain 1: 800 -86470.312 1.148 0.945
Chain 1: 900 -66819.445 1.053 0.781
Chain 1: 1000 -51626.713 0.977 0.781
Chain 1: 1100 -39120.838 0.909 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38296.250 0.480 0.379
Chain 1: 1300 -26273.943 0.448 0.379
Chain 1: 1400 -25993.929 0.355 0.320
Chain 1: 1500 -22587.307 0.342 0.320
Chain 1: 1600 -21805.511 0.292 0.294
Chain 1: 1700 -20681.918 0.202 0.294
Chain 1: 1800 -20626.571 0.164 0.151
Chain 1: 1900 -20952.580 0.136 0.054
Chain 1: 2000 -19465.501 0.115 0.054
Chain 1: 2100 -19703.712 0.084 0.036
Chain 1: 2200 -19929.924 0.083 0.036
Chain 1: 2300 -19547.354 0.039 0.020
Chain 1: 2400 -19319.533 0.039 0.020
Chain 1: 2500 -19121.517 0.025 0.016
Chain 1: 2600 -18751.957 0.023 0.016
Chain 1: 2700 -18708.959 0.018 0.012
Chain 1: 2800 -18425.909 0.019 0.015
Chain 1: 2900 -18707.013 0.019 0.015
Chain 1: 3000 -18693.208 0.012 0.012
Chain 1: 3100 -18778.209 0.011 0.012
Chain 1: 3200 -18469.018 0.012 0.015
Chain 1: 3300 -18673.610 0.011 0.012
Chain 1: 3400 -18148.788 0.012 0.015
Chain 1: 3500 -18760.309 0.015 0.015
Chain 1: 3600 -18067.375 0.017 0.015
Chain 1: 3700 -18453.909 0.018 0.017
Chain 1: 3800 -17414.259 0.023 0.021
Chain 1: 3900 -17410.394 0.021 0.021
Chain 1: 4000 -17527.701 0.022 0.021
Chain 1: 4100 -17441.550 0.022 0.021
Chain 1: 4200 -17257.875 0.021 0.021
Chain 1: 4300 -17396.201 0.021 0.021
Chain 1: 4400 -17353.150 0.018 0.011
Chain 1: 4500 -17255.683 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001332 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.32 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -14274.546 1.000 1.000
Chain 1: 200 -10883.808 0.656 1.000
Chain 1: 300 -9320.575 0.493 0.312
Chain 1: 400 -8911.243 0.381 0.312
Chain 1: 500 -9035.887 0.308 0.168
Chain 1: 600 -8790.860 0.261 0.168
Chain 1: 700 -8678.736 0.226 0.046
Chain 1: 800 -8706.592 0.198 0.046
Chain 1: 900 -8752.965 0.176 0.028
Chain 1: 1000 -8764.662 0.159 0.028
Chain 1: 1100 -8774.864 0.059 0.014
Chain 1: 1200 -8709.804 0.029 0.013
Chain 1: 1300 -8622.916 0.013 0.010
Chain 1: 1400 -8656.353 0.009 0.007 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.0014 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61917.166 1.000 1.000
Chain 1: 200 -19241.154 1.609 2.218
Chain 1: 300 -9420.447 1.420 1.042
Chain 1: 400 -8575.155 1.090 1.042
Chain 1: 500 -9279.858 0.887 1.000
Chain 1: 600 -8997.075 0.744 1.000
Chain 1: 700 -8295.124 0.650 0.099
Chain 1: 800 -9014.448 0.579 0.099
Chain 1: 900 -7609.207 0.535 0.099
Chain 1: 1000 -8157.595 0.488 0.099
Chain 1: 1100 -7710.022 0.394 0.085
Chain 1: 1200 -7915.711 0.175 0.080
Chain 1: 1300 -7923.302 0.071 0.076
Chain 1: 1400 -8194.326 0.064 0.067
Chain 1: 1500 -7824.998 0.061 0.058
Chain 1: 1600 -8113.629 0.062 0.058
Chain 1: 1700 -7799.961 0.057 0.047
Chain 1: 1800 -7794.405 0.049 0.040
Chain 1: 1900 -7773.323 0.031 0.036
Chain 1: 2000 -7886.402 0.026 0.033
Chain 1: 2100 -7691.489 0.023 0.026
Chain 1: 2200 -8123.926 0.025 0.033
Chain 1: 2300 -7708.433 0.031 0.036
Chain 1: 2400 -7879.003 0.029 0.036
Chain 1: 2500 -7742.656 0.027 0.025
Chain 1: 2600 -7688.570 0.024 0.022
Chain 1: 2700 -7615.382 0.021 0.018
Chain 1: 2800 -7547.636 0.021 0.018
Chain 1: 2900 -7537.075 0.021 0.018
Chain 1: 3000 -7677.226 0.022 0.018
Chain 1: 3100 -7670.738 0.019 0.018
Chain 1: 3200 -7985.495 0.018 0.018
Chain 1: 3300 -7534.695 0.018 0.018
Chain 1: 3400 -7912.103 0.021 0.018
Chain 1: 3500 -7620.042 0.023 0.018
Chain 1: 3600 -7737.678 0.024 0.018
Chain 1: 3700 -7529.736 0.026 0.028
Chain 1: 3800 -7765.828 0.028 0.030
Chain 1: 3900 -7533.747 0.031 0.031
Chain 1: 4000 -7518.562 0.029 0.031
Chain 1: 4100 -7533.869 0.029 0.031
Chain 1: 4200 -7687.982 0.027 0.030
Chain 1: 4300 -7511.990 0.024 0.028
Chain 1: 4400 -7565.794 0.020 0.023
Chain 1: 4500 -7733.714 0.018 0.022
Chain 1: 4600 -7590.031 0.018 0.022
Chain 1: 4700 -7576.194 0.016 0.020
Chain 1: 4800 -7508.792 0.014 0.019
Chain 1: 4900 -7816.047 0.015 0.019
Chain 1: 5000 -7739.224 0.015 0.019
Chain 1: 5100 -7618.770 0.017 0.019
Chain 1: 5200 -7564.967 0.015 0.016
Chain 1: 5300 -7627.961 0.014 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002582 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.82 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87554.310 1.000 1.000
Chain 1: 200 -14736.157 2.971 4.941
Chain 1: 300 -10914.294 2.097 1.000
Chain 1: 400 -12952.313 1.612 1.000
Chain 1: 500 -9292.659 1.369 0.394
Chain 1: 600 -9283.635 1.141 0.394
Chain 1: 700 -9181.143 0.979 0.350
Chain 1: 800 -10128.580 0.869 0.350
Chain 1: 900 -9498.850 0.779 0.157
Chain 1: 1000 -9909.688 0.706 0.157
Chain 1: 1100 -9656.525 0.608 0.094 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -9079.902 0.120 0.066
Chain 1: 1300 -9486.407 0.090 0.064
Chain 1: 1400 -9349.266 0.075 0.043
Chain 1: 1500 -9350.453 0.036 0.041
Chain 1: 1600 -9435.121 0.037 0.041
Chain 1: 1700 -9491.466 0.036 0.041
Chain 1: 1800 -9027.885 0.032 0.041
Chain 1: 1900 -9144.064 0.027 0.026
Chain 1: 2000 -9164.903 0.023 0.015
Chain 1: 2100 -9255.627 0.021 0.013
Chain 1: 2200 -9024.696 0.017 0.013
Chain 1: 2300 -9222.947 0.015 0.013
Chain 1: 2400 -9047.471 0.016 0.013
Chain 1: 2500 -9114.542 0.016 0.013
Chain 1: 2600 -9022.623 0.017 0.013
Chain 1: 2700 -9057.125 0.016 0.013
Chain 1: 2800 -9010.437 0.012 0.010
Chain 1: 2900 -9123.340 0.012 0.010
Chain 1: 3000 -9031.597 0.013 0.010
Chain 1: 3100 -8999.440 0.012 0.010
Chain 1: 3200 -8969.538 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002971 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.71 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8385965.286 1.000 1.000
Chain 1: 200 -1579788.872 2.654 4.308
Chain 1: 300 -890825.984 2.027 1.000
Chain 1: 400 -458052.977 1.757 1.000
Chain 1: 500 -359047.091 1.460 0.945
Chain 1: 600 -234110.810 1.306 0.945
Chain 1: 700 -120488.225 1.254 0.943
Chain 1: 800 -87717.046 1.144 0.943
Chain 1: 900 -68072.828 1.049 0.773
Chain 1: 1000 -52885.647 0.973 0.773
Chain 1: 1100 -40363.619 0.904 0.534 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39554.098 0.475 0.374
Chain 1: 1300 -27491.230 0.442 0.374
Chain 1: 1400 -27214.334 0.348 0.310
Chain 1: 1500 -23794.749 0.335 0.310
Chain 1: 1600 -23011.183 0.285 0.289
Chain 1: 1700 -21881.111 0.196 0.287
Chain 1: 1800 -21825.461 0.159 0.144
Chain 1: 1900 -22152.561 0.131 0.052
Chain 1: 2000 -20660.012 0.110 0.052
Chain 1: 2100 -20898.759 0.080 0.034
Chain 1: 2200 -21126.058 0.079 0.034
Chain 1: 2300 -20742.263 0.037 0.019
Chain 1: 2400 -20513.962 0.037 0.019
Chain 1: 2500 -20315.970 0.024 0.015
Chain 1: 2600 -19945.151 0.022 0.015
Chain 1: 2700 -19901.889 0.017 0.011
Chain 1: 2800 -19618.238 0.018 0.014
Chain 1: 2900 -19900.026 0.018 0.014
Chain 1: 3000 -19886.126 0.011 0.011
Chain 1: 3100 -19971.256 0.010 0.011
Chain 1: 3200 -19661.273 0.011 0.014
Chain 1: 3300 -19866.564 0.010 0.011
Chain 1: 3400 -19340.294 0.012 0.014
Chain 1: 3500 -19953.948 0.014 0.014
Chain 1: 3600 -19258.341 0.016 0.014
Chain 1: 3700 -19646.831 0.017 0.016
Chain 1: 3800 -18602.971 0.022 0.020
Chain 1: 3900 -18599.028 0.020 0.020
Chain 1: 4000 -18716.335 0.021 0.020
Chain 1: 4100 -18629.870 0.021 0.020
Chain 1: 4200 -18445.401 0.020 0.020
Chain 1: 4300 -18584.322 0.020 0.020
Chain 1: 4400 -18540.517 0.017 0.010
Chain 1: 4500 -18442.936 0.015 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001189 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 11.89 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12762.610 1.000 1.000
Chain 1: 200 -9254.676 0.690 1.000
Chain 1: 300 -7994.276 0.512 0.379
Chain 1: 400 -7994.372 0.384 0.379
Chain 1: 500 -7902.302 0.310 0.158
Chain 1: 600 -7810.484 0.260 0.158
Chain 1: 700 -7739.155 0.224 0.012
Chain 1: 800 -7719.748 0.196 0.012
Chain 1: 900 -7885.574 0.177 0.012
Chain 1: 1000 -7818.604 0.160 0.012
Chain 1: 1100 -7853.011 0.061 0.012
Chain 1: 1200 -7762.893 0.024 0.012
Chain 1: 1300 -7737.527 0.008 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001424 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.24 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57732.569 1.000 1.000
Chain 1: 200 -17287.044 1.670 2.340
Chain 1: 300 -8532.261 1.455 1.026
Chain 1: 400 -8095.537 1.105 1.026
Chain 1: 500 -8287.623 0.889 1.000
Chain 1: 600 -8754.843 0.749 1.000
Chain 1: 700 -7772.098 0.660 0.126
Chain 1: 800 -8095.321 0.583 0.126
Chain 1: 900 -7791.236 0.522 0.054
Chain 1: 1000 -7606.191 0.473 0.054
Chain 1: 1100 -7720.408 0.374 0.053
Chain 1: 1200 -7593.676 0.142 0.040
Chain 1: 1300 -7596.211 0.039 0.039
Chain 1: 1400 -7645.956 0.034 0.024
Chain 1: 1500 -7609.686 0.033 0.024
Chain 1: 1600 -7511.092 0.029 0.017
Chain 1: 1700 -7498.453 0.016 0.015
Chain 1: 1800 -7534.823 0.013 0.013
Chain 1: 1900 -7587.184 0.009 0.007 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002486 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 24.86 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85561.563 1.000 1.000
Chain 1: 200 -13177.078 3.247 5.493
Chain 1: 300 -9620.656 2.288 1.000
Chain 1: 400 -10159.084 1.729 1.000
Chain 1: 500 -8581.346 1.420 0.370
Chain 1: 600 -8139.451 1.192 0.370
Chain 1: 700 -8486.013 1.028 0.184
Chain 1: 800 -9065.145 0.907 0.184
Chain 1: 900 -8469.451 0.814 0.070
Chain 1: 1000 -8197.311 0.736 0.070
Chain 1: 1100 -8454.649 0.639 0.064 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8144.514 0.094 0.054
Chain 1: 1300 -8343.834 0.059 0.053
Chain 1: 1400 -8360.639 0.054 0.041
Chain 1: 1500 -8260.090 0.037 0.038
Chain 1: 1600 -8353.602 0.033 0.033
Chain 1: 1700 -8447.245 0.030 0.030
Chain 1: 1800 -8061.989 0.028 0.030
Chain 1: 1900 -8164.347 0.022 0.024
Chain 1: 2000 -8134.073 0.019 0.013
Chain 1: 2100 -8269.083 0.018 0.013
Chain 1: 2200 -8053.158 0.017 0.013
Chain 1: 2300 -8194.424 0.016 0.013
Chain 1: 2400 -8205.118 0.016 0.013
Chain 1: 2500 -8173.466 0.015 0.013
Chain 1: 2600 -8171.365 0.014 0.013
Chain 1: 2700 -8080.777 0.014 0.013
Chain 1: 2800 -8059.318 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002561 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.61 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8397963.737 1.000 1.000
Chain 1: 200 -1585179.590 2.649 4.298
Chain 1: 300 -890488.310 2.026 1.000
Chain 1: 400 -457304.687 1.756 1.000
Chain 1: 500 -357442.334 1.461 0.947
Chain 1: 600 -232453.631 1.307 0.947
Chain 1: 700 -118761.692 1.257 0.947
Chain 1: 800 -86008.008 1.148 0.947
Chain 1: 900 -66363.489 1.053 0.780
Chain 1: 1000 -51170.939 0.977 0.780
Chain 1: 1100 -38663.064 0.910 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37835.002 0.482 0.381
Chain 1: 1300 -25814.160 0.451 0.381
Chain 1: 1400 -25533.002 0.357 0.324
Chain 1: 1500 -22126.785 0.344 0.324
Chain 1: 1600 -21344.565 0.294 0.297
Chain 1: 1700 -20221.581 0.204 0.296
Chain 1: 1800 -20166.227 0.166 0.154
Chain 1: 1900 -20491.800 0.138 0.056
Chain 1: 2000 -19005.750 0.117 0.056
Chain 1: 2100 -19243.884 0.085 0.037
Chain 1: 2200 -19469.753 0.084 0.037
Chain 1: 2300 -19087.624 0.040 0.020
Chain 1: 2400 -18859.951 0.040 0.020
Chain 1: 2500 -18661.963 0.026 0.016
Chain 1: 2600 -18292.782 0.024 0.016
Chain 1: 2700 -18249.941 0.019 0.012
Chain 1: 2800 -17967.055 0.020 0.016
Chain 1: 2900 -18247.985 0.020 0.015
Chain 1: 3000 -18234.234 0.012 0.012
Chain 1: 3100 -18319.140 0.011 0.012
Chain 1: 3200 -18010.228 0.012 0.015
Chain 1: 3300 -18214.623 0.011 0.012
Chain 1: 3400 -17690.266 0.013 0.015
Chain 1: 3500 -18301.067 0.015 0.016
Chain 1: 3600 -17609.129 0.017 0.016
Chain 1: 3700 -17994.885 0.019 0.017
Chain 1: 3800 -16956.769 0.023 0.021
Chain 1: 3900 -16952.967 0.022 0.021
Chain 1: 4000 -17070.261 0.023 0.021
Chain 1: 4100 -16984.144 0.023 0.021
Chain 1: 4200 -16800.862 0.022 0.021
Chain 1: 4300 -16938.923 0.022 0.021
Chain 1: 4400 -16896.126 0.019 0.011
Chain 1: 4500 -16798.738 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001372 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.72 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12433.521 1.000 1.000
Chain 1: 200 -9368.807 0.664 1.000
Chain 1: 300 -8159.876 0.492 0.327
Chain 1: 400 -8288.026 0.373 0.327
Chain 1: 500 -8233.195 0.299 0.148
Chain 1: 600 -8100.546 0.252 0.148
Chain 1: 700 -8030.060 0.218 0.016
Chain 1: 800 -8033.689 0.190 0.016
Chain 1: 900 -8024.992 0.169 0.015
Chain 1: 1000 -8095.296 0.153 0.015
Chain 1: 1100 -8115.121 0.054 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001395 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.95 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -62250.418 1.000 1.000
Chain 1: 200 -17899.743 1.739 2.478
Chain 1: 300 -8908.698 1.496 1.009
Chain 1: 400 -9433.903 1.136 1.009
Chain 1: 500 -8008.084 0.944 1.000
Chain 1: 600 -8122.361 0.789 1.000
Chain 1: 700 -8009.725 0.678 0.178
Chain 1: 800 -8151.016 0.596 0.178
Chain 1: 900 -7987.218 0.532 0.056
Chain 1: 1000 -8150.313 0.481 0.056
Chain 1: 1100 -8216.068 0.381 0.021
Chain 1: 1200 -7802.099 0.139 0.021
Chain 1: 1300 -7788.354 0.038 0.020
Chain 1: 1400 -7706.806 0.034 0.017
Chain 1: 1500 -7640.565 0.017 0.014
Chain 1: 1600 -7698.235 0.016 0.014
Chain 1: 1700 -7586.666 0.016 0.015
Chain 1: 1800 -7656.351 0.015 0.011
Chain 1: 1900 -7654.587 0.013 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003514 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.14 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86222.532 1.000 1.000
Chain 1: 200 -13543.059 3.183 5.367
Chain 1: 300 -9962.711 2.242 1.000
Chain 1: 400 -10980.666 1.705 1.000
Chain 1: 500 -8909.373 1.410 0.359
Chain 1: 600 -8519.376 1.183 0.359
Chain 1: 700 -8538.944 1.014 0.232
Chain 1: 800 -8864.326 0.892 0.232
Chain 1: 900 -8773.216 0.794 0.093
Chain 1: 1000 -8544.516 0.717 0.093
Chain 1: 1100 -8678.224 0.619 0.046 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8463.839 0.085 0.037
Chain 1: 1300 -8668.331 0.051 0.027
Chain 1: 1400 -8665.220 0.042 0.025
Chain 1: 1500 -8562.299 0.020 0.024
Chain 1: 1600 -8664.589 0.016 0.015
Chain 1: 1700 -8752.455 0.017 0.015
Chain 1: 1800 -8352.674 0.018 0.015
Chain 1: 1900 -8453.262 0.019 0.015
Chain 1: 2000 -8424.209 0.016 0.012
Chain 1: 2100 -8544.623 0.016 0.012
Chain 1: 2200 -8321.190 0.016 0.012
Chain 1: 2300 -8482.493 0.016 0.012
Chain 1: 2400 -8496.673 0.016 0.012
Chain 1: 2500 -8463.048 0.015 0.012
Chain 1: 2600 -8468.831 0.014 0.012
Chain 1: 2700 -8374.463 0.014 0.012
Chain 1: 2800 -8344.890 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002481 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 24.81 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8403194.991 1.000 1.000
Chain 1: 200 -1583483.864 2.653 4.307
Chain 1: 300 -890911.597 2.028 1.000
Chain 1: 400 -458370.553 1.757 1.000
Chain 1: 500 -358795.085 1.461 0.944
Chain 1: 600 -233474.433 1.307 0.944
Chain 1: 700 -119420.407 1.257 0.944
Chain 1: 800 -86608.204 1.147 0.944
Chain 1: 900 -66899.962 1.052 0.777
Chain 1: 1000 -51665.693 0.977 0.777
Chain 1: 1100 -39122.814 0.909 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38291.120 0.480 0.379
Chain 1: 1300 -26232.111 0.448 0.379
Chain 1: 1400 -25947.317 0.355 0.321
Chain 1: 1500 -22532.440 0.342 0.321
Chain 1: 1600 -21747.974 0.292 0.295
Chain 1: 1700 -20620.110 0.202 0.295
Chain 1: 1800 -20563.735 0.165 0.152
Chain 1: 1900 -20889.550 0.137 0.055
Chain 1: 2000 -19400.765 0.115 0.055
Chain 1: 2100 -19638.868 0.084 0.036
Chain 1: 2200 -19865.455 0.083 0.036
Chain 1: 2300 -19482.642 0.039 0.020
Chain 1: 2400 -19254.874 0.039 0.020
Chain 1: 2500 -19057.097 0.025 0.016
Chain 1: 2600 -18687.398 0.023 0.016
Chain 1: 2700 -18644.334 0.018 0.012
Chain 1: 2800 -18361.506 0.020 0.015
Chain 1: 2900 -18642.595 0.019 0.015
Chain 1: 3000 -18628.672 0.012 0.012
Chain 1: 3100 -18713.701 0.011 0.012
Chain 1: 3200 -18404.510 0.012 0.015
Chain 1: 3300 -18609.110 0.011 0.012
Chain 1: 3400 -18084.437 0.013 0.015
Chain 1: 3500 -18695.787 0.015 0.015
Chain 1: 3600 -18003.092 0.017 0.015
Chain 1: 3700 -18389.508 0.018 0.017
Chain 1: 3800 -17350.277 0.023 0.021
Chain 1: 3900 -17346.485 0.021 0.021
Chain 1: 4000 -17463.735 0.022 0.021
Chain 1: 4100 -17377.639 0.022 0.021
Chain 1: 4200 -17194.064 0.021 0.021
Chain 1: 4300 -17332.285 0.021 0.021
Chain 1: 4400 -17289.289 0.019 0.011
Chain 1: 4500 -17191.904 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001334 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.34 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49478.775 1.000 1.000
Chain 1: 200 -19944.319 1.240 1.481
Chain 1: 300 -20660.244 0.838 1.000
Chain 1: 400 -19506.562 0.644 1.000
Chain 1: 500 -13446.356 0.605 0.451
Chain 1: 600 -14392.135 0.515 0.451
Chain 1: 700 -16495.202 0.460 0.127
Chain 1: 800 -15563.595 0.410 0.127
Chain 1: 900 -12211.670 0.395 0.127
Chain 1: 1000 -11472.468 0.362 0.127
Chain 1: 1100 -11870.151 0.265 0.066
Chain 1: 1200 -17527.980 0.149 0.066
Chain 1: 1300 -15081.627 0.162 0.127
Chain 1: 1400 -11082.624 0.192 0.162
Chain 1: 1500 -26748.948 0.206 0.162
Chain 1: 1600 -18468.170 0.244 0.274
Chain 1: 1700 -11733.589 0.289 0.323
Chain 1: 1800 -12671.473 0.290 0.323
Chain 1: 1900 -10793.579 0.280 0.323
Chain 1: 2000 -12310.508 0.286 0.323
Chain 1: 2100 -20549.033 0.323 0.361
Chain 1: 2200 -13419.699 0.343 0.401
Chain 1: 2300 -12307.508 0.336 0.401
Chain 1: 2400 -9679.632 0.327 0.401
Chain 1: 2500 -9743.007 0.269 0.271
Chain 1: 2600 -9951.864 0.227 0.174
Chain 1: 2700 -12575.447 0.190 0.174
Chain 1: 2800 -10881.556 0.198 0.174
Chain 1: 2900 -11217.736 0.184 0.156
Chain 1: 3000 -14135.856 0.192 0.206
Chain 1: 3100 -17739.178 0.172 0.203
Chain 1: 3200 -13713.430 0.149 0.203
Chain 1: 3300 -16087.027 0.154 0.203
Chain 1: 3400 -11630.049 0.166 0.203
Chain 1: 3500 -14724.157 0.186 0.206
Chain 1: 3600 -10630.462 0.222 0.209
Chain 1: 3700 -10477.177 0.203 0.206
Chain 1: 3800 -9204.818 0.201 0.206
Chain 1: 3900 -13225.128 0.229 0.210
Chain 1: 4000 -13097.920 0.209 0.210
Chain 1: 4100 -9288.403 0.230 0.294
Chain 1: 4200 -9130.529 0.202 0.210
Chain 1: 4300 -11286.267 0.206 0.210
Chain 1: 4400 -12486.079 0.178 0.191
Chain 1: 4500 -10711.543 0.173 0.166
Chain 1: 4600 -9132.754 0.152 0.166
Chain 1: 4700 -10396.962 0.163 0.166
Chain 1: 4800 -9393.618 0.160 0.166
Chain 1: 4900 -9807.247 0.133 0.122
Chain 1: 5000 -14658.734 0.165 0.166
Chain 1: 5100 -9836.968 0.173 0.166
Chain 1: 5200 -9192.224 0.179 0.166
Chain 1: 5300 -13792.640 0.193 0.166
Chain 1: 5400 -10874.349 0.210 0.173
Chain 1: 5500 -13931.268 0.216 0.219
Chain 1: 5600 -13533.195 0.201 0.219
Chain 1: 5700 -15446.974 0.201 0.219
Chain 1: 5800 -18810.055 0.209 0.219
Chain 1: 5900 -9485.954 0.303 0.268
Chain 1: 6000 -10611.716 0.280 0.219
Chain 1: 6100 -10199.604 0.235 0.179
Chain 1: 6200 -12002.415 0.243 0.179
Chain 1: 6300 -8869.834 0.245 0.179
Chain 1: 6400 -8707.024 0.220 0.150
Chain 1: 6500 -10304.958 0.214 0.150
Chain 1: 6600 -11743.863 0.223 0.150
Chain 1: 6700 -9746.213 0.231 0.155
Chain 1: 6800 -9029.484 0.221 0.150
Chain 1: 6900 -8950.298 0.124 0.123
Chain 1: 7000 -9125.194 0.115 0.123
Chain 1: 7100 -8793.447 0.115 0.123
Chain 1: 7200 -12874.174 0.132 0.123
Chain 1: 7300 -8778.043 0.143 0.123
Chain 1: 7400 -9662.489 0.150 0.123
Chain 1: 7500 -9239.920 0.139 0.092
Chain 1: 7600 -11863.709 0.149 0.092
Chain 1: 7700 -9116.075 0.159 0.092
Chain 1: 7800 -10087.989 0.161 0.096
Chain 1: 7900 -8957.494 0.172 0.126
Chain 1: 8000 -8734.941 0.173 0.126
Chain 1: 8100 -10633.893 0.187 0.179
Chain 1: 8200 -12106.588 0.167 0.126
Chain 1: 8300 -11406.718 0.127 0.122
Chain 1: 8400 -9016.163 0.144 0.126
Chain 1: 8500 -8913.900 0.141 0.126
Chain 1: 8600 -8654.339 0.122 0.122
Chain 1: 8700 -8627.237 0.092 0.096
Chain 1: 8800 -9151.280 0.088 0.061
Chain 1: 8900 -12700.850 0.103 0.061
Chain 1: 9000 -11976.064 0.107 0.061
Chain 1: 9100 -9267.666 0.118 0.061
Chain 1: 9200 -11027.467 0.122 0.061
Chain 1: 9300 -9309.327 0.134 0.160
Chain 1: 9400 -9150.917 0.110 0.061
Chain 1: 9500 -9525.019 0.112 0.061
Chain 1: 9600 -8841.221 0.117 0.077
Chain 1: 9700 -8795.826 0.117 0.077
Chain 1: 9800 -12008.427 0.138 0.160
Chain 1: 9900 -11385.671 0.116 0.077
Chain 1: 10000 -8652.741 0.141 0.160
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001455 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.55 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -62623.467 1.000 1.000
Chain 1: 200 -18523.191 1.690 2.381
Chain 1: 300 -9174.538 1.467 1.019
Chain 1: 400 -8906.997 1.107 1.019
Chain 1: 500 -8733.771 0.890 1.000
Chain 1: 600 -9481.739 0.755 1.000
Chain 1: 700 -8302.295 0.667 0.142
Chain 1: 800 -8029.890 0.588 0.142
Chain 1: 900 -8377.987 0.527 0.079
Chain 1: 1000 -7914.861 0.480 0.079
Chain 1: 1100 -7811.312 0.382 0.059
Chain 1: 1200 -7843.932 0.144 0.042
Chain 1: 1300 -7672.250 0.044 0.034
Chain 1: 1400 -7912.386 0.044 0.034
Chain 1: 1500 -7596.716 0.047 0.042
Chain 1: 1600 -7856.497 0.042 0.034
Chain 1: 1700 -7635.316 0.031 0.033
Chain 1: 1800 -7684.803 0.028 0.030
Chain 1: 1900 -7645.844 0.024 0.029
Chain 1: 2000 -7758.165 0.020 0.022
Chain 1: 2100 -7604.707 0.021 0.022
Chain 1: 2200 -7815.037 0.023 0.027
Chain 1: 2300 -7631.115 0.023 0.027
Chain 1: 2400 -7777.778 0.022 0.024
Chain 1: 2500 -7669.631 0.019 0.020
Chain 1: 2600 -7568.316 0.017 0.019
Chain 1: 2700 -7556.992 0.015 0.014
Chain 1: 2800 -7557.025 0.014 0.014
Chain 1: 2900 -7408.114 0.015 0.019
Chain 1: 3000 -7563.962 0.016 0.020
Chain 1: 3100 -7561.780 0.014 0.019
Chain 1: 3200 -7786.840 0.014 0.019
Chain 1: 3300 -7501.559 0.016 0.019
Chain 1: 3400 -7754.547 0.017 0.020
Chain 1: 3500 -7475.376 0.019 0.021
Chain 1: 3600 -7538.993 0.019 0.021
Chain 1: 3700 -7490.748 0.019 0.021
Chain 1: 3800 -7473.549 0.020 0.021
Chain 1: 3900 -7442.642 0.018 0.021
Chain 1: 4000 -7437.762 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002608 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 26.08 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86543.806 1.000 1.000
Chain 1: 200 -14130.081 3.062 5.125
Chain 1: 300 -10456.839 2.159 1.000
Chain 1: 400 -11618.109 1.644 1.000
Chain 1: 500 -9445.397 1.361 0.351
Chain 1: 600 -9228.951 1.138 0.351
Chain 1: 700 -9019.205 0.979 0.230
Chain 1: 800 -9455.161 0.862 0.230
Chain 1: 900 -9267.811 0.769 0.100
Chain 1: 1000 -8902.064 0.696 0.100
Chain 1: 1100 -9277.511 0.600 0.046 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8892.394 0.092 0.043
Chain 1: 1300 -9168.117 0.060 0.041
Chain 1: 1400 -9163.640 0.050 0.040
Chain 1: 1500 -9007.117 0.029 0.030
Chain 1: 1600 -9119.085 0.027 0.030
Chain 1: 1700 -9198.673 0.026 0.030
Chain 1: 1800 -8773.491 0.026 0.030
Chain 1: 1900 -8875.240 0.025 0.030
Chain 1: 2000 -8850.058 0.022 0.017
Chain 1: 2100 -8975.949 0.019 0.014
Chain 1: 2200 -8777.034 0.017 0.014
Chain 1: 2300 -8870.216 0.015 0.012
Chain 1: 2400 -8938.722 0.016 0.012
Chain 1: 2500 -8885.052 0.014 0.011
Chain 1: 2600 -8886.843 0.013 0.011
Chain 1: 2700 -8803.311 0.013 0.011
Chain 1: 2800 -8762.650 0.009 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00251 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.1 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8401959.792 1.000 1.000
Chain 1: 200 -1583752.225 2.653 4.305
Chain 1: 300 -891933.944 2.027 1.000
Chain 1: 400 -458787.470 1.756 1.000
Chain 1: 500 -359254.836 1.460 0.944
Chain 1: 600 -234062.460 1.306 0.944
Chain 1: 700 -120064.188 1.255 0.944
Chain 1: 800 -87245.208 1.145 0.944
Chain 1: 900 -67542.085 1.050 0.776
Chain 1: 1000 -52310.000 0.975 0.776
Chain 1: 1100 -39759.153 0.906 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38933.912 0.478 0.376
Chain 1: 1300 -26854.841 0.445 0.376
Chain 1: 1400 -26571.990 0.352 0.316
Chain 1: 1500 -23150.214 0.339 0.316
Chain 1: 1600 -22364.734 0.289 0.292
Chain 1: 1700 -21233.679 0.199 0.291
Chain 1: 1800 -21176.910 0.162 0.148
Chain 1: 1900 -21503.221 0.134 0.053
Chain 1: 2000 -20011.802 0.113 0.053
Chain 1: 2100 -20250.253 0.082 0.035
Chain 1: 2200 -20477.293 0.081 0.035
Chain 1: 2300 -20093.951 0.038 0.019
Chain 1: 2400 -19865.942 0.038 0.019
Chain 1: 2500 -19668.228 0.024 0.015
Chain 1: 2600 -19298.057 0.023 0.015
Chain 1: 2700 -19254.885 0.018 0.012
Chain 1: 2800 -18971.800 0.019 0.015
Chain 1: 2900 -19253.161 0.019 0.015
Chain 1: 3000 -19239.274 0.012 0.012
Chain 1: 3100 -19324.321 0.011 0.011
Chain 1: 3200 -19014.833 0.011 0.015
Chain 1: 3300 -19219.694 0.010 0.011
Chain 1: 3400 -18694.453 0.012 0.015
Chain 1: 3500 -19306.653 0.014 0.015
Chain 1: 3600 -18612.902 0.016 0.015
Chain 1: 3700 -19000.073 0.018 0.016
Chain 1: 3800 -17959.203 0.022 0.020
Chain 1: 3900 -17955.363 0.021 0.020
Chain 1: 4000 -18072.623 0.021 0.020
Chain 1: 4100 -17986.395 0.021 0.020
Chain 1: 4200 -17802.502 0.021 0.020
Chain 1: 4300 -17940.959 0.020 0.020
Chain 1: 4400 -17897.674 0.018 0.010
Chain 1: 4500 -17800.222 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001315 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.15 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48891.832 1.000 1.000
Chain 1: 200 -21801.668 1.121 1.243
Chain 1: 300 -24400.871 0.783 1.000
Chain 1: 400 -38310.252 0.678 1.000
Chain 1: 500 -14098.437 0.886 1.000
Chain 1: 600 -14356.346 0.741 1.000
Chain 1: 700 -11593.534 0.669 0.363
Chain 1: 800 -14668.762 0.612 0.363
Chain 1: 900 -11812.325 0.571 0.242
Chain 1: 1000 -11274.674 0.518 0.242
Chain 1: 1100 -9779.337 0.434 0.238
Chain 1: 1200 -10929.048 0.320 0.210
Chain 1: 1300 -10165.027 0.317 0.210
Chain 1: 1400 -12679.725 0.300 0.198
Chain 1: 1500 -12079.218 0.134 0.153
Chain 1: 1600 -12293.254 0.134 0.153
Chain 1: 1700 -12101.031 0.111 0.105
Chain 1: 1800 -11401.379 0.097 0.075
Chain 1: 1900 -9865.210 0.088 0.075
Chain 1: 2000 -14079.067 0.113 0.105
Chain 1: 2100 -13166.838 0.105 0.075
Chain 1: 2200 -10127.810 0.124 0.075
Chain 1: 2300 -12164.355 0.133 0.156
Chain 1: 2400 -19453.708 0.151 0.156
Chain 1: 2500 -9077.002 0.260 0.167
Chain 1: 2600 -12430.334 0.286 0.270
Chain 1: 2700 -11653.191 0.291 0.270
Chain 1: 2800 -9000.183 0.314 0.295
Chain 1: 2900 -13341.456 0.331 0.299
Chain 1: 3000 -8722.905 0.354 0.300
Chain 1: 3100 -8994.287 0.350 0.300
Chain 1: 3200 -9705.557 0.327 0.295
Chain 1: 3300 -9846.339 0.312 0.295
Chain 1: 3400 -9139.301 0.282 0.270
Chain 1: 3500 -9175.235 0.169 0.077
Chain 1: 3600 -9303.178 0.143 0.073
Chain 1: 3700 -8750.217 0.143 0.073
Chain 1: 3800 -8733.714 0.113 0.063
Chain 1: 3900 -9047.416 0.084 0.035
Chain 1: 4000 -8679.292 0.035 0.035
Chain 1: 4100 -9368.596 0.040 0.042
Chain 1: 4200 -9165.222 0.035 0.035
Chain 1: 4300 -10626.499 0.047 0.042
Chain 1: 4400 -8819.615 0.060 0.042
Chain 1: 4500 -9420.235 0.066 0.063
Chain 1: 4600 -8813.680 0.071 0.064
Chain 1: 4700 -9966.900 0.077 0.069
Chain 1: 4800 -8716.879 0.091 0.074
Chain 1: 4900 -8461.591 0.090 0.074
Chain 1: 5000 -9872.287 0.100 0.116
Chain 1: 5100 -8540.687 0.109 0.138
Chain 1: 5200 -12720.206 0.139 0.143
Chain 1: 5300 -14031.218 0.135 0.143
Chain 1: 5400 -8850.160 0.173 0.143
Chain 1: 5500 -9806.741 0.176 0.143
Chain 1: 5600 -8409.558 0.186 0.143
Chain 1: 5700 -8871.553 0.180 0.143
Chain 1: 5800 -9379.766 0.171 0.143
Chain 1: 5900 -12804.683 0.194 0.156
Chain 1: 6000 -10891.805 0.198 0.166
Chain 1: 6100 -8332.806 0.213 0.176
Chain 1: 6200 -8255.914 0.181 0.166
Chain 1: 6300 -9523.963 0.185 0.166
Chain 1: 6400 -11535.790 0.144 0.166
Chain 1: 6500 -9131.670 0.160 0.174
Chain 1: 6600 -8544.897 0.151 0.174
Chain 1: 6700 -9446.642 0.155 0.174
Chain 1: 6800 -13006.647 0.177 0.176
Chain 1: 6900 -8146.554 0.210 0.176
Chain 1: 7000 -8509.314 0.196 0.174
Chain 1: 7100 -8139.078 0.170 0.133
Chain 1: 7200 -10577.224 0.192 0.174
Chain 1: 7300 -10834.875 0.181 0.174
Chain 1: 7400 -9274.305 0.181 0.168
Chain 1: 7500 -8642.803 0.162 0.095
Chain 1: 7600 -10939.021 0.176 0.168
Chain 1: 7700 -8811.868 0.191 0.210
Chain 1: 7800 -9492.663 0.170 0.168
Chain 1: 7900 -9246.155 0.113 0.073
Chain 1: 8000 -8924.393 0.113 0.073
Chain 1: 8100 -9295.948 0.112 0.073
Chain 1: 8200 -8782.570 0.095 0.072
Chain 1: 8300 -8311.419 0.098 0.072
Chain 1: 8400 -8144.678 0.083 0.058
Chain 1: 8500 -8110.902 0.077 0.057
Chain 1: 8600 -8689.020 0.062 0.057
Chain 1: 8700 -10995.527 0.059 0.057
Chain 1: 8800 -8097.391 0.088 0.057
Chain 1: 8900 -10735.617 0.110 0.058
Chain 1: 9000 -9949.902 0.114 0.067
Chain 1: 9100 -9283.842 0.117 0.072
Chain 1: 9200 -11734.113 0.132 0.079
Chain 1: 9300 -8871.278 0.159 0.209
Chain 1: 9400 -9311.068 0.161 0.209
Chain 1: 9500 -10346.223 0.171 0.209
Chain 1: 9600 -10248.873 0.165 0.209
Chain 1: 9700 -8279.708 0.168 0.209
Chain 1: 9800 -8538.059 0.135 0.100
Chain 1: 9900 -9234.375 0.118 0.079
Chain 1: 10000 -8211.585 0.123 0.100
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001375 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.75 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56864.177 1.000 1.000
Chain 1: 200 -17333.201 1.640 2.281
Chain 1: 300 -8680.386 1.426 1.000
Chain 1: 400 -8339.907 1.080 1.000
Chain 1: 500 -8290.669 0.865 0.997
Chain 1: 600 -8247.779 0.722 0.997
Chain 1: 700 -7967.456 0.624 0.041
Chain 1: 800 -8154.776 0.548 0.041
Chain 1: 900 -7793.654 0.493 0.041
Chain 1: 1000 -7754.773 0.444 0.041
Chain 1: 1100 -7746.479 0.344 0.035
Chain 1: 1200 -7578.123 0.118 0.023
Chain 1: 1300 -7492.280 0.020 0.022
Chain 1: 1400 -7612.348 0.017 0.016
Chain 1: 1500 -7589.228 0.017 0.016
Chain 1: 1600 -7533.851 0.017 0.016
Chain 1: 1700 -7499.462 0.014 0.011
Chain 1: 1800 -7566.707 0.013 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002613 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 26.13 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86172.286 1.000 1.000
Chain 1: 200 -13408.958 3.213 5.426
Chain 1: 300 -9809.436 2.264 1.000
Chain 1: 400 -10692.927 1.719 1.000
Chain 1: 500 -8772.949 1.419 0.367
Chain 1: 600 -8319.479 1.192 0.367
Chain 1: 700 -8328.522 1.021 0.219
Chain 1: 800 -8610.641 0.898 0.219
Chain 1: 900 -8644.353 0.799 0.083
Chain 1: 1000 -8382.794 0.722 0.083
Chain 1: 1100 -8654.049 0.625 0.055 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8300.518 0.087 0.043
Chain 1: 1300 -8525.928 0.053 0.033
Chain 1: 1400 -8523.496 0.044 0.031
Chain 1: 1500 -8420.710 0.024 0.031
Chain 1: 1600 -8521.141 0.019 0.026
Chain 1: 1700 -8610.393 0.020 0.026
Chain 1: 1800 -8209.895 0.022 0.026
Chain 1: 1900 -8310.568 0.023 0.026
Chain 1: 2000 -8281.446 0.020 0.012
Chain 1: 2100 -8401.871 0.018 0.012
Chain 1: 2200 -8178.473 0.017 0.012
Chain 1: 2300 -8339.785 0.016 0.012
Chain 1: 2400 -8221.610 0.017 0.014
Chain 1: 2500 -8285.781 0.017 0.014
Chain 1: 2600 -8306.661 0.016 0.014
Chain 1: 2700 -8226.224 0.016 0.014
Chain 1: 2800 -8201.001 0.011 0.012
Chain 1: 2900 -8255.664 0.011 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002502 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.02 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8401498.896 1.000 1.000
Chain 1: 200 -1585168.770 2.650 4.300
Chain 1: 300 -891472.746 2.026 1.000
Chain 1: 400 -457640.044 1.757 1.000
Chain 1: 500 -358008.454 1.461 0.948
Chain 1: 600 -232929.064 1.307 0.948
Chain 1: 700 -119144.781 1.257 0.948
Chain 1: 800 -86337.371 1.147 0.948
Chain 1: 900 -66680.133 1.052 0.778
Chain 1: 1000 -51471.282 0.977 0.778
Chain 1: 1100 -38944.664 0.909 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38117.549 0.481 0.380
Chain 1: 1300 -26077.664 0.449 0.380
Chain 1: 1400 -25795.283 0.356 0.322
Chain 1: 1500 -22383.290 0.343 0.322
Chain 1: 1600 -21599.375 0.293 0.295
Chain 1: 1700 -20474.125 0.203 0.295
Chain 1: 1800 -20418.210 0.165 0.152
Chain 1: 1900 -20744.058 0.137 0.055
Chain 1: 2000 -19256.016 0.116 0.055
Chain 1: 2100 -19494.461 0.085 0.036
Chain 1: 2200 -19720.603 0.084 0.036
Chain 1: 2300 -19338.116 0.039 0.020
Chain 1: 2400 -19110.319 0.039 0.020
Chain 1: 2500 -18912.251 0.025 0.016
Chain 1: 2600 -18542.920 0.024 0.016
Chain 1: 2700 -18499.950 0.018 0.012
Chain 1: 2800 -18216.942 0.020 0.016
Chain 1: 2900 -18497.986 0.020 0.015
Chain 1: 3000 -18484.267 0.012 0.012
Chain 1: 3100 -18569.227 0.011 0.012
Chain 1: 3200 -18260.129 0.012 0.015
Chain 1: 3300 -18464.633 0.011 0.012
Chain 1: 3400 -17939.961 0.013 0.015
Chain 1: 3500 -18551.270 0.015 0.016
Chain 1: 3600 -17858.638 0.017 0.016
Chain 1: 3700 -18244.939 0.019 0.017
Chain 1: 3800 -17205.768 0.023 0.021
Chain 1: 3900 -17201.909 0.022 0.021
Chain 1: 4000 -17319.216 0.022 0.021
Chain 1: 4100 -17233.075 0.022 0.021
Chain 1: 4200 -17049.518 0.022 0.021
Chain 1: 4300 -17187.790 0.021 0.021
Chain 1: 4400 -17144.822 0.019 0.011
Chain 1: 4500 -17047.359 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001298 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.98 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12489.493 1.000 1.000
Chain 1: 200 -9408.082 0.664 1.000
Chain 1: 300 -8127.255 0.495 0.328
Chain 1: 400 -8215.479 0.374 0.328
Chain 1: 500 -8170.172 0.300 0.158
Chain 1: 600 -7996.075 0.254 0.158
Chain 1: 700 -7958.577 0.218 0.022
Chain 1: 800 -7970.566 0.191 0.022
Chain 1: 900 -7939.334 0.170 0.011
Chain 1: 1000 -8028.540 0.154 0.011
Chain 1: 1100 -8080.497 0.055 0.011
Chain 1: 1200 -7975.976 0.024 0.011
Chain 1: 1300 -7927.229 0.009 0.006 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001389 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.89 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58157.192 1.000 1.000
Chain 1: 200 -17734.061 1.640 2.279
Chain 1: 300 -8739.955 1.436 1.029
Chain 1: 400 -8254.674 1.092 1.029
Chain 1: 500 -8417.031 0.877 1.000
Chain 1: 600 -8414.463 0.731 1.000
Chain 1: 700 -7982.283 0.634 0.059
Chain 1: 800 -8235.476 0.559 0.059
Chain 1: 900 -7935.551 0.501 0.054
Chain 1: 1000 -7876.753 0.452 0.054
Chain 1: 1100 -7793.824 0.353 0.038
Chain 1: 1200 -7622.670 0.127 0.031
Chain 1: 1300 -7803.949 0.026 0.023
Chain 1: 1400 -7902.715 0.022 0.022
Chain 1: 1500 -7668.178 0.023 0.023
Chain 1: 1600 -7802.006 0.025 0.023
Chain 1: 1700 -7563.877 0.022 0.023
Chain 1: 1800 -7660.411 0.021 0.022
Chain 1: 1900 -7603.815 0.018 0.017
Chain 1: 2000 -7679.054 0.018 0.017
Chain 1: 2100 -7697.378 0.017 0.017
Chain 1: 2200 -7755.938 0.015 0.013
Chain 1: 2300 -7655.004 0.014 0.013
Chain 1: 2400 -7712.634 0.014 0.013
Chain 1: 2500 -7628.735 0.012 0.011
Chain 1: 2600 -7605.713 0.011 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003137 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.37 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86356.703 1.000 1.000
Chain 1: 200 -13555.545 3.185 5.371
Chain 1: 300 -9931.779 2.245 1.000
Chain 1: 400 -10809.353 1.704 1.000
Chain 1: 500 -8898.056 1.406 0.365
Chain 1: 600 -8362.434 1.183 0.365
Chain 1: 700 -8531.061 1.016 0.215
Chain 1: 800 -9217.211 0.899 0.215
Chain 1: 900 -8727.570 0.805 0.081
Chain 1: 1000 -8558.694 0.727 0.081
Chain 1: 1100 -8692.233 0.628 0.074 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8464.301 0.094 0.064
Chain 1: 1300 -8574.710 0.059 0.056
Chain 1: 1400 -8609.991 0.051 0.027
Chain 1: 1500 -8501.822 0.031 0.020
Chain 1: 1600 -8608.078 0.025 0.020
Chain 1: 1700 -8696.033 0.024 0.015
Chain 1: 1800 -8286.299 0.022 0.015
Chain 1: 1900 -8382.234 0.018 0.013
Chain 1: 2000 -8354.936 0.016 0.013
Chain 1: 2100 -8476.626 0.016 0.013
Chain 1: 2200 -8318.565 0.015 0.013
Chain 1: 2300 -8379.675 0.014 0.012
Chain 1: 2400 -8446.849 0.015 0.012
Chain 1: 2500 -8392.670 0.014 0.011
Chain 1: 2600 -8390.976 0.013 0.010
Chain 1: 2700 -8308.175 0.013 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003406 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.06 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8408319.755 1.000 1.000
Chain 1: 200 -1584308.905 2.654 4.307
Chain 1: 300 -890246.392 2.029 1.000
Chain 1: 400 -457715.296 1.758 1.000
Chain 1: 500 -358106.395 1.462 0.945
Chain 1: 600 -233146.936 1.308 0.945
Chain 1: 700 -119311.867 1.257 0.945
Chain 1: 800 -86512.196 1.147 0.945
Chain 1: 900 -66845.520 1.053 0.780
Chain 1: 1000 -51635.786 0.977 0.780
Chain 1: 1100 -39109.234 0.909 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38283.129 0.480 0.379
Chain 1: 1300 -26239.512 0.448 0.379
Chain 1: 1400 -25957.466 0.355 0.320
Chain 1: 1500 -22545.522 0.342 0.320
Chain 1: 1600 -21761.954 0.292 0.295
Chain 1: 1700 -20635.908 0.202 0.294
Chain 1: 1800 -20580.007 0.165 0.151
Chain 1: 1900 -20906.071 0.137 0.055
Chain 1: 2000 -19417.489 0.115 0.055
Chain 1: 2100 -19655.806 0.084 0.036
Chain 1: 2200 -19882.279 0.083 0.036
Chain 1: 2300 -19499.470 0.039 0.020
Chain 1: 2400 -19271.616 0.039 0.020
Chain 1: 2500 -19073.657 0.025 0.016
Chain 1: 2600 -18703.973 0.023 0.016
Chain 1: 2700 -18660.869 0.018 0.012
Chain 1: 2800 -18377.855 0.020 0.015
Chain 1: 2900 -18659.029 0.019 0.015
Chain 1: 3000 -18645.181 0.012 0.012
Chain 1: 3100 -18730.214 0.011 0.012
Chain 1: 3200 -18420.926 0.012 0.015
Chain 1: 3300 -18625.578 0.011 0.012
Chain 1: 3400 -18100.657 0.012 0.015
Chain 1: 3500 -18712.333 0.015 0.015
Chain 1: 3600 -18019.191 0.017 0.015
Chain 1: 3700 -18405.942 0.018 0.017
Chain 1: 3800 -17365.950 0.023 0.021
Chain 1: 3900 -17362.087 0.021 0.021
Chain 1: 4000 -17479.392 0.022 0.021
Chain 1: 4100 -17393.234 0.022 0.021
Chain 1: 4200 -17209.474 0.021 0.021
Chain 1: 4300 -17347.866 0.021 0.021
Chain 1: 4400 -17304.751 0.019 0.011
Chain 1: 4500 -17207.284 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001402 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.02 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49170.424 1.000 1.000
Chain 1: 200 -23141.351 1.062 1.125
Chain 1: 300 -19074.570 0.779 1.000
Chain 1: 400 -13826.679 0.679 1.000
Chain 1: 500 -22460.984 0.620 0.384
Chain 1: 600 -12960.242 0.639 0.733
Chain 1: 700 -15121.171 0.568 0.384
Chain 1: 800 -14662.266 0.501 0.384
Chain 1: 900 -14277.636 0.448 0.380
Chain 1: 1000 -12447.289 0.418 0.380
Chain 1: 1100 -11173.756 0.330 0.213
Chain 1: 1200 -11457.764 0.220 0.147
Chain 1: 1300 -13966.041 0.216 0.147
Chain 1: 1400 -15029.777 0.185 0.143
Chain 1: 1500 -11222.506 0.181 0.143
Chain 1: 1600 -12878.006 0.121 0.129
Chain 1: 1700 -9711.591 0.139 0.129
Chain 1: 1800 -10508.780 0.143 0.129
Chain 1: 1900 -11188.017 0.147 0.129
Chain 1: 2000 -17947.709 0.170 0.129
Chain 1: 2100 -16577.854 0.166 0.129
Chain 1: 2200 -10491.406 0.222 0.180
Chain 1: 2300 -10254.279 0.206 0.129
Chain 1: 2400 -9423.568 0.208 0.129
Chain 1: 2500 -9568.771 0.176 0.088
Chain 1: 2600 -9856.584 0.166 0.083
Chain 1: 2700 -11501.762 0.147 0.083
Chain 1: 2800 -10594.134 0.148 0.086
Chain 1: 2900 -10019.360 0.148 0.086
Chain 1: 3000 -10145.046 0.112 0.083
Chain 1: 3100 -15934.456 0.140 0.086
Chain 1: 3200 -10236.747 0.137 0.086
Chain 1: 3300 -10006.838 0.137 0.086
Chain 1: 3400 -9487.966 0.134 0.057
Chain 1: 3500 -9479.075 0.133 0.057
Chain 1: 3600 -18162.593 0.178 0.086
Chain 1: 3700 -14633.113 0.187 0.086
Chain 1: 3800 -13039.691 0.191 0.122
Chain 1: 3900 -9718.932 0.219 0.241
Chain 1: 4000 -9041.745 0.226 0.241
Chain 1: 4100 -9413.575 0.193 0.122
Chain 1: 4200 -17357.337 0.183 0.122
Chain 1: 4300 -8926.531 0.276 0.241
Chain 1: 4400 -9400.725 0.275 0.241
Chain 1: 4500 -9033.292 0.279 0.241
Chain 1: 4600 -9452.456 0.236 0.122
Chain 1: 4700 -8725.596 0.220 0.083
Chain 1: 4800 -9033.254 0.211 0.075
Chain 1: 4900 -9174.260 0.178 0.050
Chain 1: 5000 -9844.616 0.178 0.050
Chain 1: 5100 -8945.072 0.184 0.068
Chain 1: 5200 -10333.437 0.152 0.068
Chain 1: 5300 -13829.220 0.082 0.068
Chain 1: 5400 -9271.468 0.127 0.083
Chain 1: 5500 -14377.570 0.158 0.101
Chain 1: 5600 -9407.279 0.206 0.134
Chain 1: 5700 -13460.224 0.228 0.253
Chain 1: 5800 -9273.256 0.270 0.301
Chain 1: 5900 -9252.988 0.269 0.301
Chain 1: 6000 -11001.601 0.278 0.301
Chain 1: 6100 -8815.085 0.292 0.301
Chain 1: 6200 -11673.391 0.303 0.301
Chain 1: 6300 -12799.629 0.287 0.301
Chain 1: 6400 -9207.419 0.277 0.301
Chain 1: 6500 -9162.816 0.242 0.248
Chain 1: 6600 -8705.645 0.194 0.245
Chain 1: 6700 -9380.950 0.171 0.159
Chain 1: 6800 -8518.139 0.136 0.101
Chain 1: 6900 -12514.387 0.168 0.159
Chain 1: 7000 -9151.902 0.189 0.245
Chain 1: 7100 -12521.113 0.191 0.245
Chain 1: 7200 -11915.140 0.172 0.101
Chain 1: 7300 -10882.767 0.172 0.101
Chain 1: 7400 -9215.122 0.151 0.101
Chain 1: 7500 -9031.919 0.153 0.101
Chain 1: 7600 -13365.320 0.180 0.181
Chain 1: 7700 -8522.577 0.230 0.269
Chain 1: 7800 -11662.122 0.246 0.269
Chain 1: 7900 -10110.504 0.230 0.269
Chain 1: 8000 -8828.006 0.208 0.181
Chain 1: 8100 -11938.673 0.207 0.181
Chain 1: 8200 -10053.894 0.220 0.187
Chain 1: 8300 -8536.637 0.229 0.187
Chain 1: 8400 -8561.253 0.211 0.187
Chain 1: 8500 -8479.467 0.210 0.187
Chain 1: 8600 -8834.450 0.181 0.178
Chain 1: 8700 -8653.242 0.127 0.153
Chain 1: 8800 -12450.242 0.130 0.153
Chain 1: 8900 -10218.730 0.137 0.178
Chain 1: 9000 -9462.246 0.130 0.178
Chain 1: 9100 -8652.458 0.114 0.094
Chain 1: 9200 -8651.066 0.095 0.080
Chain 1: 9300 -8479.938 0.079 0.040
Chain 1: 9400 -8553.557 0.080 0.040
Chain 1: 9500 -8454.555 0.080 0.040
Chain 1: 9600 -10367.225 0.094 0.080
Chain 1: 9700 -8348.894 0.116 0.094
Chain 1: 9800 -12444.263 0.119 0.094
Chain 1: 9900 -8649.865 0.141 0.094
Chain 1: 10000 -8755.596 0.134 0.094
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001396 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.96 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57635.895 1.000 1.000
Chain 1: 200 -17863.942 1.613 2.226
Chain 1: 300 -8929.563 1.409 1.001
Chain 1: 400 -8256.226 1.077 1.001
Chain 1: 500 -9152.620 0.881 1.000
Chain 1: 600 -8471.122 0.748 1.000
Chain 1: 700 -8144.853 0.647 0.098
Chain 1: 800 -8724.316 0.574 0.098
Chain 1: 900 -8130.316 0.518 0.082
Chain 1: 1000 -7735.779 0.472 0.082
Chain 1: 1100 -7693.837 0.372 0.080
Chain 1: 1200 -7649.363 0.150 0.073
Chain 1: 1300 -7659.251 0.050 0.066
Chain 1: 1400 -8073.353 0.047 0.051
Chain 1: 1500 -7667.455 0.043 0.051
Chain 1: 1600 -7871.211 0.037 0.051
Chain 1: 1700 -7625.265 0.037 0.051
Chain 1: 1800 -7673.155 0.031 0.032
Chain 1: 1900 -7675.605 0.023 0.026
Chain 1: 2000 -7741.894 0.019 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002633 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 26.33 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85509.757 1.000 1.000
Chain 1: 200 -13871.791 3.082 5.164
Chain 1: 300 -10189.562 2.175 1.000
Chain 1: 400 -11226.587 1.655 1.000
Chain 1: 500 -9003.202 1.373 0.361
Chain 1: 600 -9515.329 1.153 0.361
Chain 1: 700 -8649.620 1.003 0.247
Chain 1: 800 -9406.242 0.887 0.247
Chain 1: 900 -8919.107 0.795 0.100
Chain 1: 1000 -9029.956 0.717 0.100
Chain 1: 1100 -8764.039 0.620 0.092 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8554.043 0.106 0.080
Chain 1: 1300 -8862.335 0.073 0.055
Chain 1: 1400 -8815.503 0.064 0.054
Chain 1: 1500 -8700.229 0.041 0.035
Chain 1: 1600 -8807.158 0.037 0.030
Chain 1: 1700 -8879.128 0.028 0.025
Chain 1: 1800 -8445.322 0.025 0.025
Chain 1: 1900 -8549.578 0.020 0.013
Chain 1: 2000 -8525.070 0.019 0.013
Chain 1: 2100 -8473.176 0.017 0.012
Chain 1: 2200 -8467.135 0.015 0.012
Chain 1: 2300 -8603.695 0.013 0.012
Chain 1: 2400 -8450.954 0.014 0.012
Chain 1: 2500 -8519.792 0.014 0.012
Chain 1: 2600 -8438.614 0.013 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002565 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.65 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8376427.667 1.000 1.000
Chain 1: 200 -1581074.100 2.649 4.298
Chain 1: 300 -891525.243 2.024 1.000
Chain 1: 400 -458619.648 1.754 1.000
Chain 1: 500 -359451.049 1.458 0.944
Chain 1: 600 -234344.331 1.304 0.944
Chain 1: 700 -120080.868 1.254 0.944
Chain 1: 800 -87213.384 1.144 0.944
Chain 1: 900 -67457.062 1.050 0.773
Chain 1: 1000 -52190.440 0.974 0.773
Chain 1: 1100 -39601.975 0.906 0.534 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38771.379 0.478 0.377
Chain 1: 1300 -26648.169 0.446 0.377
Chain 1: 1400 -26361.813 0.353 0.318
Chain 1: 1500 -22928.677 0.340 0.318
Chain 1: 1600 -22139.849 0.290 0.293
Chain 1: 1700 -21003.260 0.201 0.293
Chain 1: 1800 -20945.290 0.163 0.150
Chain 1: 1900 -21271.772 0.136 0.054
Chain 1: 2000 -19776.985 0.114 0.054
Chain 1: 2100 -20015.620 0.083 0.036
Chain 1: 2200 -20243.324 0.082 0.036
Chain 1: 2300 -19859.332 0.039 0.019
Chain 1: 2400 -19631.161 0.039 0.019
Chain 1: 2500 -19433.680 0.025 0.015
Chain 1: 2600 -19063.089 0.023 0.015
Chain 1: 2700 -19019.772 0.018 0.012
Chain 1: 2800 -18736.712 0.019 0.015
Chain 1: 2900 -19018.182 0.019 0.015
Chain 1: 3000 -19004.270 0.012 0.012
Chain 1: 3100 -19089.370 0.011 0.012
Chain 1: 3200 -18779.708 0.011 0.015
Chain 1: 3300 -18984.665 0.011 0.012
Chain 1: 3400 -18459.205 0.012 0.015
Chain 1: 3500 -19071.890 0.014 0.015
Chain 1: 3600 -18377.490 0.016 0.015
Chain 1: 3700 -18765.160 0.018 0.016
Chain 1: 3800 -17723.402 0.023 0.021
Chain 1: 3900 -17719.567 0.021 0.021
Chain 1: 4000 -17836.792 0.022 0.021
Chain 1: 4100 -17750.561 0.022 0.021
Chain 1: 4200 -17566.440 0.021 0.021
Chain 1: 4300 -17705.035 0.021 0.021
Chain 1: 4400 -17661.566 0.018 0.010
Chain 1: 4500 -17564.112 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001377 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.77 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13153.057 1.000 1.000
Chain 1: 200 -9534.503 0.690 1.000
Chain 1: 300 -8245.000 0.512 0.380
Chain 1: 400 -8291.430 0.385 0.380
Chain 1: 500 -8097.802 0.313 0.156
Chain 1: 600 -8195.840 0.263 0.156
Chain 1: 700 -8063.396 0.228 0.024
Chain 1: 800 -8099.588 0.200 0.024
Chain 1: 900 -8171.017 0.179 0.016
Chain 1: 1000 -8180.432 0.161 0.016
Chain 1: 1100 -8127.269 0.061 0.012
Chain 1: 1200 -8067.946 0.024 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001448 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.48 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -50778.838 1.000 1.000
Chain 1: 200 -16913.600 1.501 2.002
Chain 1: 300 -9104.612 1.287 1.000
Chain 1: 400 -8534.510 0.982 1.000
Chain 1: 500 -8879.343 0.793 0.858
Chain 1: 600 -8505.505 0.668 0.858
Chain 1: 700 -8415.489 0.574 0.067
Chain 1: 800 -8434.991 0.503 0.067
Chain 1: 900 -7855.284 0.455 0.067
Chain 1: 1000 -7749.259 0.411 0.067
Chain 1: 1100 -7909.568 0.313 0.044
Chain 1: 1200 -7696.489 0.116 0.039
Chain 1: 1300 -7579.238 0.031 0.028
Chain 1: 1400 -8378.550 0.034 0.028
Chain 1: 1500 -7525.162 0.042 0.028
Chain 1: 1600 -7850.850 0.041 0.028
Chain 1: 1700 -7744.553 0.042 0.028
Chain 1: 1800 -7536.806 0.044 0.028
Chain 1: 1900 -7561.029 0.037 0.028
Chain 1: 2000 -7602.438 0.036 0.028
Chain 1: 2100 -7542.126 0.035 0.028
Chain 1: 2200 -7835.550 0.036 0.028
Chain 1: 2300 -7630.390 0.037 0.028
Chain 1: 2400 -7641.807 0.028 0.027
Chain 1: 2500 -7544.246 0.018 0.014
Chain 1: 2600 -7492.681 0.014 0.013
Chain 1: 2700 -7432.884 0.014 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002933 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.33 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86724.095 1.000 1.000
Chain 1: 200 -14233.281 3.047 5.093
Chain 1: 300 -10363.867 2.155 1.000
Chain 1: 400 -12537.183 1.660 1.000
Chain 1: 500 -8851.467 1.411 0.416
Chain 1: 600 -8646.792 1.180 0.416
Chain 1: 700 -8844.563 1.015 0.373
Chain 1: 800 -9825.750 0.900 0.373
Chain 1: 900 -9152.122 0.808 0.173
Chain 1: 1000 -8825.392 0.731 0.173
Chain 1: 1100 -9096.607 0.634 0.100 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8672.342 0.130 0.074
Chain 1: 1300 -8881.707 0.095 0.049
Chain 1: 1400 -8774.702 0.079 0.037
Chain 1: 1500 -8787.510 0.037 0.030
Chain 1: 1600 -8856.346 0.036 0.030
Chain 1: 1700 -8896.754 0.034 0.030
Chain 1: 1800 -8419.186 0.030 0.030
Chain 1: 1900 -8537.813 0.024 0.024
Chain 1: 2000 -8547.245 0.020 0.014
Chain 1: 2100 -8666.252 0.018 0.014
Chain 1: 2200 -8409.006 0.017 0.014
Chain 1: 2300 -8501.076 0.015 0.012
Chain 1: 2400 -8589.432 0.015 0.011
Chain 1: 2500 -8507.216 0.016 0.011
Chain 1: 2600 -8534.103 0.015 0.011
Chain 1: 2700 -8447.212 0.016 0.011
Chain 1: 2800 -8411.343 0.011 0.010
Chain 1: 2900 -8507.036 0.011 0.010
Chain 1: 3000 -8426.435 0.011 0.010
Chain 1: 3100 -8385.204 0.010 0.010
Chain 1: 3200 -8348.824 0.008 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003496 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.96 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8412302.809 1.000 1.000
Chain 1: 200 -1586491.381 2.651 4.302
Chain 1: 300 -892919.302 2.026 1.000
Chain 1: 400 -458837.836 1.756 1.000
Chain 1: 500 -359003.345 1.461 0.946
Chain 1: 600 -233783.655 1.306 0.946
Chain 1: 700 -120026.429 1.255 0.946
Chain 1: 800 -87219.970 1.145 0.946
Chain 1: 900 -67570.130 1.050 0.777
Chain 1: 1000 -52389.930 0.974 0.777
Chain 1: 1100 -39873.829 0.906 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39064.343 0.478 0.376
Chain 1: 1300 -27007.980 0.445 0.376
Chain 1: 1400 -26731.996 0.351 0.314
Chain 1: 1500 -23314.372 0.338 0.314
Chain 1: 1600 -22531.449 0.288 0.291
Chain 1: 1700 -21402.468 0.198 0.290
Chain 1: 1800 -21346.901 0.161 0.147
Chain 1: 1900 -21674.230 0.133 0.053
Chain 1: 2000 -20181.684 0.112 0.053
Chain 1: 2100 -20420.435 0.081 0.035
Chain 1: 2200 -20647.820 0.081 0.035
Chain 1: 2300 -20263.926 0.038 0.019
Chain 1: 2400 -20035.551 0.038 0.019
Chain 1: 2500 -19837.503 0.024 0.015
Chain 1: 2600 -19466.486 0.023 0.015
Chain 1: 2700 -19423.166 0.018 0.012
Chain 1: 2800 -19139.365 0.019 0.015
Chain 1: 2900 -19421.271 0.019 0.015
Chain 1: 3000 -19407.388 0.011 0.012
Chain 1: 3100 -19492.522 0.011 0.011
Chain 1: 3200 -19182.400 0.011 0.015
Chain 1: 3300 -19387.811 0.010 0.011
Chain 1: 3400 -18861.195 0.012 0.015
Chain 1: 3500 -19475.261 0.014 0.015
Chain 1: 3600 -18779.166 0.016 0.015
Chain 1: 3700 -19168.000 0.018 0.016
Chain 1: 3800 -18123.282 0.022 0.020
Chain 1: 3900 -18119.309 0.021 0.020
Chain 1: 4000 -18236.656 0.021 0.020
Chain 1: 4100 -18150.101 0.021 0.020
Chain 1: 4200 -17965.449 0.021 0.020
Chain 1: 4300 -18104.500 0.020 0.020
Chain 1: 4400 -18060.546 0.018 0.010
Chain 1: 4500 -17962.919 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001177 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 11.77 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48771.228 1.000 1.000
Chain 1: 200 -22963.585 1.062 1.124
Chain 1: 300 -13191.593 0.955 1.000
Chain 1: 400 -12679.817 0.726 1.000
Chain 1: 500 -20298.171 0.656 0.741
Chain 1: 600 -11934.648 0.664 0.741
Chain 1: 700 -11968.889 0.569 0.701
Chain 1: 800 -13951.366 0.516 0.701
Chain 1: 900 -12363.471 0.473 0.375
Chain 1: 1000 -30316.985 0.485 0.592
Chain 1: 1100 -25885.975 0.402 0.375
Chain 1: 1200 -12969.316 0.389 0.375
Chain 1: 1300 -12989.764 0.315 0.171
Chain 1: 1400 -16708.132 0.333 0.223
Chain 1: 1500 -12122.657 0.334 0.223
Chain 1: 1600 -12188.645 0.264 0.171
Chain 1: 1700 -14892.832 0.282 0.182
Chain 1: 1800 -10092.776 0.315 0.223
Chain 1: 1900 -9582.696 0.308 0.223
Chain 1: 2000 -9510.916 0.249 0.182
Chain 1: 2100 -10802.787 0.244 0.182
Chain 1: 2200 -9497.670 0.158 0.137
Chain 1: 2300 -15482.199 0.197 0.182
Chain 1: 2400 -9322.010 0.241 0.182
Chain 1: 2500 -9129.714 0.205 0.137
Chain 1: 2600 -9867.447 0.212 0.137
Chain 1: 2700 -9576.567 0.197 0.120
Chain 1: 2800 -9631.253 0.150 0.075
Chain 1: 2900 -11985.622 0.164 0.120
Chain 1: 3000 -9068.747 0.195 0.137
Chain 1: 3100 -9374.778 0.187 0.137
Chain 1: 3200 -9176.148 0.175 0.075
Chain 1: 3300 -9055.321 0.138 0.033
Chain 1: 3400 -14953.027 0.111 0.033
Chain 1: 3500 -10755.074 0.148 0.075
Chain 1: 3600 -9473.394 0.154 0.135
Chain 1: 3700 -8725.370 0.160 0.135
Chain 1: 3800 -9039.745 0.163 0.135
Chain 1: 3900 -11666.398 0.165 0.135
Chain 1: 4000 -8626.074 0.169 0.135
Chain 1: 4100 -8955.851 0.169 0.135
Chain 1: 4200 -11209.462 0.187 0.201
Chain 1: 4300 -12260.378 0.194 0.201
Chain 1: 4400 -12209.164 0.155 0.135
Chain 1: 4500 -8840.296 0.154 0.135
Chain 1: 4600 -11750.697 0.165 0.201
Chain 1: 4700 -14159.375 0.174 0.201
Chain 1: 4800 -8753.763 0.232 0.225
Chain 1: 4900 -9214.565 0.215 0.201
Chain 1: 5000 -9497.614 0.182 0.170
Chain 1: 5100 -11345.672 0.195 0.170
Chain 1: 5200 -16445.022 0.206 0.170
Chain 1: 5300 -10827.887 0.249 0.248
Chain 1: 5400 -13608.703 0.269 0.248
Chain 1: 5500 -8835.133 0.285 0.248
Chain 1: 5600 -8403.321 0.266 0.204
Chain 1: 5700 -8989.808 0.255 0.204
Chain 1: 5800 -8518.874 0.199 0.163
Chain 1: 5900 -9019.170 0.199 0.163
Chain 1: 6000 -8847.419 0.198 0.163
Chain 1: 6100 -8375.840 0.188 0.065
Chain 1: 6200 -8126.683 0.160 0.056
Chain 1: 6300 -15454.686 0.155 0.056
Chain 1: 6400 -13921.178 0.146 0.056
Chain 1: 6500 -9766.471 0.134 0.056
Chain 1: 6600 -8259.381 0.147 0.065
Chain 1: 6700 -10375.467 0.161 0.110
Chain 1: 6800 -9866.262 0.161 0.110
Chain 1: 6900 -12269.873 0.175 0.182
Chain 1: 7000 -8184.371 0.223 0.196
Chain 1: 7100 -8933.344 0.226 0.196
Chain 1: 7200 -8393.939 0.229 0.196
Chain 1: 7300 -9697.750 0.195 0.182
Chain 1: 7400 -12540.648 0.207 0.196
Chain 1: 7500 -10817.611 0.180 0.182
Chain 1: 7600 -8735.053 0.186 0.196
Chain 1: 7700 -8223.874 0.172 0.159
Chain 1: 7800 -10692.395 0.190 0.196
Chain 1: 7900 -8207.075 0.200 0.227
Chain 1: 8000 -8269.022 0.151 0.159
Chain 1: 8100 -8850.813 0.149 0.159
Chain 1: 8200 -8673.872 0.145 0.159
Chain 1: 8300 -8165.217 0.138 0.159
Chain 1: 8400 -9524.067 0.129 0.143
Chain 1: 8500 -10392.646 0.122 0.084
Chain 1: 8600 -11446.910 0.107 0.084
Chain 1: 8700 -7997.722 0.144 0.092
Chain 1: 8800 -8115.009 0.122 0.084
Chain 1: 8900 -8515.601 0.097 0.066
Chain 1: 9000 -9364.592 0.105 0.084
Chain 1: 9100 -7989.732 0.116 0.091
Chain 1: 9200 -10141.490 0.135 0.092
Chain 1: 9300 -8576.405 0.147 0.143
Chain 1: 9400 -7957.572 0.140 0.092
Chain 1: 9500 -8085.213 0.134 0.092
Chain 1: 9600 -8194.210 0.126 0.091
Chain 1: 9700 -8405.921 0.085 0.078
Chain 1: 9800 -8465.194 0.084 0.078
Chain 1: 9900 -10822.221 0.101 0.091
Chain 1: 10000 -8330.855 0.122 0.172
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001377 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.77 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58147.838 1.000 1.000
Chain 1: 200 -17623.264 1.650 2.299
Chain 1: 300 -8609.952 1.449 1.047
Chain 1: 400 -8171.981 1.100 1.047
Chain 1: 500 -8314.276 0.883 1.000
Chain 1: 600 -8350.595 0.737 1.000
Chain 1: 700 -7875.625 0.640 0.060
Chain 1: 800 -8131.068 0.564 0.060
Chain 1: 900 -7745.866 0.507 0.054
Chain 1: 1000 -7874.924 0.458 0.054
Chain 1: 1100 -7678.941 0.360 0.050
Chain 1: 1200 -7609.698 0.131 0.031
Chain 1: 1300 -7602.171 0.027 0.026
Chain 1: 1400 -7577.129 0.022 0.017
Chain 1: 1500 -7542.747 0.021 0.016
Chain 1: 1600 -7716.602 0.022 0.023
Chain 1: 1700 -7456.393 0.020 0.023
Chain 1: 1800 -7592.593 0.018 0.018
Chain 1: 1900 -7570.301 0.014 0.016
Chain 1: 2000 -7589.905 0.012 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003046 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.46 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86276.329 1.000 1.000
Chain 1: 200 -13386.319 3.223 5.445
Chain 1: 300 -9762.832 2.272 1.000
Chain 1: 400 -10863.312 1.729 1.000
Chain 1: 500 -8643.982 1.435 0.371
Chain 1: 600 -8393.532 1.201 0.371
Chain 1: 700 -8483.942 1.031 0.257
Chain 1: 800 -9130.479 0.911 0.257
Chain 1: 900 -8557.397 0.817 0.101
Chain 1: 1000 -8416.516 0.737 0.101
Chain 1: 1100 -8546.580 0.638 0.071 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8124.558 0.099 0.067
Chain 1: 1300 -8445.824 0.066 0.052
Chain 1: 1400 -8476.258 0.056 0.038
Chain 1: 1500 -8330.731 0.032 0.030
Chain 1: 1600 -8449.781 0.031 0.017
Chain 1: 1700 -8531.230 0.030 0.017
Chain 1: 1800 -8121.344 0.028 0.017
Chain 1: 1900 -8217.269 0.023 0.017
Chain 1: 2000 -8190.087 0.022 0.015
Chain 1: 2100 -8311.878 0.021 0.015
Chain 1: 2200 -8152.869 0.018 0.015
Chain 1: 2300 -8214.813 0.015 0.014
Chain 1: 2400 -8281.999 0.016 0.014
Chain 1: 2500 -8227.803 0.015 0.012
Chain 1: 2600 -8226.127 0.013 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004359 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 43.59 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8404817.796 1.000 1.000
Chain 1: 200 -1585003.568 2.651 4.303
Chain 1: 300 -890247.601 2.028 1.000
Chain 1: 400 -457461.986 1.757 1.000
Chain 1: 500 -357752.560 1.462 0.946
Chain 1: 600 -232826.297 1.307 0.946
Chain 1: 700 -119080.393 1.257 0.946
Chain 1: 800 -86295.765 1.147 0.946
Chain 1: 900 -66647.382 1.053 0.780
Chain 1: 1000 -51447.664 0.977 0.780
Chain 1: 1100 -38928.906 0.909 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38105.328 0.481 0.380
Chain 1: 1300 -26069.779 0.449 0.380
Chain 1: 1400 -25788.982 0.356 0.322
Chain 1: 1500 -22378.454 0.343 0.322
Chain 1: 1600 -21595.401 0.293 0.295
Chain 1: 1700 -20470.227 0.203 0.295
Chain 1: 1800 -20414.653 0.165 0.152
Chain 1: 1900 -20740.692 0.137 0.055
Chain 1: 2000 -19252.518 0.116 0.055
Chain 1: 2100 -19490.934 0.085 0.036
Chain 1: 2200 -19717.212 0.084 0.036
Chain 1: 2300 -19334.574 0.039 0.020
Chain 1: 2400 -19106.705 0.039 0.020
Chain 1: 2500 -18908.671 0.025 0.016
Chain 1: 2600 -18539.038 0.024 0.016
Chain 1: 2700 -18496.023 0.018 0.012
Chain 1: 2800 -18212.895 0.020 0.016
Chain 1: 2900 -18494.106 0.020 0.015
Chain 1: 3000 -18480.298 0.012 0.012
Chain 1: 3100 -18565.289 0.011 0.012
Chain 1: 3200 -18256.044 0.012 0.015
Chain 1: 3300 -18460.702 0.011 0.012
Chain 1: 3400 -17935.748 0.013 0.015
Chain 1: 3500 -18547.436 0.015 0.016
Chain 1: 3600 -17854.342 0.017 0.016
Chain 1: 3700 -18240.977 0.019 0.017
Chain 1: 3800 -17201.039 0.023 0.021
Chain 1: 3900 -17197.179 0.022 0.021
Chain 1: 4000 -17314.488 0.022 0.021
Chain 1: 4100 -17228.265 0.022 0.021
Chain 1: 4200 -17044.577 0.022 0.021
Chain 1: 4300 -17182.938 0.021 0.021
Chain 1: 4400 -17139.831 0.019 0.011
Chain 1: 4500 -17042.354 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001264 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.64 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12193.357 1.000 1.000
Chain 1: 200 -9088.474 0.671 1.000
Chain 1: 300 -7911.045 0.497 0.342
Chain 1: 400 -8007.452 0.376 0.342
Chain 1: 500 -7875.838 0.304 0.149
Chain 1: 600 -7790.141 0.255 0.149
Chain 1: 700 -7704.158 0.220 0.017
Chain 1: 800 -7712.126 0.193 0.017
Chain 1: 900 -7613.199 0.173 0.013
Chain 1: 1000 -7756.116 0.157 0.017
Chain 1: 1100 -7718.610 0.058 0.013
Chain 1: 1200 -7743.921 0.024 0.012
Chain 1: 1300 -7668.955 0.010 0.011
Chain 1: 1400 -7696.400 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001417 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.17 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49126.441 1.000 1.000
Chain 1: 200 -15717.778 1.563 2.126
Chain 1: 300 -8585.804 1.319 1.000
Chain 1: 400 -8316.164 0.997 1.000
Chain 1: 500 -7830.728 0.810 0.831
Chain 1: 600 -8932.462 0.696 0.831
Chain 1: 700 -8161.370 0.610 0.123
Chain 1: 800 -7803.384 0.539 0.123
Chain 1: 900 -8043.559 0.483 0.094
Chain 1: 1000 -7954.908 0.436 0.094
Chain 1: 1100 -7610.002 0.340 0.062
Chain 1: 1200 -7580.232 0.128 0.046
Chain 1: 1300 -7732.783 0.047 0.045
Chain 1: 1400 -7771.622 0.044 0.045
Chain 1: 1500 -7564.875 0.041 0.030
Chain 1: 1600 -7524.145 0.029 0.027
Chain 1: 1700 -7473.459 0.020 0.020
Chain 1: 1800 -7511.219 0.016 0.011
Chain 1: 1900 -7525.434 0.013 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002533 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.33 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86360.086 1.000 1.000
Chain 1: 200 -13291.021 3.249 5.498
Chain 1: 300 -9664.668 2.291 1.000
Chain 1: 400 -10397.062 1.736 1.000
Chain 1: 500 -8640.617 1.429 0.375
Chain 1: 600 -8118.722 1.202 0.375
Chain 1: 700 -8377.706 1.035 0.203
Chain 1: 800 -9078.980 0.915 0.203
Chain 1: 900 -8428.205 0.822 0.077
Chain 1: 1000 -8284.819 0.741 0.077
Chain 1: 1100 -8500.873 0.644 0.077 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8034.188 0.100 0.070
Chain 1: 1300 -8247.402 0.065 0.064
Chain 1: 1400 -8371.948 0.059 0.058
Chain 1: 1500 -8233.931 0.041 0.031
Chain 1: 1600 -8346.425 0.036 0.026
Chain 1: 1700 -8428.813 0.034 0.025
Chain 1: 1800 -8018.550 0.031 0.025
Chain 1: 1900 -8114.658 0.024 0.017
Chain 1: 2000 -8087.380 0.023 0.017
Chain 1: 2100 -8209.174 0.022 0.015
Chain 1: 2200 -8047.591 0.018 0.015
Chain 1: 2300 -8111.982 0.016 0.015
Chain 1: 2400 -8179.299 0.016 0.013
Chain 1: 2500 -8125.079 0.015 0.012
Chain 1: 2600 -8123.436 0.013 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003168 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.68 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8402625.421 1.000 1.000
Chain 1: 200 -1585653.489 2.650 4.299
Chain 1: 300 -891796.335 2.026 1.000
Chain 1: 400 -458228.132 1.756 1.000
Chain 1: 500 -358443.445 1.460 0.946
Chain 1: 600 -232975.823 1.307 0.946
Chain 1: 700 -119086.995 1.257 0.946
Chain 1: 800 -86247.994 1.147 0.946
Chain 1: 900 -66572.180 1.053 0.778
Chain 1: 1000 -51361.835 0.977 0.778
Chain 1: 1100 -38834.528 0.909 0.539 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38008.148 0.481 0.381
Chain 1: 1300 -25969.991 0.450 0.381
Chain 1: 1400 -25687.559 0.356 0.323
Chain 1: 1500 -22276.598 0.344 0.323
Chain 1: 1600 -21492.896 0.294 0.296
Chain 1: 1700 -20368.153 0.204 0.296
Chain 1: 1800 -20312.418 0.166 0.153
Chain 1: 1900 -20638.372 0.138 0.055
Chain 1: 2000 -19150.442 0.116 0.055
Chain 1: 2100 -19388.746 0.085 0.036
Chain 1: 2200 -19614.969 0.084 0.036
Chain 1: 2300 -19232.413 0.040 0.020
Chain 1: 2400 -19004.601 0.040 0.020
Chain 1: 2500 -18806.437 0.025 0.016
Chain 1: 2600 -18436.888 0.024 0.016
Chain 1: 2700 -18393.936 0.018 0.012
Chain 1: 2800 -18110.805 0.020 0.016
Chain 1: 2900 -18391.974 0.020 0.015
Chain 1: 3000 -18378.195 0.012 0.012
Chain 1: 3100 -18463.150 0.011 0.012
Chain 1: 3200 -18153.944 0.012 0.015
Chain 1: 3300 -18358.566 0.011 0.012
Chain 1: 3400 -17833.631 0.013 0.015
Chain 1: 3500 -18445.246 0.015 0.016
Chain 1: 3600 -17752.296 0.017 0.016
Chain 1: 3700 -18138.814 0.019 0.017
Chain 1: 3800 -17099.013 0.023 0.021
Chain 1: 3900 -17095.158 0.022 0.021
Chain 1: 4000 -17212.483 0.022 0.021
Chain 1: 4100 -17126.266 0.022 0.021
Chain 1: 4200 -16942.616 0.022 0.021
Chain 1: 4300 -17080.951 0.021 0.021
Chain 1: 4400 -17037.873 0.019 0.011
Chain 1: 4500 -16940.420 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001341 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.41 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12343.227 1.000 1.000
Chain 1: 200 -9158.693 0.674 1.000
Chain 1: 300 -8115.314 0.492 0.348
Chain 1: 400 -8225.362 0.372 0.348
Chain 1: 500 -8113.857 0.301 0.129
Chain 1: 600 -8037.635 0.252 0.129
Chain 1: 700 -7959.788 0.218 0.014
Chain 1: 800 -7972.174 0.191 0.014
Chain 1: 900 -8087.628 0.171 0.014
Chain 1: 1000 -8022.339 0.155 0.014
Chain 1: 1100 -8091.973 0.056 0.013
Chain 1: 1200 -7970.468 0.022 0.013
Chain 1: 1300 -7946.527 0.010 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001379 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.79 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57073.000 1.000 1.000
Chain 1: 200 -17373.311 1.643 2.285
Chain 1: 300 -8772.343 1.422 1.000
Chain 1: 400 -8424.539 1.077 1.000
Chain 1: 500 -8861.858 0.871 0.980
Chain 1: 600 -8306.655 0.737 0.980
Chain 1: 700 -8015.343 0.637 0.067
Chain 1: 800 -8128.533 0.559 0.067
Chain 1: 900 -8120.684 0.497 0.049
Chain 1: 1000 -7721.617 0.453 0.052
Chain 1: 1100 -7796.446 0.354 0.049
Chain 1: 1200 -7730.180 0.126 0.041
Chain 1: 1300 -7763.468 0.028 0.036
Chain 1: 1400 -7990.288 0.027 0.028
Chain 1: 1500 -7683.406 0.026 0.028
Chain 1: 1600 -7709.475 0.020 0.014
Chain 1: 1700 -7607.989 0.017 0.013
Chain 1: 1800 -7672.106 0.017 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002522 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.22 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85906.460 1.000 1.000
Chain 1: 200 -13445.880 3.195 5.389
Chain 1: 300 -9874.555 2.250 1.000
Chain 1: 400 -10760.042 1.708 1.000
Chain 1: 500 -8820.299 1.411 0.362
Chain 1: 600 -8734.946 1.177 0.362
Chain 1: 700 -8542.542 1.012 0.220
Chain 1: 800 -8679.405 0.888 0.220
Chain 1: 900 -8646.086 0.789 0.082
Chain 1: 1000 -8456.066 0.713 0.082
Chain 1: 1100 -8726.855 0.616 0.031 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8443.616 0.080 0.031
Chain 1: 1300 -8613.046 0.046 0.023
Chain 1: 1400 -8612.187 0.038 0.022
Chain 1: 1500 -8476.312 0.017 0.020
Chain 1: 1600 -8584.665 0.018 0.020
Chain 1: 1700 -8670.967 0.017 0.016
Chain 1: 1800 -8276.150 0.020 0.020
Chain 1: 1900 -8376.699 0.021 0.020
Chain 1: 2000 -8347.500 0.019 0.016
Chain 1: 2100 -8469.368 0.017 0.014
Chain 1: 2200 -8249.393 0.016 0.014
Chain 1: 2300 -8405.570 0.016 0.014
Chain 1: 2400 -8419.147 0.016 0.014
Chain 1: 2500 -8389.063 0.015 0.013
Chain 1: 2600 -8391.750 0.014 0.012
Chain 1: 2700 -8297.965 0.014 0.012
Chain 1: 2800 -8268.801 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003319 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.19 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8418143.304 1.000 1.000
Chain 1: 200 -1583638.732 2.658 4.316
Chain 1: 300 -890598.092 2.031 1.000
Chain 1: 400 -457935.722 1.760 1.000
Chain 1: 500 -358099.260 1.463 0.945
Chain 1: 600 -232981.295 1.309 0.945
Chain 1: 700 -119148.644 1.259 0.945
Chain 1: 800 -86386.904 1.149 0.945
Chain 1: 900 -66717.361 1.054 0.778
Chain 1: 1000 -51511.474 0.978 0.778
Chain 1: 1100 -38993.785 0.910 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38165.429 0.481 0.379
Chain 1: 1300 -26125.785 0.449 0.379
Chain 1: 1400 -25843.939 0.355 0.321
Chain 1: 1500 -22433.523 0.343 0.321
Chain 1: 1600 -21650.704 0.293 0.295
Chain 1: 1700 -20524.833 0.203 0.295
Chain 1: 1800 -20469.018 0.165 0.152
Chain 1: 1900 -20794.817 0.137 0.055
Chain 1: 2000 -19307.250 0.115 0.055
Chain 1: 2100 -19545.253 0.084 0.036
Chain 1: 2200 -19771.634 0.083 0.036
Chain 1: 2300 -19389.045 0.039 0.020
Chain 1: 2400 -19161.289 0.039 0.020
Chain 1: 2500 -18963.475 0.025 0.016
Chain 1: 2600 -18593.788 0.024 0.016
Chain 1: 2700 -18550.860 0.018 0.012
Chain 1: 2800 -18267.977 0.020 0.015
Chain 1: 2900 -18549.024 0.020 0.015
Chain 1: 3000 -18535.183 0.012 0.012
Chain 1: 3100 -18620.143 0.011 0.012
Chain 1: 3200 -18311.009 0.012 0.015
Chain 1: 3300 -18515.600 0.011 0.012
Chain 1: 3400 -17990.948 0.013 0.015
Chain 1: 3500 -18602.219 0.015 0.015
Chain 1: 3600 -17909.691 0.017 0.015
Chain 1: 3700 -18295.898 0.019 0.017
Chain 1: 3800 -17256.895 0.023 0.021
Chain 1: 3900 -17253.117 0.022 0.021
Chain 1: 4000 -17370.372 0.022 0.021
Chain 1: 4100 -17284.223 0.022 0.021
Chain 1: 4200 -17100.770 0.022 0.021
Chain 1: 4300 -17238.925 0.021 0.021
Chain 1: 4400 -17195.959 0.019 0.011
Chain 1: 4500 -17098.581 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001278 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.78 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12331.556 1.000 1.000
Chain 1: 200 -8965.088 0.688 1.000
Chain 1: 300 -7930.780 0.502 0.376
Chain 1: 400 -8097.689 0.382 0.376
Chain 1: 500 -8065.896 0.306 0.130
Chain 1: 600 -7870.274 0.259 0.130
Chain 1: 700 -7914.067 0.223 0.025
Chain 1: 800 -7807.379 0.197 0.025
Chain 1: 900 -7888.878 0.176 0.021
Chain 1: 1000 -7856.444 0.159 0.021
Chain 1: 1100 -7916.715 0.060 0.014
Chain 1: 1200 -7818.535 0.023 0.013
Chain 1: 1300 -7841.112 0.011 0.010
Chain 1: 1400 -7798.823 0.009 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001434 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.34 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -62806.572 1.000 1.000
Chain 1: 200 -17972.972 1.747 2.495
Chain 1: 300 -8664.332 1.523 1.074
Chain 1: 400 -8314.953 1.153 1.074
Chain 1: 500 -8447.457 0.925 1.000
Chain 1: 600 -8848.666 0.779 1.000
Chain 1: 700 -8323.597 0.676 0.063
Chain 1: 800 -7874.663 0.599 0.063
Chain 1: 900 -7753.685 0.534 0.057
Chain 1: 1000 -7764.207 0.481 0.057
Chain 1: 1100 -7621.633 0.383 0.045
Chain 1: 1200 -7599.514 0.134 0.042
Chain 1: 1300 -7539.089 0.027 0.019
Chain 1: 1400 -7868.506 0.027 0.019
Chain 1: 1500 -7571.043 0.029 0.039
Chain 1: 1600 -7473.853 0.026 0.019
Chain 1: 1700 -7469.342 0.020 0.016
Chain 1: 1800 -7568.181 0.015 0.013
Chain 1: 1900 -7574.132 0.014 0.013
Chain 1: 2000 -7566.460 0.014 0.013
Chain 1: 2100 -7569.183 0.012 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003292 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.92 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85811.892 1.000 1.000
Chain 1: 200 -13277.408 3.232 5.463
Chain 1: 300 -9713.781 2.277 1.000
Chain 1: 400 -10443.312 1.725 1.000
Chain 1: 500 -8620.278 1.422 0.367
Chain 1: 600 -8302.776 1.192 0.367
Chain 1: 700 -8622.659 1.027 0.211
Chain 1: 800 -8583.613 0.899 0.211
Chain 1: 900 -8555.150 0.799 0.070
Chain 1: 1000 -8373.522 0.722 0.070
Chain 1: 1100 -8580.259 0.624 0.038 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8246.869 0.082 0.038
Chain 1: 1300 -8429.836 0.047 0.037
Chain 1: 1400 -8429.218 0.040 0.024
Chain 1: 1500 -8328.076 0.020 0.022
Chain 1: 1600 -8427.644 0.018 0.022
Chain 1: 1700 -8515.114 0.015 0.012
Chain 1: 1800 -8122.520 0.019 0.022
Chain 1: 1900 -8224.581 0.020 0.022
Chain 1: 2000 -8194.869 0.018 0.012
Chain 1: 2100 -8320.832 0.018 0.012
Chain 1: 2200 -8106.356 0.016 0.012
Chain 1: 2300 -8253.324 0.016 0.012
Chain 1: 2400 -8268.775 0.016 0.012
Chain 1: 2500 -8235.443 0.015 0.012
Chain 1: 2600 -8237.404 0.014 0.012
Chain 1: 2700 -8144.251 0.014 0.012
Chain 1: 2800 -8117.150 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002508 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.08 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8419708.778 1.000 1.000
Chain 1: 200 -1586064.897 2.654 4.309
Chain 1: 300 -890928.198 2.030 1.000
Chain 1: 400 -457683.273 1.759 1.000
Chain 1: 500 -357645.474 1.463 0.947
Chain 1: 600 -232673.472 1.309 0.947
Chain 1: 700 -118917.381 1.258 0.947
Chain 1: 800 -86134.515 1.149 0.947
Chain 1: 900 -66487.082 1.054 0.780
Chain 1: 1000 -51293.019 0.978 0.780
Chain 1: 1100 -38783.359 0.910 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37957.158 0.482 0.381
Chain 1: 1300 -25936.053 0.450 0.381
Chain 1: 1400 -25655.548 0.356 0.323
Chain 1: 1500 -22248.988 0.344 0.323
Chain 1: 1600 -21466.857 0.294 0.296
Chain 1: 1700 -20343.871 0.204 0.296
Chain 1: 1800 -20288.549 0.166 0.153
Chain 1: 1900 -20614.252 0.138 0.055
Chain 1: 2000 -19127.963 0.116 0.055
Chain 1: 2100 -19366.140 0.085 0.036
Chain 1: 2200 -19592.020 0.084 0.036
Chain 1: 2300 -19209.876 0.040 0.020
Chain 1: 2400 -18982.152 0.040 0.020
Chain 1: 2500 -18784.078 0.025 0.016
Chain 1: 2600 -18414.747 0.024 0.016
Chain 1: 2700 -18371.909 0.018 0.012
Chain 1: 2800 -18088.887 0.020 0.016
Chain 1: 2900 -18369.939 0.020 0.015
Chain 1: 3000 -18356.182 0.012 0.012
Chain 1: 3100 -18441.073 0.011 0.012
Chain 1: 3200 -18132.056 0.012 0.015
Chain 1: 3300 -18336.560 0.011 0.012
Chain 1: 3400 -17811.972 0.013 0.015
Chain 1: 3500 -18423.040 0.015 0.016
Chain 1: 3600 -17730.811 0.017 0.016
Chain 1: 3700 -18116.760 0.019 0.017
Chain 1: 3800 -17078.088 0.023 0.021
Chain 1: 3900 -17074.267 0.022 0.021
Chain 1: 4000 -17191.590 0.022 0.021
Chain 1: 4100 -17105.394 0.022 0.021
Chain 1: 4200 -16922.022 0.022 0.021
Chain 1: 4300 -17060.153 0.021 0.021
Chain 1: 4400 -17017.270 0.019 0.011
Chain 1: 4500 -16919.856 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001442 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.42 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -11959.651 1.000 1.000
Chain 1: 200 -8904.938 0.672 1.000
Chain 1: 300 -7733.487 0.498 0.343
Chain 1: 400 -7911.954 0.379 0.343
Chain 1: 500 -7954.925 0.304 0.151
Chain 1: 600 -7811.393 0.257 0.151
Chain 1: 700 -7601.833 0.224 0.028
Chain 1: 800 -7620.132 0.196 0.028
Chain 1: 900 -7578.550 0.175 0.023
Chain 1: 1000 -7654.008 0.159 0.023
Chain 1: 1100 -7721.495 0.059 0.018
Chain 1: 1200 -7610.895 0.027 0.015
Chain 1: 1300 -7636.008 0.012 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001388 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.88 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -45968.437 1.000 1.000
Chain 1: 200 -15127.219 1.519 2.039
Chain 1: 300 -8470.176 1.275 1.000
Chain 1: 400 -8325.955 0.961 1.000
Chain 1: 500 -8129.377 0.773 0.786
Chain 1: 600 -7785.524 0.652 0.786
Chain 1: 700 -8238.468 0.566 0.055
Chain 1: 800 -8169.382 0.497 0.055
Chain 1: 900 -7816.902 0.447 0.045
Chain 1: 1000 -7790.323 0.402 0.045
Chain 1: 1100 -7592.038 0.305 0.044
Chain 1: 1200 -7686.821 0.102 0.026
Chain 1: 1300 -7623.366 0.024 0.024
Chain 1: 1400 -7605.098 0.023 0.024
Chain 1: 1500 -7566.165 0.021 0.012
Chain 1: 1600 -7482.889 0.018 0.011
Chain 1: 1700 -7481.854 0.012 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003015 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.15 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85932.404 1.000 1.000
Chain 1: 200 -13048.643 3.293 5.586
Chain 1: 300 -9482.312 2.321 1.000
Chain 1: 400 -10311.089 1.761 1.000
Chain 1: 500 -8415.583 1.453 0.376
Chain 1: 600 -8032.948 1.219 0.376
Chain 1: 700 -8338.094 1.050 0.225
Chain 1: 800 -8469.815 0.921 0.225
Chain 1: 900 -8388.539 0.820 0.080
Chain 1: 1000 -8066.158 0.742 0.080
Chain 1: 1100 -8382.891 0.645 0.048 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8035.295 0.091 0.043
Chain 1: 1300 -8117.027 0.055 0.040
Chain 1: 1400 -8262.250 0.048 0.038
Chain 1: 1500 -8103.693 0.028 0.037
Chain 1: 1600 -8214.672 0.024 0.020
Chain 1: 1700 -8295.419 0.022 0.018
Chain 1: 1800 -7905.003 0.025 0.020
Chain 1: 1900 -8008.333 0.025 0.020
Chain 1: 2000 -7977.834 0.022 0.018
Chain 1: 2100 -8107.105 0.020 0.016
Chain 1: 2200 -7893.691 0.018 0.016
Chain 1: 2300 -8036.845 0.019 0.018
Chain 1: 2400 -8050.800 0.017 0.016
Chain 1: 2500 -8018.003 0.016 0.014
Chain 1: 2600 -8019.039 0.014 0.013
Chain 1: 2700 -7926.512 0.014 0.013
Chain 1: 2800 -7901.061 0.010 0.012 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002531 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.31 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8377178.433 1.000 1.000
Chain 1: 200 -1579989.473 2.651 4.302
Chain 1: 300 -889870.950 2.026 1.000
Chain 1: 400 -457146.052 1.756 1.000
Chain 1: 500 -357831.927 1.460 0.947
Chain 1: 600 -232862.969 1.306 0.947
Chain 1: 700 -118950.103 1.257 0.947
Chain 1: 800 -86103.943 1.147 0.947
Chain 1: 900 -66410.339 1.053 0.776
Chain 1: 1000 -51175.584 0.977 0.776
Chain 1: 1100 -38622.510 0.910 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37793.666 0.482 0.381
Chain 1: 1300 -25728.657 0.451 0.381
Chain 1: 1400 -25443.811 0.357 0.325
Chain 1: 1500 -22025.368 0.345 0.325
Chain 1: 1600 -21239.388 0.295 0.298
Chain 1: 1700 -20111.102 0.205 0.297
Chain 1: 1800 -20054.583 0.167 0.155
Chain 1: 1900 -20380.284 0.139 0.056
Chain 1: 2000 -18891.169 0.117 0.056
Chain 1: 2100 -19129.570 0.086 0.037
Chain 1: 2200 -19355.812 0.085 0.037
Chain 1: 2300 -18973.342 0.040 0.020
Chain 1: 2400 -18745.565 0.040 0.020
Chain 1: 2500 -18547.642 0.026 0.016
Chain 1: 2600 -18178.236 0.024 0.016
Chain 1: 2700 -18135.383 0.019 0.012
Chain 1: 2800 -17852.416 0.020 0.016
Chain 1: 2900 -18133.566 0.020 0.016
Chain 1: 3000 -18119.736 0.012 0.012
Chain 1: 3100 -18204.623 0.011 0.012
Chain 1: 3200 -17895.652 0.012 0.016
Chain 1: 3300 -18100.138 0.011 0.012
Chain 1: 3400 -17575.629 0.013 0.016
Chain 1: 3500 -18186.674 0.015 0.016
Chain 1: 3600 -17494.530 0.017 0.016
Chain 1: 3700 -17880.457 0.019 0.017
Chain 1: 3800 -16841.938 0.024 0.022
Chain 1: 3900 -16838.154 0.022 0.022
Chain 1: 4000 -16955.444 0.023 0.022
Chain 1: 4100 -16869.249 0.023 0.022
Chain 1: 4200 -16685.925 0.022 0.022
Chain 1: 4300 -16824.016 0.022 0.022
Chain 1: 4400 -16781.167 0.019 0.011
Chain 1: 4500 -16683.771 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001406 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.06 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12896.666 1.000 1.000
Chain 1: 200 -9773.187 0.660 1.000
Chain 1: 300 -8443.890 0.492 0.320
Chain 1: 400 -8612.393 0.374 0.320
Chain 1: 500 -8592.927 0.300 0.157
Chain 1: 600 -8367.205 0.254 0.157
Chain 1: 700 -8448.625 0.219 0.027
Chain 1: 800 -8274.410 0.195 0.027
Chain 1: 900 -8399.505 0.175 0.021
Chain 1: 1000 -8375.973 0.157 0.021
Chain 1: 1100 -8401.752 0.058 0.020
Chain 1: 1200 -8304.344 0.027 0.015
Chain 1: 1300 -8391.403 0.012 0.012
Chain 1: 1400 -8283.684 0.012 0.012
Chain 1: 1500 -8388.188 0.013 0.012
Chain 1: 1600 -8313.123 0.011 0.012
Chain 1: 1700 -8265.409 0.010 0.012
Chain 1: 1800 -8239.737 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001404 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.04 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56701.544 1.000 1.000
Chain 1: 200 -17753.129 1.597 2.194
Chain 1: 300 -8919.232 1.395 1.000
Chain 1: 400 -8347.419 1.063 1.000
Chain 1: 500 -8789.820 0.861 0.990
Chain 1: 600 -9114.681 0.723 0.990
Chain 1: 700 -8014.449 0.639 0.137
Chain 1: 800 -8106.425 0.561 0.137
Chain 1: 900 -7961.890 0.501 0.069
Chain 1: 1000 -8014.324 0.451 0.069
Chain 1: 1100 -7802.554 0.354 0.050
Chain 1: 1200 -7678.233 0.136 0.036
Chain 1: 1300 -7646.835 0.038 0.027
Chain 1: 1400 -7791.847 0.033 0.019
Chain 1: 1500 -7557.354 0.031 0.019
Chain 1: 1600 -7629.681 0.028 0.018
Chain 1: 1700 -7624.840 0.014 0.016
Chain 1: 1800 -7551.525 0.014 0.016
Chain 1: 1900 -7594.060 0.013 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002538 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.38 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86580.763 1.000 1.000
Chain 1: 200 -14063.169 3.078 5.157
Chain 1: 300 -10359.806 2.171 1.000
Chain 1: 400 -11553.965 1.654 1.000
Chain 1: 500 -9345.330 1.371 0.357
Chain 1: 600 -8945.776 1.150 0.357
Chain 1: 700 -8783.712 0.988 0.236
Chain 1: 800 -9339.151 0.872 0.236
Chain 1: 900 -9100.245 0.778 0.103
Chain 1: 1000 -8910.439 0.702 0.103
Chain 1: 1100 -9120.990 0.605 0.059 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8627.583 0.095 0.057
Chain 1: 1300 -8862.696 0.062 0.045
Chain 1: 1400 -8975.805 0.053 0.027
Chain 1: 1500 -8887.993 0.030 0.026
Chain 1: 1600 -8998.110 0.027 0.023
Chain 1: 1700 -9056.732 0.026 0.023
Chain 1: 1800 -8622.554 0.025 0.023
Chain 1: 1900 -8726.792 0.023 0.021
Chain 1: 2000 -8702.052 0.021 0.013
Chain 1: 2100 -8670.418 0.019 0.012
Chain 1: 2200 -8645.077 0.014 0.012
Chain 1: 2300 -8780.549 0.013 0.012
Chain 1: 2400 -8627.499 0.013 0.012
Chain 1: 2500 -8696.769 0.013 0.012
Chain 1: 2600 -8614.911 0.013 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002595 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.95 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8393494.247 1.000 1.000
Chain 1: 200 -1584941.882 2.648 4.296
Chain 1: 300 -891708.669 2.024 1.000
Chain 1: 400 -458443.859 1.755 1.000
Chain 1: 500 -358775.947 1.459 0.945
Chain 1: 600 -233681.528 1.305 0.945
Chain 1: 700 -119864.243 1.254 0.945
Chain 1: 800 -87066.599 1.145 0.945
Chain 1: 900 -67402.628 1.050 0.777
Chain 1: 1000 -52197.316 0.974 0.777
Chain 1: 1100 -39666.396 0.906 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38847.849 0.478 0.377
Chain 1: 1300 -26788.899 0.445 0.377
Chain 1: 1400 -26508.985 0.352 0.316
Chain 1: 1500 -23092.070 0.339 0.316
Chain 1: 1600 -22307.944 0.289 0.292
Chain 1: 1700 -21179.467 0.199 0.291
Chain 1: 1800 -21123.500 0.162 0.148
Chain 1: 1900 -21449.994 0.134 0.053
Chain 1: 2000 -19959.412 0.113 0.053
Chain 1: 2100 -20197.878 0.082 0.035
Chain 1: 2200 -20424.759 0.081 0.035
Chain 1: 2300 -20041.528 0.038 0.019
Chain 1: 2400 -19813.473 0.038 0.019
Chain 1: 2500 -19615.548 0.024 0.015
Chain 1: 2600 -19245.281 0.023 0.015
Chain 1: 2700 -19202.179 0.018 0.012
Chain 1: 2800 -18918.853 0.019 0.015
Chain 1: 2900 -19200.354 0.019 0.015
Chain 1: 3000 -19186.487 0.012 0.012
Chain 1: 3100 -19271.496 0.011 0.012
Chain 1: 3200 -18961.937 0.011 0.015
Chain 1: 3300 -19166.893 0.010 0.012
Chain 1: 3400 -18641.355 0.012 0.015
Chain 1: 3500 -19253.885 0.014 0.015
Chain 1: 3600 -18559.798 0.016 0.015
Chain 1: 3700 -18947.157 0.018 0.016
Chain 1: 3800 -17905.594 0.022 0.020
Chain 1: 3900 -17901.730 0.021 0.020
Chain 1: 4000 -18019.039 0.021 0.020
Chain 1: 4100 -17932.686 0.021 0.020
Chain 1: 4200 -17748.705 0.021 0.020
Chain 1: 4300 -17887.249 0.021 0.020
Chain 1: 4400 -17843.848 0.018 0.010
Chain 1: 4500 -17746.366 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001253 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.53 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48676.708 1.000 1.000
Chain 1: 200 -16979.975 1.433 1.867
Chain 1: 300 -13020.513 1.057 1.000
Chain 1: 400 -18748.355 0.869 1.000
Chain 1: 500 -13203.468 0.779 0.420
Chain 1: 600 -15825.010 0.677 0.420
Chain 1: 700 -14099.523 0.598 0.306
Chain 1: 800 -13521.491 0.528 0.306
Chain 1: 900 -15283.978 0.482 0.304
Chain 1: 1000 -20728.602 0.461 0.304
Chain 1: 1100 -10136.848 0.465 0.304
Chain 1: 1200 -19473.280 0.326 0.304
Chain 1: 1300 -11329.829 0.368 0.306
Chain 1: 1400 -9795.132 0.353 0.263
Chain 1: 1500 -9993.968 0.313 0.166
Chain 1: 1600 -12064.656 0.313 0.172
Chain 1: 1700 -9754.746 0.325 0.237
Chain 1: 1800 -11958.676 0.339 0.237
Chain 1: 1900 -11086.069 0.335 0.237
Chain 1: 2000 -10819.137 0.312 0.184
Chain 1: 2100 -9654.973 0.219 0.172
Chain 1: 2200 -10351.403 0.178 0.157
Chain 1: 2300 -11926.693 0.119 0.132
Chain 1: 2400 -9234.265 0.133 0.132
Chain 1: 2500 -9532.362 0.134 0.132
Chain 1: 2600 -9204.444 0.120 0.121
Chain 1: 2700 -9120.623 0.098 0.079
Chain 1: 2800 -9539.362 0.083 0.067
Chain 1: 2900 -9721.449 0.077 0.044
Chain 1: 3000 -9297.046 0.080 0.046
Chain 1: 3100 -9640.434 0.071 0.044
Chain 1: 3200 -8740.948 0.075 0.044
Chain 1: 3300 -9521.616 0.070 0.044
Chain 1: 3400 -10383.817 0.049 0.044
Chain 1: 3500 -9026.610 0.061 0.046
Chain 1: 3600 -9912.389 0.066 0.082
Chain 1: 3700 -8796.545 0.078 0.083
Chain 1: 3800 -10263.382 0.088 0.089
Chain 1: 3900 -8850.485 0.102 0.103
Chain 1: 4000 -9101.907 0.100 0.103
Chain 1: 4100 -8861.281 0.099 0.103
Chain 1: 4200 -11784.489 0.114 0.127
Chain 1: 4300 -9322.198 0.132 0.143
Chain 1: 4400 -8620.681 0.132 0.143
Chain 1: 4500 -9173.670 0.123 0.127
Chain 1: 4600 -10408.055 0.126 0.127
Chain 1: 4700 -8916.812 0.130 0.143
Chain 1: 4800 -8612.765 0.119 0.119
Chain 1: 4900 -13619.456 0.140 0.119
Chain 1: 5000 -11052.460 0.160 0.167
Chain 1: 5100 -8908.399 0.182 0.232
Chain 1: 5200 -14448.724 0.195 0.232
Chain 1: 5300 -13936.967 0.172 0.167
Chain 1: 5400 -8483.710 0.228 0.232
Chain 1: 5500 -8417.294 0.223 0.232
Chain 1: 5600 -8512.270 0.213 0.232
Chain 1: 5700 -10377.926 0.214 0.232
Chain 1: 5800 -8503.713 0.232 0.232
Chain 1: 5900 -16585.030 0.244 0.232
Chain 1: 6000 -9031.503 0.305 0.241
Chain 1: 6100 -8890.589 0.282 0.220
Chain 1: 6200 -8954.073 0.245 0.180
Chain 1: 6300 -13108.043 0.273 0.220
Chain 1: 6400 -9616.120 0.245 0.220
Chain 1: 6500 -9381.664 0.246 0.220
Chain 1: 6600 -10410.593 0.255 0.220
Chain 1: 6700 -8764.281 0.256 0.220
Chain 1: 6800 -8386.639 0.238 0.188
Chain 1: 6900 -9079.064 0.197 0.099
Chain 1: 7000 -8441.446 0.121 0.076
Chain 1: 7100 -8568.189 0.121 0.076
Chain 1: 7200 -9312.427 0.128 0.080
Chain 1: 7300 -8363.087 0.108 0.080
Chain 1: 7400 -8600.526 0.074 0.076
Chain 1: 7500 -9169.179 0.078 0.076
Chain 1: 7600 -8426.849 0.077 0.076
Chain 1: 7700 -8577.834 0.060 0.076
Chain 1: 7800 -10298.422 0.072 0.076
Chain 1: 7900 -8264.886 0.089 0.080
Chain 1: 8000 -9165.352 0.091 0.088
Chain 1: 8100 -12914.854 0.119 0.098
Chain 1: 8200 -8258.604 0.167 0.114
Chain 1: 8300 -12472.265 0.190 0.167
Chain 1: 8400 -10422.553 0.207 0.197
Chain 1: 8500 -12613.396 0.218 0.197
Chain 1: 8600 -8788.817 0.253 0.246
Chain 1: 8700 -9719.817 0.260 0.246
Chain 1: 8800 -8976.449 0.252 0.246
Chain 1: 8900 -8672.759 0.231 0.197
Chain 1: 9000 -9237.332 0.227 0.197
Chain 1: 9100 -8574.130 0.206 0.174
Chain 1: 9200 -9873.282 0.163 0.132
Chain 1: 9300 -9693.678 0.131 0.096
Chain 1: 9400 -9503.921 0.113 0.083
Chain 1: 9500 -8355.144 0.109 0.083
Chain 1: 9600 -9027.533 0.073 0.077
Chain 1: 9700 -8470.316 0.070 0.074
Chain 1: 9800 -9131.897 0.069 0.072
Chain 1: 9900 -8551.594 0.073 0.072
Chain 1: 10000 -8295.369 0.070 0.072
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001372 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.72 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56961.448 1.000 1.000
Chain 1: 200 -17283.471 1.648 2.296
Chain 1: 300 -8657.571 1.431 1.000
Chain 1: 400 -8354.595 1.082 1.000
Chain 1: 500 -8540.017 0.870 0.996
Chain 1: 600 -8842.776 0.731 0.996
Chain 1: 700 -8347.774 0.635 0.059
Chain 1: 800 -7969.139 0.561 0.059
Chain 1: 900 -7837.734 0.501 0.048
Chain 1: 1000 -7869.797 0.451 0.048
Chain 1: 1100 -7620.084 0.354 0.036
Chain 1: 1200 -7734.625 0.126 0.034
Chain 1: 1300 -7728.201 0.027 0.033
Chain 1: 1400 -7667.763 0.024 0.022
Chain 1: 1500 -7582.353 0.023 0.017
Chain 1: 1600 -7758.655 0.022 0.017
Chain 1: 1700 -7491.591 0.019 0.017
Chain 1: 1800 -7585.572 0.016 0.015
Chain 1: 1900 -7616.018 0.015 0.012
Chain 1: 2000 -7643.665 0.015 0.012
Chain 1: 2100 -7601.964 0.012 0.011
Chain 1: 2200 -7685.550 0.011 0.011
Chain 1: 2300 -7591.698 0.013 0.011
Chain 1: 2400 -7630.190 0.012 0.011
Chain 1: 2500 -7566.629 0.012 0.011
Chain 1: 2600 -7514.412 0.010 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002997 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.97 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86312.179 1.000 1.000
Chain 1: 200 -13405.906 3.219 5.438
Chain 1: 300 -9849.817 2.266 1.000
Chain 1: 400 -10692.472 1.720 1.000
Chain 1: 500 -8714.128 1.421 0.361
Chain 1: 600 -8553.829 1.187 0.361
Chain 1: 700 -8576.390 1.018 0.227
Chain 1: 800 -8782.636 0.894 0.227
Chain 1: 900 -8670.259 0.796 0.079
Chain 1: 1000 -8447.156 0.719 0.079
Chain 1: 1100 -8714.606 0.622 0.031 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8454.950 0.081 0.031
Chain 1: 1300 -8568.715 0.046 0.026
Chain 1: 1400 -8574.321 0.039 0.023
Chain 1: 1500 -8447.312 0.017 0.019
Chain 1: 1600 -8554.319 0.017 0.015
Chain 1: 1700 -8639.616 0.018 0.015
Chain 1: 1800 -8248.195 0.020 0.015
Chain 1: 1900 -8349.536 0.020 0.015
Chain 1: 2000 -8320.047 0.018 0.013
Chain 1: 2100 -8445.184 0.016 0.013
Chain 1: 2200 -8230.017 0.016 0.013
Chain 1: 2300 -8378.376 0.016 0.015
Chain 1: 2400 -8393.654 0.016 0.015
Chain 1: 2500 -8360.799 0.015 0.013
Chain 1: 2600 -8362.892 0.014 0.012
Chain 1: 2700 -8269.592 0.014 0.012
Chain 1: 2800 -8242.131 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003007 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.07 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8401566.255 1.000 1.000
Chain 1: 200 -1583625.342 2.653 4.305
Chain 1: 300 -889790.641 2.028 1.000
Chain 1: 400 -456615.034 1.758 1.000
Chain 1: 500 -357108.412 1.462 0.949
Chain 1: 600 -232210.686 1.308 0.949
Chain 1: 700 -118799.347 1.258 0.949
Chain 1: 800 -86114.279 1.148 0.949
Chain 1: 900 -66520.084 1.053 0.780
Chain 1: 1000 -51360.559 0.977 0.780
Chain 1: 1100 -38878.841 0.910 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38056.888 0.481 0.380
Chain 1: 1300 -26054.185 0.449 0.380
Chain 1: 1400 -25775.474 0.355 0.321
Chain 1: 1500 -22374.114 0.343 0.321
Chain 1: 1600 -21593.731 0.293 0.295
Chain 1: 1700 -20472.495 0.203 0.295
Chain 1: 1800 -20417.617 0.165 0.152
Chain 1: 1900 -20743.479 0.137 0.055
Chain 1: 2000 -19257.983 0.115 0.055
Chain 1: 2100 -19496.059 0.084 0.036
Chain 1: 2200 -19721.999 0.083 0.036
Chain 1: 2300 -19339.747 0.039 0.020
Chain 1: 2400 -19112.034 0.039 0.020
Chain 1: 2500 -18913.972 0.025 0.016
Chain 1: 2600 -18544.680 0.024 0.016
Chain 1: 2700 -18501.789 0.018 0.012
Chain 1: 2800 -18218.861 0.020 0.016
Chain 1: 2900 -18499.799 0.020 0.015
Chain 1: 3000 -18486.035 0.012 0.012
Chain 1: 3100 -18570.991 0.011 0.012
Chain 1: 3200 -18261.955 0.012 0.015
Chain 1: 3300 -18466.445 0.011 0.012
Chain 1: 3400 -17941.887 0.013 0.015
Chain 1: 3500 -18552.988 0.015 0.016
Chain 1: 3600 -17860.621 0.017 0.016
Chain 1: 3700 -18246.724 0.019 0.017
Chain 1: 3800 -17207.945 0.023 0.021
Chain 1: 3900 -17204.120 0.022 0.021
Chain 1: 4000 -17321.414 0.022 0.021
Chain 1: 4100 -17235.302 0.022 0.021
Chain 1: 4200 -17051.858 0.022 0.021
Chain 1: 4300 -17190.039 0.021 0.021
Chain 1: 4400 -17147.129 0.019 0.011
Chain 1: 4500 -17049.705 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001509 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.09 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12901.149 1.000 1.000
Chain 1: 200 -9748.897 0.662 1.000
Chain 1: 300 -8324.739 0.498 0.323
Chain 1: 400 -8552.903 0.380 0.323
Chain 1: 500 -8451.229 0.307 0.171
Chain 1: 600 -8279.859 0.259 0.171
Chain 1: 700 -8177.726 0.224 0.027
Chain 1: 800 -8245.688 0.197 0.027
Chain 1: 900 -8088.297 0.177 0.021
Chain 1: 1000 -8305.349 0.162 0.026
Chain 1: 1100 -8335.030 0.062 0.021
Chain 1: 1200 -8225.081 0.031 0.019
Chain 1: 1300 -8159.726 0.015 0.013
Chain 1: 1400 -8173.511 0.013 0.012
Chain 1: 1500 -8265.698 0.012 0.012
Chain 1: 1600 -8192.296 0.011 0.011
Chain 1: 1700 -8152.893 0.011 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00144 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.4 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -50621.417 1.000 1.000
Chain 1: 200 -16600.298 1.525 2.049
Chain 1: 300 -9031.118 1.296 1.000
Chain 1: 400 -8300.280 0.994 1.000
Chain 1: 500 -8656.401 0.803 0.838
Chain 1: 600 -8880.166 0.674 0.838
Chain 1: 700 -8391.717 0.586 0.088
Chain 1: 800 -8027.268 0.518 0.088
Chain 1: 900 -8300.257 0.464 0.058
Chain 1: 1000 -7953.451 0.422 0.058
Chain 1: 1100 -7937.355 0.322 0.045
Chain 1: 1200 -8058.131 0.119 0.044
Chain 1: 1300 -7859.137 0.038 0.041
Chain 1: 1400 -7819.338 0.029 0.033
Chain 1: 1500 -7706.842 0.027 0.025
Chain 1: 1600 -7842.980 0.026 0.025
Chain 1: 1700 -7694.797 0.022 0.019
Chain 1: 1800 -7767.029 0.018 0.017
Chain 1: 1900 -7713.456 0.016 0.015
Chain 1: 2000 -7834.352 0.013 0.015
Chain 1: 2100 -7685.827 0.015 0.015
Chain 1: 2200 -7853.794 0.015 0.017
Chain 1: 2300 -7678.853 0.015 0.017
Chain 1: 2400 -7729.841 0.015 0.017
Chain 1: 2500 -7565.892 0.016 0.019
Chain 1: 2600 -7644.538 0.015 0.019
Chain 1: 2700 -7632.495 0.014 0.015
Chain 1: 2800 -7655.936 0.013 0.015
Chain 1: 2900 -7494.860 0.014 0.019
Chain 1: 3000 -7646.842 0.015 0.020
Chain 1: 3100 -7644.722 0.013 0.020
Chain 1: 3200 -7867.288 0.014 0.020
Chain 1: 3300 -7551.164 0.015 0.020
Chain 1: 3400 -7796.086 0.018 0.021
Chain 1: 3500 -7555.932 0.019 0.021
Chain 1: 3600 -7622.924 0.019 0.021
Chain 1: 3700 -7572.311 0.019 0.021
Chain 1: 3800 -7573.047 0.019 0.021
Chain 1: 3900 -7532.067 0.017 0.020
Chain 1: 4000 -7523.836 0.016 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002544 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.44 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86653.663 1.000 1.000
Chain 1: 200 -14046.298 3.085 5.169
Chain 1: 300 -10305.484 2.177 1.000
Chain 1: 400 -11717.358 1.663 1.000
Chain 1: 500 -9230.598 1.384 0.363
Chain 1: 600 -9027.732 1.157 0.363
Chain 1: 700 -9696.424 1.002 0.269
Chain 1: 800 -8768.608 0.890 0.269
Chain 1: 900 -8628.528 0.793 0.120
Chain 1: 1000 -9321.183 0.721 0.120
Chain 1: 1100 -8796.425 0.627 0.106 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -9254.348 0.115 0.074
Chain 1: 1300 -8679.403 0.085 0.069
Chain 1: 1400 -8826.462 0.075 0.066
Chain 1: 1500 -8710.779 0.049 0.060
Chain 1: 1600 -8717.707 0.047 0.060
Chain 1: 1700 -8604.305 0.042 0.049
Chain 1: 1800 -8663.067 0.032 0.017
Chain 1: 1900 -8547.276 0.031 0.017
Chain 1: 2000 -8606.375 0.025 0.014
Chain 1: 2100 -8761.323 0.020 0.014
Chain 1: 2200 -8540.615 0.018 0.014
Chain 1: 2300 -8678.151 0.013 0.014
Chain 1: 2400 -8537.227 0.013 0.014
Chain 1: 2500 -8606.673 0.013 0.014
Chain 1: 2600 -8520.023 0.013 0.014
Chain 1: 2700 -8551.458 0.012 0.014
Chain 1: 2800 -8508.005 0.012 0.014
Chain 1: 2900 -8608.923 0.012 0.012
Chain 1: 3000 -8457.715 0.013 0.016
Chain 1: 3100 -8594.101 0.013 0.016
Chain 1: 3200 -8464.046 0.012 0.015
Chain 1: 3300 -8485.171 0.011 0.012
Chain 1: 3400 -8680.337 0.011 0.012
Chain 1: 3500 -8638.954 0.011 0.012
Chain 1: 3600 -8419.371 0.013 0.015
Chain 1: 3700 -8570.741 0.014 0.016
Chain 1: 3800 -8425.000 0.015 0.017
Chain 1: 3900 -8357.671 0.015 0.017
Chain 1: 4000 -8452.765 0.014 0.016
Chain 1: 4100 -8431.178 0.013 0.015
Chain 1: 4200 -8416.912 0.011 0.011
Chain 1: 4300 -8450.118 0.012 0.011
Chain 1: 4400 -8406.858 0.010 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002619 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 26.19 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8396196.607 1.000 1.000
Chain 1: 200 -1581141.028 2.655 4.310
Chain 1: 300 -890966.436 2.028 1.000
Chain 1: 400 -458080.463 1.757 1.000
Chain 1: 500 -358747.443 1.461 0.945
Chain 1: 600 -233693.796 1.307 0.945
Chain 1: 700 -119870.910 1.256 0.945
Chain 1: 800 -87056.515 1.146 0.945
Chain 1: 900 -67385.613 1.051 0.775
Chain 1: 1000 -52183.472 0.975 0.775
Chain 1: 1100 -39652.824 0.907 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38834.353 0.478 0.377
Chain 1: 1300 -26773.160 0.445 0.377
Chain 1: 1400 -26493.812 0.352 0.316
Chain 1: 1500 -23075.491 0.339 0.316
Chain 1: 1600 -22291.014 0.289 0.292
Chain 1: 1700 -21162.007 0.199 0.291
Chain 1: 1800 -21106.037 0.162 0.148
Chain 1: 1900 -21432.670 0.134 0.053
Chain 1: 2000 -19941.307 0.113 0.053
Chain 1: 2100 -20179.957 0.082 0.035
Chain 1: 2200 -20406.928 0.081 0.035
Chain 1: 2300 -20023.543 0.038 0.019
Chain 1: 2400 -19795.393 0.038 0.019
Chain 1: 2500 -19597.401 0.024 0.015
Chain 1: 2600 -19227.066 0.023 0.015
Chain 1: 2700 -19183.899 0.018 0.012
Chain 1: 2800 -18900.498 0.019 0.015
Chain 1: 2900 -19181.998 0.019 0.015
Chain 1: 3000 -19168.201 0.012 0.012
Chain 1: 3100 -19253.256 0.011 0.012
Chain 1: 3200 -18943.583 0.011 0.015
Chain 1: 3300 -19148.569 0.010 0.012
Chain 1: 3400 -18622.871 0.012 0.015
Chain 1: 3500 -19235.720 0.014 0.015
Chain 1: 3600 -18541.127 0.016 0.015
Chain 1: 3700 -18928.865 0.018 0.016
Chain 1: 3800 -17886.615 0.022 0.020
Chain 1: 3900 -17882.696 0.021 0.020
Chain 1: 4000 -18000.006 0.021 0.020
Chain 1: 4100 -17913.665 0.021 0.020
Chain 1: 4200 -17729.489 0.021 0.020
Chain 1: 4300 -17868.194 0.021 0.020
Chain 1: 4400 -17824.657 0.018 0.010
Chain 1: 4500 -17727.126 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001343 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.43 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12219.665 1.000 1.000
Chain 1: 200 -9101.333 0.671 1.000
Chain 1: 300 -7884.928 0.499 0.343
Chain 1: 400 -7993.591 0.378 0.343
Chain 1: 500 -7937.294 0.304 0.154
Chain 1: 600 -7806.918 0.256 0.154
Chain 1: 700 -7978.158 0.222 0.021
Chain 1: 800 -7766.407 0.198 0.027
Chain 1: 900 -7726.809 0.176 0.021
Chain 1: 1000 -7800.513 0.160 0.021
Chain 1: 1100 -7847.796 0.060 0.017
Chain 1: 1200 -7766.024 0.027 0.014
Chain 1: 1300 -7724.808 0.012 0.011
Chain 1: 1400 -7732.351 0.011 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001389 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.89 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57660.202 1.000 1.000
Chain 1: 200 -17416.523 1.655 2.311
Chain 1: 300 -8533.749 1.451 1.041
Chain 1: 400 -8061.914 1.103 1.041
Chain 1: 500 -8367.378 0.889 1.000
Chain 1: 600 -8114.396 0.746 1.000
Chain 1: 700 -8074.468 0.640 0.059
Chain 1: 800 -8019.934 0.561 0.059
Chain 1: 900 -7928.894 0.500 0.037
Chain 1: 1000 -7862.666 0.451 0.037
Chain 1: 1100 -7779.677 0.352 0.031
Chain 1: 1200 -7786.613 0.121 0.011
Chain 1: 1300 -7747.993 0.017 0.011
Chain 1: 1400 -7613.508 0.013 0.011
Chain 1: 1500 -7519.864 0.011 0.011
Chain 1: 1600 -7500.068 0.008 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003049 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.49 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85954.504 1.000 1.000
Chain 1: 200 -13271.029 3.238 5.477
Chain 1: 300 -9667.270 2.283 1.000
Chain 1: 400 -10621.241 1.735 1.000
Chain 1: 500 -8612.007 1.435 0.373
Chain 1: 600 -8262.961 1.202 0.373
Chain 1: 700 -8335.133 1.032 0.233
Chain 1: 800 -8732.229 0.909 0.233
Chain 1: 900 -8545.919 0.810 0.090
Chain 1: 1000 -8226.561 0.733 0.090
Chain 1: 1100 -8550.790 0.637 0.045 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8155.568 0.094 0.045
Chain 1: 1300 -8374.347 0.059 0.042
Chain 1: 1400 -8363.051 0.050 0.039
Chain 1: 1500 -8265.486 0.028 0.038
Chain 1: 1600 -8371.128 0.025 0.026
Chain 1: 1700 -8460.386 0.025 0.026
Chain 1: 1800 -8057.486 0.026 0.026
Chain 1: 1900 -8156.552 0.025 0.026
Chain 1: 2000 -8127.811 0.021 0.013
Chain 1: 2100 -8247.685 0.019 0.013
Chain 1: 2200 -8039.902 0.017 0.013
Chain 1: 2300 -8189.593 0.016 0.013
Chain 1: 2400 -8067.500 0.017 0.015
Chain 1: 2500 -8131.618 0.017 0.015
Chain 1: 2600 -8154.531 0.016 0.015
Chain 1: 2700 -8073.235 0.016 0.015
Chain 1: 2800 -8046.321 0.011 0.012
Chain 1: 2900 -8101.743 0.011 0.010
Chain 1: 3000 -7985.440 0.012 0.015
Chain 1: 3100 -8123.830 0.012 0.015
Chain 1: 3200 -8003.436 0.011 0.015
Chain 1: 3300 -8025.323 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003204 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.04 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8392125.080 1.000 1.000
Chain 1: 200 -1586191.363 2.645 4.291
Chain 1: 300 -891580.228 2.023 1.000
Chain 1: 400 -457576.119 1.755 1.000
Chain 1: 500 -357871.975 1.459 0.948
Chain 1: 600 -232767.437 1.306 0.948
Chain 1: 700 -118991.807 1.256 0.948
Chain 1: 800 -86188.272 1.146 0.948
Chain 1: 900 -66533.672 1.052 0.779
Chain 1: 1000 -51331.451 0.976 0.779
Chain 1: 1100 -38807.755 0.909 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37983.757 0.482 0.381
Chain 1: 1300 -25946.707 0.450 0.381
Chain 1: 1400 -25665.555 0.356 0.323
Chain 1: 1500 -22253.875 0.344 0.323
Chain 1: 1600 -21470.117 0.294 0.296
Chain 1: 1700 -20345.132 0.204 0.295
Chain 1: 1800 -20289.458 0.166 0.153
Chain 1: 1900 -20615.335 0.138 0.055
Chain 1: 2000 -19127.346 0.116 0.055
Chain 1: 2100 -19365.841 0.085 0.037
Chain 1: 2200 -19591.919 0.084 0.037
Chain 1: 2300 -19209.483 0.040 0.020
Chain 1: 2400 -18981.654 0.040 0.020
Chain 1: 2500 -18783.567 0.025 0.016
Chain 1: 2600 -18414.201 0.024 0.016
Chain 1: 2700 -18371.265 0.019 0.012
Chain 1: 2800 -18088.163 0.020 0.016
Chain 1: 2900 -18369.309 0.020 0.015
Chain 1: 3000 -18355.562 0.012 0.012
Chain 1: 3100 -18440.502 0.011 0.012
Chain 1: 3200 -18131.400 0.012 0.015
Chain 1: 3300 -18335.937 0.011 0.012
Chain 1: 3400 -17811.197 0.013 0.015
Chain 1: 3500 -18422.568 0.015 0.016
Chain 1: 3600 -17729.906 0.017 0.016
Chain 1: 3700 -18116.208 0.019 0.017
Chain 1: 3800 -17076.924 0.023 0.021
Chain 1: 3900 -17073.064 0.022 0.021
Chain 1: 4000 -17190.393 0.022 0.021
Chain 1: 4100 -17104.188 0.022 0.021
Chain 1: 4200 -16920.651 0.022 0.021
Chain 1: 4300 -17058.912 0.021 0.021
Chain 1: 4400 -17015.924 0.019 0.011
Chain 1: 4500 -16918.468 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001245 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.45 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49340.325 1.000 1.000
Chain 1: 200 -24298.982 1.015 1.031
Chain 1: 300 -19607.977 0.757 1.000
Chain 1: 400 -19899.263 0.571 1.000
Chain 1: 500 -14771.464 0.526 0.347
Chain 1: 600 -26811.648 0.513 0.449
Chain 1: 700 -13371.093 0.584 0.449
Chain 1: 800 -24514.549 0.568 0.455
Chain 1: 900 -11007.494 0.641 0.455
Chain 1: 1000 -13308.269 0.594 0.455
Chain 1: 1100 -14040.327 0.499 0.449
Chain 1: 1200 -16825.015 0.413 0.347
Chain 1: 1300 -12418.931 0.424 0.355
Chain 1: 1400 -10828.595 0.438 0.355
Chain 1: 1500 -10641.875 0.405 0.355
Chain 1: 1600 -11792.732 0.369 0.173
Chain 1: 1700 -10586.043 0.280 0.166
Chain 1: 1800 -12158.419 0.248 0.147
Chain 1: 1900 -12401.842 0.127 0.129
Chain 1: 2000 -11032.610 0.122 0.124
Chain 1: 2100 -11301.606 0.119 0.124
Chain 1: 2200 -11683.744 0.106 0.114
Chain 1: 2300 -10335.726 0.084 0.114
Chain 1: 2400 -11717.289 0.081 0.114
Chain 1: 2500 -10617.219 0.089 0.114
Chain 1: 2600 -13012.391 0.098 0.118
Chain 1: 2700 -13661.741 0.091 0.118
Chain 1: 2800 -16617.191 0.096 0.118
Chain 1: 2900 -14557.882 0.108 0.124
Chain 1: 3000 -16972.476 0.110 0.130
Chain 1: 3100 -12750.360 0.141 0.141
Chain 1: 3200 -10435.589 0.160 0.142
Chain 1: 3300 -9797.670 0.153 0.142
Chain 1: 3400 -10655.996 0.150 0.142
Chain 1: 3500 -10438.818 0.141 0.142
Chain 1: 3600 -18501.275 0.166 0.142
Chain 1: 3700 -10267.012 0.242 0.178
Chain 1: 3800 -9180.492 0.236 0.142
Chain 1: 3900 -9303.724 0.223 0.142
Chain 1: 4000 -9430.508 0.210 0.118
Chain 1: 4100 -9470.621 0.178 0.081
Chain 1: 4200 -10925.262 0.169 0.081
Chain 1: 4300 -13705.157 0.182 0.118
Chain 1: 4400 -9736.312 0.215 0.133
Chain 1: 4500 -11246.258 0.226 0.134
Chain 1: 4600 -10177.074 0.193 0.133
Chain 1: 4700 -11842.618 0.127 0.133
Chain 1: 4800 -9060.171 0.146 0.134
Chain 1: 4900 -9495.890 0.149 0.134
Chain 1: 5000 -15451.632 0.187 0.141
Chain 1: 5100 -9358.847 0.251 0.203
Chain 1: 5200 -9584.459 0.240 0.203
Chain 1: 5300 -14628.818 0.255 0.307
Chain 1: 5400 -8846.853 0.279 0.307
Chain 1: 5500 -14917.191 0.306 0.345
Chain 1: 5600 -13587.456 0.306 0.345
Chain 1: 5700 -12764.963 0.298 0.345
Chain 1: 5800 -13479.273 0.273 0.345
Chain 1: 5900 -11324.206 0.287 0.345
Chain 1: 6000 -8912.693 0.276 0.271
Chain 1: 6100 -9822.457 0.220 0.190
Chain 1: 6200 -10365.862 0.223 0.190
Chain 1: 6300 -9498.923 0.197 0.098
Chain 1: 6400 -12781.810 0.158 0.098
Chain 1: 6500 -12812.484 0.117 0.093
Chain 1: 6600 -8916.847 0.151 0.093
Chain 1: 6700 -9518.501 0.151 0.093
Chain 1: 6800 -9320.539 0.148 0.093
Chain 1: 6900 -9517.915 0.131 0.091
Chain 1: 7000 -8778.888 0.112 0.084
Chain 1: 7100 -8684.132 0.104 0.063
Chain 1: 7200 -10858.032 0.119 0.084
Chain 1: 7300 -11966.535 0.119 0.084
Chain 1: 7400 -9615.719 0.118 0.084
Chain 1: 7500 -11148.988 0.131 0.093
Chain 1: 7600 -8654.914 0.116 0.093
Chain 1: 7700 -8864.849 0.112 0.093
Chain 1: 7800 -9460.619 0.117 0.093
Chain 1: 7900 -8995.896 0.120 0.093
Chain 1: 8000 -8526.051 0.117 0.093
Chain 1: 8100 -9479.361 0.126 0.101
Chain 1: 8200 -8936.850 0.112 0.093
Chain 1: 8300 -8859.576 0.103 0.063
Chain 1: 8400 -10338.356 0.093 0.063
Chain 1: 8500 -9511.503 0.088 0.063
Chain 1: 8600 -8937.156 0.066 0.063
Chain 1: 8700 -9295.154 0.067 0.063
Chain 1: 8800 -8718.603 0.068 0.064
Chain 1: 8900 -14132.523 0.101 0.066
Chain 1: 9000 -9418.813 0.145 0.087
Chain 1: 9100 -8625.799 0.144 0.087
Chain 1: 9200 -8871.349 0.141 0.087
Chain 1: 9300 -8960.434 0.141 0.087
Chain 1: 9400 -9217.794 0.130 0.066
Chain 1: 9500 -9424.335 0.123 0.064
Chain 1: 9600 -9312.728 0.118 0.039
Chain 1: 9700 -8561.758 0.123 0.066
Chain 1: 9800 -10026.388 0.131 0.088
Chain 1: 9900 -11966.677 0.109 0.088
Chain 1: 10000 -8568.409 0.098 0.088
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001389 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.89 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -47166.505 1.000 1.000
Chain 1: 200 -16093.626 1.465 1.931
Chain 1: 300 -8974.766 1.241 1.000
Chain 1: 400 -8825.062 0.935 1.000
Chain 1: 500 -9073.969 0.754 0.793
Chain 1: 600 -8880.709 0.632 0.793
Chain 1: 700 -8165.150 0.554 0.088
Chain 1: 800 -8318.259 0.487 0.088
Chain 1: 900 -8044.030 0.437 0.034
Chain 1: 1000 -7871.469 0.395 0.034
Chain 1: 1100 -7776.916 0.296 0.027
Chain 1: 1200 -7576.578 0.106 0.026
Chain 1: 1300 -7603.948 0.027 0.022
Chain 1: 1400 -8119.820 0.032 0.026
Chain 1: 1500 -7571.440 0.036 0.026
Chain 1: 1600 -7864.814 0.038 0.034
Chain 1: 1700 -7575.901 0.033 0.034
Chain 1: 1800 -7632.387 0.032 0.034
Chain 1: 1900 -7590.627 0.029 0.026
Chain 1: 2000 -7767.421 0.029 0.026
Chain 1: 2100 -7622.023 0.030 0.026
Chain 1: 2200 -7812.322 0.029 0.024
Chain 1: 2300 -7677.421 0.031 0.024
Chain 1: 2400 -7586.258 0.026 0.023
Chain 1: 2500 -7709.196 0.020 0.019
Chain 1: 2600 -7584.205 0.018 0.018
Chain 1: 2700 -7483.924 0.015 0.016
Chain 1: 2800 -7690.261 0.017 0.018
Chain 1: 2900 -7485.464 0.020 0.019
Chain 1: 3000 -7597.260 0.019 0.018
Chain 1: 3100 -7584.547 0.017 0.016
Chain 1: 3200 -7788.753 0.017 0.016
Chain 1: 3300 -7491.102 0.019 0.016
Chain 1: 3400 -7736.198 0.021 0.026
Chain 1: 3500 -7484.801 0.023 0.027
Chain 1: 3600 -7551.601 0.022 0.027
Chain 1: 3700 -7502.092 0.022 0.027
Chain 1: 3800 -7475.451 0.019 0.026
Chain 1: 3900 -7453.681 0.017 0.015
Chain 1: 4000 -7449.320 0.016 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003201 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.01 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85687.695 1.000 1.000
Chain 1: 200 -14105.095 3.037 5.075
Chain 1: 300 -10396.529 2.144 1.000
Chain 1: 400 -11522.683 1.632 1.000
Chain 1: 500 -9304.334 1.354 0.357
Chain 1: 600 -9761.256 1.136 0.357
Chain 1: 700 -9020.933 0.985 0.238
Chain 1: 800 -9310.550 0.866 0.238
Chain 1: 900 -9117.770 0.772 0.098
Chain 1: 1000 -9182.596 0.696 0.098
Chain 1: 1100 -8943.116 0.598 0.082 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8761.233 0.093 0.047
Chain 1: 1300 -9066.150 0.061 0.034
Chain 1: 1400 -9023.756 0.051 0.031
Chain 1: 1500 -8888.501 0.029 0.027
Chain 1: 1600 -8998.393 0.025 0.021
Chain 1: 1700 -9063.917 0.018 0.021
Chain 1: 1800 -8624.231 0.020 0.021
Chain 1: 1900 -8729.744 0.019 0.015
Chain 1: 2000 -8710.996 0.019 0.015
Chain 1: 2100 -8837.254 0.017 0.014
Chain 1: 2200 -8629.915 0.018 0.014
Chain 1: 2300 -8722.824 0.015 0.012
Chain 1: 2400 -8789.560 0.016 0.012
Chain 1: 2500 -8738.445 0.015 0.012
Chain 1: 2600 -8750.088 0.014 0.011
Chain 1: 2700 -8659.176 0.014 0.011
Chain 1: 2800 -8608.455 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002531 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.31 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8367669.252 1.000 1.000
Chain 1: 200 -1578406.142 2.651 4.301
Chain 1: 300 -890971.059 2.024 1.000
Chain 1: 400 -458231.685 1.754 1.000
Chain 1: 500 -359393.495 1.458 0.944
Chain 1: 600 -234489.296 1.304 0.944
Chain 1: 700 -120326.347 1.253 0.944
Chain 1: 800 -87459.176 1.144 0.944
Chain 1: 900 -67713.400 1.049 0.772
Chain 1: 1000 -52445.253 0.973 0.772
Chain 1: 1100 -39852.404 0.905 0.533 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39023.288 0.477 0.376
Chain 1: 1300 -26894.550 0.445 0.376
Chain 1: 1400 -26607.844 0.351 0.316
Chain 1: 1500 -23173.085 0.339 0.316
Chain 1: 1600 -22383.809 0.289 0.292
Chain 1: 1700 -21246.581 0.199 0.291
Chain 1: 1800 -21188.502 0.162 0.148
Chain 1: 1900 -21515.143 0.135 0.054
Chain 1: 2000 -20019.713 0.113 0.054
Chain 1: 2100 -20258.390 0.082 0.035
Chain 1: 2200 -20486.226 0.081 0.035
Chain 1: 2300 -20102.104 0.038 0.019
Chain 1: 2400 -19873.887 0.038 0.019
Chain 1: 2500 -19676.363 0.024 0.015
Chain 1: 2600 -19305.653 0.023 0.015
Chain 1: 2700 -19262.316 0.018 0.012
Chain 1: 2800 -18979.179 0.019 0.015
Chain 1: 2900 -19260.749 0.019 0.015
Chain 1: 3000 -19246.802 0.012 0.012
Chain 1: 3100 -19331.899 0.011 0.011
Chain 1: 3200 -19022.166 0.011 0.015
Chain 1: 3300 -19227.203 0.010 0.011
Chain 1: 3400 -18701.554 0.012 0.015
Chain 1: 3500 -19314.489 0.014 0.015
Chain 1: 3600 -18619.811 0.016 0.015
Chain 1: 3700 -19007.696 0.018 0.016
Chain 1: 3800 -17965.437 0.022 0.020
Chain 1: 3900 -17961.596 0.021 0.020
Chain 1: 4000 -18078.828 0.021 0.020
Chain 1: 4100 -17992.548 0.021 0.020
Chain 1: 4200 -17808.331 0.021 0.020
Chain 1: 4300 -17947.004 0.021 0.020
Chain 1: 4400 -17903.458 0.018 0.010
Chain 1: 4500 -17805.978 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001169 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 11.69 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12443.718 1.000 1.000
Chain 1: 200 -9369.201 0.664 1.000
Chain 1: 300 -8142.888 0.493 0.328
Chain 1: 400 -8359.471 0.376 0.328
Chain 1: 500 -8184.910 0.305 0.151
Chain 1: 600 -8098.140 0.256 0.151
Chain 1: 700 -8009.880 0.221 0.026
Chain 1: 800 -8015.639 0.194 0.026
Chain 1: 900 -7925.632 0.173 0.021
Chain 1: 1000 -8116.437 0.158 0.024
Chain 1: 1100 -8151.062 0.059 0.021
Chain 1: 1200 -8044.558 0.027 0.013
Chain 1: 1300 -7986.591 0.013 0.011
Chain 1: 1400 -8003.802 0.011 0.011
Chain 1: 1500 -8090.630 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001387 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.87 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61854.908 1.000 1.000
Chain 1: 200 -18041.242 1.714 2.429
Chain 1: 300 -8919.832 1.484 1.023
Chain 1: 400 -9573.322 1.130 1.023
Chain 1: 500 -7889.539 0.947 1.000
Chain 1: 600 -9023.567 0.810 1.000
Chain 1: 700 -8384.726 0.705 0.213
Chain 1: 800 -8605.387 0.620 0.213
Chain 1: 900 -7900.964 0.561 0.126
Chain 1: 1000 -7750.071 0.507 0.126
Chain 1: 1100 -7757.298 0.407 0.089
Chain 1: 1200 -7661.504 0.165 0.076
Chain 1: 1300 -7777.212 0.065 0.068
Chain 1: 1400 -7643.479 0.060 0.026
Chain 1: 1500 -7583.576 0.039 0.019
Chain 1: 1600 -7738.110 0.028 0.019
Chain 1: 1700 -7528.481 0.024 0.019
Chain 1: 1800 -7620.073 0.022 0.017
Chain 1: 1900 -7603.030 0.014 0.015
Chain 1: 2000 -7661.811 0.012 0.013
Chain 1: 2100 -7604.821 0.013 0.013
Chain 1: 2200 -7711.344 0.013 0.014
Chain 1: 2300 -7564.939 0.014 0.014
Chain 1: 2400 -7616.960 0.013 0.012
Chain 1: 2500 -7621.879 0.012 0.012
Chain 1: 2600 -7532.486 0.011 0.012
Chain 1: 2700 -7554.759 0.008 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002824 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 28.24 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85717.965 1.000 1.000
Chain 1: 200 -13636.327 3.143 5.286
Chain 1: 300 -9993.566 2.217 1.000
Chain 1: 400 -10903.019 1.683 1.000
Chain 1: 500 -8969.812 1.390 0.365
Chain 1: 600 -8488.784 1.168 0.365
Chain 1: 700 -8722.424 1.005 0.216
Chain 1: 800 -8921.657 0.882 0.216
Chain 1: 900 -8806.658 0.785 0.083
Chain 1: 1000 -8700.129 0.708 0.083
Chain 1: 1100 -8853.466 0.610 0.057 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8462.789 0.086 0.046
Chain 1: 1300 -8687.121 0.052 0.027
Chain 1: 1400 -8698.810 0.044 0.026
Chain 1: 1500 -8551.230 0.024 0.022
Chain 1: 1600 -8664.643 0.020 0.017
Chain 1: 1700 -8745.707 0.018 0.017
Chain 1: 1800 -8329.478 0.021 0.017
Chain 1: 1900 -8427.019 0.020 0.017
Chain 1: 2000 -8400.725 0.019 0.017
Chain 1: 2100 -8524.153 0.019 0.014
Chain 1: 2200 -8341.758 0.017 0.014
Chain 1: 2300 -8421.636 0.015 0.013
Chain 1: 2400 -8491.312 0.016 0.013
Chain 1: 2500 -8437.101 0.015 0.012
Chain 1: 2600 -8437.075 0.013 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00255 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.5 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8385687.024 1.000 1.000
Chain 1: 200 -1584103.873 2.647 4.294
Chain 1: 300 -892066.447 2.023 1.000
Chain 1: 400 -458256.427 1.754 1.000
Chain 1: 500 -358816.020 1.459 0.947
Chain 1: 600 -233675.200 1.305 0.947
Chain 1: 700 -119644.057 1.255 0.947
Chain 1: 800 -86771.522 1.145 0.947
Chain 1: 900 -67069.210 1.050 0.776
Chain 1: 1000 -51832.393 0.975 0.776
Chain 1: 1100 -39268.356 0.907 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38440.677 0.480 0.379
Chain 1: 1300 -26356.730 0.448 0.379
Chain 1: 1400 -26071.656 0.354 0.320
Chain 1: 1500 -22648.465 0.342 0.320
Chain 1: 1600 -21861.516 0.292 0.294
Chain 1: 1700 -20730.740 0.202 0.294
Chain 1: 1800 -20673.766 0.164 0.151
Chain 1: 1900 -20999.904 0.136 0.055
Chain 1: 2000 -19508.565 0.115 0.055
Chain 1: 2100 -19747.038 0.084 0.036
Chain 1: 2200 -19973.908 0.083 0.036
Chain 1: 2300 -19590.781 0.039 0.020
Chain 1: 2400 -19362.860 0.039 0.020
Chain 1: 2500 -19164.985 0.025 0.016
Chain 1: 2600 -18795.079 0.023 0.016
Chain 1: 2700 -18751.996 0.018 0.012
Chain 1: 2800 -18468.899 0.019 0.015
Chain 1: 2900 -18750.219 0.019 0.015
Chain 1: 3000 -18736.395 0.012 0.012
Chain 1: 3100 -18821.375 0.011 0.012
Chain 1: 3200 -18512.026 0.012 0.015
Chain 1: 3300 -18716.767 0.011 0.012
Chain 1: 3400 -18191.672 0.012 0.015
Chain 1: 3500 -18803.624 0.015 0.015
Chain 1: 3600 -18110.289 0.017 0.015
Chain 1: 3700 -18497.131 0.018 0.017
Chain 1: 3800 -17456.793 0.023 0.021
Chain 1: 3900 -17452.971 0.021 0.021
Chain 1: 4000 -17570.249 0.022 0.021
Chain 1: 4100 -17484.018 0.022 0.021
Chain 1: 4200 -17300.264 0.021 0.021
Chain 1: 4300 -17438.638 0.021 0.021
Chain 1: 4400 -17395.460 0.018 0.011
Chain 1: 4500 -17298.039 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001256 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.56 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12687.882 1.000 1.000
Chain 1: 200 -9573.784 0.663 1.000
Chain 1: 300 -8187.872 0.498 0.325
Chain 1: 400 -8450.125 0.381 0.325
Chain 1: 500 -8306.594 0.309 0.169
Chain 1: 600 -8155.304 0.260 0.169
Chain 1: 700 -8053.476 0.225 0.031
Chain 1: 800 -8056.386 0.197 0.031
Chain 1: 900 -7984.959 0.176 0.019
Chain 1: 1000 -8179.525 0.161 0.024
Chain 1: 1100 -8203.439 0.061 0.019
Chain 1: 1200 -8070.390 0.030 0.017
Chain 1: 1300 -8039.636 0.014 0.016
Chain 1: 1400 -8047.153 0.011 0.013
Chain 1: 1500 -8137.219 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001446 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.46 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57372.728 1.000 1.000
Chain 1: 200 -17741.172 1.617 2.234
Chain 1: 300 -8898.269 1.409 1.000
Chain 1: 400 -8236.961 1.077 1.000
Chain 1: 500 -8958.323 0.878 0.994
Chain 1: 600 -8799.973 0.734 0.994
Chain 1: 700 -8553.782 0.634 0.081
Chain 1: 800 -8134.088 0.561 0.081
Chain 1: 900 -7681.956 0.505 0.080
Chain 1: 1000 -7746.593 0.455 0.080
Chain 1: 1100 -7898.347 0.357 0.059
Chain 1: 1200 -7540.854 0.139 0.052
Chain 1: 1300 -7777.979 0.042 0.047
Chain 1: 1400 -7581.358 0.037 0.030
Chain 1: 1500 -7592.238 0.029 0.029
Chain 1: 1600 -7778.451 0.030 0.029
Chain 1: 1700 -7593.503 0.029 0.026
Chain 1: 1800 -7694.674 0.025 0.024
Chain 1: 1900 -7590.518 0.021 0.024
Chain 1: 2000 -7739.713 0.022 0.024
Chain 1: 2100 -7576.455 0.022 0.024
Chain 1: 2200 -7756.993 0.020 0.023
Chain 1: 2300 -7546.258 0.019 0.023
Chain 1: 2400 -7550.922 0.017 0.022
Chain 1: 2500 -7441.068 0.018 0.022
Chain 1: 2600 -7537.862 0.017 0.019
Chain 1: 2700 -7542.521 0.015 0.015
Chain 1: 2800 -7522.115 0.014 0.015
Chain 1: 2900 -7396.734 0.014 0.017
Chain 1: 3000 -7544.256 0.014 0.017
Chain 1: 3100 -7538.210 0.012 0.015
Chain 1: 3200 -7737.396 0.012 0.015
Chain 1: 3300 -7458.086 0.013 0.015
Chain 1: 3400 -7684.900 0.016 0.017
Chain 1: 3500 -7442.037 0.018 0.020
Chain 1: 3600 -7508.462 0.017 0.020
Chain 1: 3700 -7457.579 0.018 0.020
Chain 1: 3800 -7456.067 0.018 0.020
Chain 1: 3900 -7421.910 0.017 0.020
Chain 1: 4000 -7418.802 0.015 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003345 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.45 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87049.163 1.000 1.000
Chain 1: 200 -13819.518 3.150 5.299
Chain 1: 300 -10115.490 2.222 1.000
Chain 1: 400 -11483.303 1.696 1.000
Chain 1: 500 -9123.867 1.409 0.366
Chain 1: 600 -8510.181 1.186 0.366
Chain 1: 700 -8567.828 1.017 0.259
Chain 1: 800 -8854.677 0.894 0.259
Chain 1: 900 -9016.360 0.797 0.119
Chain 1: 1000 -8911.959 0.718 0.119
Chain 1: 1100 -8730.902 0.620 0.072 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8504.947 0.093 0.032
Chain 1: 1300 -8810.789 0.060 0.032
Chain 1: 1400 -8793.705 0.048 0.027
Chain 1: 1500 -8643.624 0.024 0.021
Chain 1: 1600 -8753.590 0.018 0.018
Chain 1: 1700 -8827.574 0.018 0.018
Chain 1: 1800 -8394.719 0.020 0.018
Chain 1: 1900 -8498.916 0.020 0.017
Chain 1: 2000 -8474.351 0.019 0.017
Chain 1: 2100 -8453.538 0.017 0.013
Chain 1: 2200 -8417.916 0.015 0.012
Chain 1: 2300 -8546.865 0.013 0.012
Chain 1: 2400 -8401.415 0.014 0.013
Chain 1: 2500 -8469.649 0.013 0.012
Chain 1: 2600 -8389.425 0.013 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003755 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 37.55 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8432328.872 1.000 1.000
Chain 1: 200 -1592318.892 2.648 4.296
Chain 1: 300 -893250.448 2.026 1.000
Chain 1: 400 -458942.125 1.756 1.000
Chain 1: 500 -358582.918 1.461 0.946
Chain 1: 600 -233083.109 1.307 0.946
Chain 1: 700 -119368.203 1.257 0.946
Chain 1: 800 -86619.699 1.147 0.946
Chain 1: 900 -66995.710 1.052 0.783
Chain 1: 1000 -51831.794 0.976 0.783
Chain 1: 1100 -39344.467 0.908 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38525.867 0.480 0.378
Chain 1: 1300 -26512.433 0.447 0.378
Chain 1: 1400 -26235.538 0.354 0.317
Chain 1: 1500 -22830.726 0.341 0.317
Chain 1: 1600 -22050.165 0.290 0.293
Chain 1: 1700 -20927.234 0.200 0.293
Chain 1: 1800 -20872.274 0.163 0.149
Chain 1: 1900 -21198.763 0.135 0.054
Chain 1: 2000 -19710.958 0.113 0.054
Chain 1: 2100 -19949.319 0.083 0.035
Chain 1: 2200 -20175.809 0.082 0.035
Chain 1: 2300 -19792.860 0.038 0.019
Chain 1: 2400 -19564.862 0.039 0.019
Chain 1: 2500 -19366.750 0.025 0.015
Chain 1: 2600 -18996.801 0.023 0.015
Chain 1: 2700 -18953.644 0.018 0.012
Chain 1: 2800 -18670.354 0.019 0.015
Chain 1: 2900 -18951.638 0.019 0.015
Chain 1: 3000 -18937.856 0.012 0.012
Chain 1: 3100 -19022.943 0.011 0.012
Chain 1: 3200 -18713.417 0.011 0.015
Chain 1: 3300 -18918.257 0.011 0.012
Chain 1: 3400 -18392.820 0.012 0.015
Chain 1: 3500 -19005.246 0.015 0.015
Chain 1: 3600 -18311.114 0.016 0.015
Chain 1: 3700 -18698.509 0.018 0.017
Chain 1: 3800 -17657.025 0.023 0.021
Chain 1: 3900 -17653.090 0.021 0.021
Chain 1: 4000 -17770.420 0.022 0.021
Chain 1: 4100 -17684.188 0.022 0.021
Chain 1: 4200 -17500.090 0.021 0.021
Chain 1: 4300 -17638.733 0.021 0.021
Chain 1: 4400 -17595.343 0.018 0.011
Chain 1: 4500 -17497.791 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001135 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 11.35 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49530.385 1.000 1.000
Chain 1: 200 -20346.898 1.217 1.434
Chain 1: 300 -14833.190 0.935 1.000
Chain 1: 400 -15146.575 0.707 1.000
Chain 1: 500 -13972.568 0.582 0.372
Chain 1: 600 -13417.867 0.492 0.372
Chain 1: 700 -15781.310 0.443 0.150
Chain 1: 800 -14096.218 0.403 0.150
Chain 1: 900 -17668.958 0.380 0.150
Chain 1: 1000 -12222.783 0.387 0.202
Chain 1: 1100 -13845.401 0.299 0.150
Chain 1: 1200 -10500.860 0.187 0.150
Chain 1: 1300 -10255.282 0.152 0.120
Chain 1: 1400 -10597.538 0.153 0.120
Chain 1: 1500 -10637.958 0.145 0.120
Chain 1: 1600 -10572.893 0.142 0.120
Chain 1: 1700 -13837.339 0.151 0.120
Chain 1: 1800 -19673.397 0.168 0.202
Chain 1: 1900 -10602.275 0.234 0.236
Chain 1: 2000 -10396.384 0.191 0.117
Chain 1: 2100 -9903.275 0.184 0.050
Chain 1: 2200 -9815.112 0.153 0.032
Chain 1: 2300 -9531.874 0.154 0.032
Chain 1: 2400 -9460.297 0.151 0.030
Chain 1: 2500 -18181.088 0.199 0.050
Chain 1: 2600 -17361.694 0.203 0.050
Chain 1: 2700 -12670.860 0.217 0.050
Chain 1: 2800 -9539.030 0.220 0.050
Chain 1: 2900 -10773.512 0.146 0.050
Chain 1: 3000 -8987.424 0.163 0.115
Chain 1: 3100 -14622.496 0.197 0.199
Chain 1: 3200 -9394.956 0.252 0.328
Chain 1: 3300 -9611.092 0.251 0.328
Chain 1: 3400 -8931.505 0.258 0.328
Chain 1: 3500 -9678.101 0.218 0.199
Chain 1: 3600 -9507.541 0.215 0.199
Chain 1: 3700 -9899.228 0.182 0.115
Chain 1: 3800 -9203.012 0.156 0.077
Chain 1: 3900 -9061.486 0.147 0.076
Chain 1: 4000 -9015.890 0.127 0.076
Chain 1: 4100 -9747.308 0.096 0.075
Chain 1: 4200 -9008.768 0.049 0.075
Chain 1: 4300 -9960.357 0.056 0.076
Chain 1: 4400 -9780.905 0.050 0.075
Chain 1: 4500 -9277.518 0.048 0.054
Chain 1: 4600 -9489.877 0.048 0.054
Chain 1: 4700 -8763.770 0.053 0.075
Chain 1: 4800 -13570.191 0.081 0.075
Chain 1: 4900 -10172.511 0.112 0.082
Chain 1: 5000 -9699.026 0.117 0.082
Chain 1: 5100 -9036.548 0.117 0.082
Chain 1: 5200 -9304.136 0.111 0.073
Chain 1: 5300 -9511.336 0.104 0.054
Chain 1: 5400 -10121.655 0.108 0.060
Chain 1: 5500 -8771.943 0.118 0.073
Chain 1: 5600 -14778.853 0.156 0.083
Chain 1: 5700 -8899.366 0.214 0.154
Chain 1: 5800 -8833.354 0.180 0.073
Chain 1: 5900 -9399.814 0.152 0.060
Chain 1: 6000 -9532.914 0.149 0.060
Chain 1: 6100 -9322.701 0.144 0.060
Chain 1: 6200 -8521.619 0.150 0.060
Chain 1: 6300 -12399.885 0.179 0.094
Chain 1: 6400 -14558.680 0.188 0.148
Chain 1: 6500 -8645.572 0.241 0.148
Chain 1: 6600 -9373.782 0.208 0.094
Chain 1: 6700 -9031.974 0.146 0.078
Chain 1: 6800 -8991.313 0.146 0.078
Chain 1: 6900 -8673.881 0.143 0.078
Chain 1: 7000 -10974.022 0.163 0.094
Chain 1: 7100 -12129.789 0.170 0.095
Chain 1: 7200 -8748.042 0.199 0.148
Chain 1: 7300 -9049.040 0.171 0.095
Chain 1: 7400 -10311.872 0.169 0.095
Chain 1: 7500 -11133.208 0.108 0.078
Chain 1: 7600 -9838.663 0.113 0.095
Chain 1: 7700 -8693.213 0.123 0.122
Chain 1: 7800 -9808.493 0.133 0.122
Chain 1: 7900 -8469.628 0.146 0.132
Chain 1: 8000 -8437.371 0.125 0.122
Chain 1: 8100 -8430.560 0.116 0.122
Chain 1: 8200 -10167.789 0.094 0.122
Chain 1: 8300 -8613.909 0.109 0.132
Chain 1: 8400 -9472.735 0.106 0.132
Chain 1: 8500 -8574.627 0.109 0.132
Chain 1: 8600 -8528.454 0.096 0.114
Chain 1: 8700 -8910.326 0.087 0.105
Chain 1: 8800 -8389.521 0.082 0.091
Chain 1: 8900 -10150.372 0.084 0.091
Chain 1: 9000 -11373.166 0.094 0.105
Chain 1: 9100 -8318.174 0.131 0.108
Chain 1: 9200 -8660.400 0.117 0.105
Chain 1: 9300 -9663.909 0.110 0.104
Chain 1: 9400 -13665.939 0.130 0.105
Chain 1: 9500 -8723.521 0.176 0.108
Chain 1: 9600 -8453.194 0.179 0.108
Chain 1: 9700 -9741.148 0.188 0.132
Chain 1: 9800 -9311.628 0.186 0.132
Chain 1: 9900 -10908.556 0.183 0.132
Chain 1: 10000 -8509.376 0.201 0.146
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00147 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.7 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -55901.673 1.000 1.000
Chain 1: 200 -17477.949 1.599 2.198
Chain 1: 300 -8786.873 1.396 1.000
Chain 1: 400 -8449.431 1.057 1.000
Chain 1: 500 -9018.239 0.858 0.989
Chain 1: 600 -8633.093 0.723 0.989
Chain 1: 700 -8358.293 0.624 0.063
Chain 1: 800 -7519.206 0.560 0.112
Chain 1: 900 -8095.889 0.506 0.071
Chain 1: 1000 -7896.198 0.458 0.071
Chain 1: 1100 -7886.691 0.358 0.063
Chain 1: 1200 -7527.562 0.143 0.048
Chain 1: 1300 -7742.968 0.047 0.045
Chain 1: 1400 -7657.502 0.044 0.045
Chain 1: 1500 -7518.697 0.039 0.033
Chain 1: 1600 -7878.402 0.039 0.033
Chain 1: 1700 -7690.420 0.038 0.028
Chain 1: 1800 -7621.976 0.028 0.025
Chain 1: 1900 -7751.701 0.023 0.024
Chain 1: 2000 -7605.994 0.022 0.019
Chain 1: 2100 -7548.705 0.023 0.019
Chain 1: 2200 -7711.530 0.020 0.019
Chain 1: 2300 -7534.442 0.020 0.019
Chain 1: 2400 -7586.077 0.019 0.019
Chain 1: 2500 -7598.259 0.018 0.019
Chain 1: 2600 -7498.442 0.014 0.017
Chain 1: 2700 -7525.380 0.012 0.013
Chain 1: 2800 -7475.024 0.012 0.013
Chain 1: 2900 -7385.804 0.012 0.012
Chain 1: 3000 -7517.605 0.011 0.012
Chain 1: 3100 -7508.895 0.011 0.012
Chain 1: 3200 -7709.063 0.011 0.012
Chain 1: 3300 -7435.847 0.013 0.012
Chain 1: 3400 -7652.735 0.015 0.013
Chain 1: 3500 -7415.716 0.018 0.018
Chain 1: 3600 -7482.012 0.017 0.018
Chain 1: 3700 -7430.603 0.018 0.018
Chain 1: 3800 -7430.170 0.017 0.018
Chain 1: 3900 -7397.198 0.016 0.018
Chain 1: 4000 -7391.820 0.015 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002618 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 26.18 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86170.790 1.000 1.000
Chain 1: 200 -13727.086 3.139 5.277
Chain 1: 300 -10091.083 2.213 1.000
Chain 1: 400 -11165.712 1.683 1.000
Chain 1: 500 -8939.192 1.397 0.360
Chain 1: 600 -8532.379 1.172 0.360
Chain 1: 700 -8964.193 1.011 0.249
Chain 1: 800 -8847.313 0.887 0.249
Chain 1: 900 -8924.538 0.789 0.096
Chain 1: 1000 -8835.740 0.711 0.096
Chain 1: 1100 -8922.203 0.612 0.048 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8724.001 0.087 0.048
Chain 1: 1300 -8786.403 0.051 0.023
Chain 1: 1400 -8781.870 0.042 0.013
Chain 1: 1500 -8651.963 0.018 0.013
Chain 1: 1600 -8759.980 0.015 0.012
Chain 1: 1700 -8843.126 0.011 0.010
Chain 1: 1800 -8426.563 0.014 0.010
Chain 1: 1900 -8524.095 0.015 0.011
Chain 1: 2000 -8497.882 0.014 0.011
Chain 1: 2100 -8621.431 0.015 0.012
Chain 1: 2200 -8438.329 0.014 0.012
Chain 1: 2300 -8518.728 0.015 0.012
Chain 1: 2400 -8588.387 0.015 0.012
Chain 1: 2500 -8534.224 0.015 0.011
Chain 1: 2600 -8534.262 0.013 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003114 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.14 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8407984.250 1.000 1.000
Chain 1: 200 -1584473.449 2.653 4.306
Chain 1: 300 -891413.572 2.028 1.000
Chain 1: 400 -458378.801 1.757 1.000
Chain 1: 500 -358754.466 1.461 0.945
Chain 1: 600 -233567.145 1.307 0.945
Chain 1: 700 -119583.888 1.257 0.945
Chain 1: 800 -86801.067 1.147 0.945
Chain 1: 900 -67096.906 1.052 0.777
Chain 1: 1000 -51873.324 0.976 0.777
Chain 1: 1100 -39335.630 0.908 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38506.979 0.479 0.378
Chain 1: 1300 -26443.244 0.447 0.378
Chain 1: 1400 -26160.119 0.354 0.319
Chain 1: 1500 -22743.688 0.341 0.319
Chain 1: 1600 -21959.470 0.291 0.294
Chain 1: 1700 -20830.362 0.201 0.293
Chain 1: 1800 -20774.013 0.164 0.150
Chain 1: 1900 -21100.129 0.136 0.054
Chain 1: 2000 -19610.218 0.114 0.054
Chain 1: 2100 -19848.408 0.083 0.036
Chain 1: 2200 -20075.310 0.082 0.036
Chain 1: 2300 -19692.148 0.039 0.019
Chain 1: 2400 -19464.214 0.039 0.019
Chain 1: 2500 -19266.512 0.025 0.015
Chain 1: 2600 -18896.411 0.023 0.015
Chain 1: 2700 -18853.277 0.018 0.012
Chain 1: 2800 -18570.290 0.019 0.015
Chain 1: 2900 -18851.546 0.019 0.015
Chain 1: 3000 -18837.605 0.012 0.012
Chain 1: 3100 -18922.666 0.011 0.012
Chain 1: 3200 -18613.265 0.012 0.015
Chain 1: 3300 -18818.044 0.011 0.012
Chain 1: 3400 -18292.982 0.012 0.015
Chain 1: 3500 -18904.950 0.015 0.015
Chain 1: 3600 -18211.457 0.016 0.015
Chain 1: 3700 -18598.418 0.018 0.017
Chain 1: 3800 -17558.008 0.023 0.021
Chain 1: 3900 -17554.187 0.021 0.021
Chain 1: 4000 -17671.432 0.022 0.021
Chain 1: 4100 -17585.258 0.022 0.021
Chain 1: 4200 -17401.439 0.021 0.021
Chain 1: 4300 -17539.832 0.021 0.021
Chain 1: 4400 -17496.603 0.018 0.011
Chain 1: 4500 -17399.178 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001269 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.69 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12453.052 1.000 1.000
Chain 1: 200 -9463.986 0.658 1.000
Chain 1: 300 -7890.167 0.505 0.316
Chain 1: 400 -8149.632 0.387 0.316
Chain 1: 500 -8024.731 0.313 0.199
Chain 1: 600 -7882.142 0.263 0.199
Chain 1: 700 -7989.969 0.228 0.032
Chain 1: 800 -7815.151 0.202 0.032
Chain 1: 900 -7985.529 0.182 0.022
Chain 1: 1000 -7790.033 0.166 0.025
Chain 1: 1100 -7891.671 0.068 0.022
Chain 1: 1200 -7803.828 0.037 0.021
Chain 1: 1300 -7727.794 0.018 0.018
Chain 1: 1400 -7766.940 0.015 0.016
Chain 1: 1500 -7860.787 0.015 0.013
Chain 1: 1600 -7778.086 0.014 0.013
Chain 1: 1700 -7736.446 0.014 0.012
Chain 1: 1800 -7707.662 0.012 0.011
Chain 1: 1900 -7734.518 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001569 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.69 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57267.214 1.000 1.000
Chain 1: 200 -17685.287 1.619 2.238
Chain 1: 300 -8810.365 1.415 1.007
Chain 1: 400 -8238.502 1.079 1.007
Chain 1: 500 -8446.443 0.868 1.000
Chain 1: 600 -8492.695 0.724 1.000
Chain 1: 700 -7834.322 0.633 0.084
Chain 1: 800 -8195.494 0.559 0.084
Chain 1: 900 -7837.062 0.502 0.069
Chain 1: 1000 -7649.525 0.454 0.069
Chain 1: 1100 -7705.522 0.355 0.046
Chain 1: 1200 -7645.305 0.132 0.044
Chain 1: 1300 -7784.871 0.033 0.025
Chain 1: 1400 -7923.529 0.028 0.025
Chain 1: 1500 -7551.179 0.030 0.025
Chain 1: 1600 -7730.524 0.032 0.025
Chain 1: 1700 -7693.314 0.024 0.023
Chain 1: 1800 -7645.842 0.020 0.018
Chain 1: 1900 -7577.491 0.017 0.017
Chain 1: 2000 -7625.514 0.015 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00262 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 26.2 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86115.500 1.000 1.000
Chain 1: 200 -13657.309 3.153 5.305
Chain 1: 300 -9909.030 2.228 1.000
Chain 1: 400 -11515.397 1.706 1.000
Chain 1: 500 -8612.363 1.432 0.378
Chain 1: 600 -8654.174 1.194 0.378
Chain 1: 700 -8286.424 1.030 0.337
Chain 1: 800 -8506.707 0.904 0.337
Chain 1: 900 -8639.529 0.806 0.139
Chain 1: 1000 -8832.386 0.727 0.139
Chain 1: 1100 -8590.478 0.630 0.044 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8236.613 0.104 0.043
Chain 1: 1300 -8544.859 0.070 0.036
Chain 1: 1400 -8487.792 0.056 0.028
Chain 1: 1500 -8393.700 0.024 0.026
Chain 1: 1600 -8506.424 0.025 0.026
Chain 1: 1700 -8560.170 0.021 0.022
Chain 1: 1800 -8115.045 0.024 0.022
Chain 1: 1900 -8220.655 0.023 0.022
Chain 1: 2000 -8203.917 0.021 0.013
Chain 1: 2100 -8342.845 0.020 0.013
Chain 1: 2200 -8114.835 0.019 0.013
Chain 1: 2300 -8268.696 0.017 0.013
Chain 1: 2400 -8113.502 0.018 0.017
Chain 1: 2500 -8191.236 0.018 0.017
Chain 1: 2600 -8111.916 0.018 0.017
Chain 1: 2700 -8137.192 0.017 0.017
Chain 1: 2800 -8093.225 0.013 0.013
Chain 1: 2900 -8196.671 0.012 0.013
Chain 1: 3000 -8134.257 0.013 0.013
Chain 1: 3100 -8082.604 0.012 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002524 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.24 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8357963.104 1.000 1.000
Chain 1: 200 -1575800.690 2.652 4.304
Chain 1: 300 -890461.630 2.025 1.000
Chain 1: 400 -458251.830 1.754 1.000
Chain 1: 500 -359547.696 1.458 0.943
Chain 1: 600 -234498.036 1.304 0.943
Chain 1: 700 -120130.470 1.254 0.943
Chain 1: 800 -87185.188 1.144 0.943
Chain 1: 900 -67394.706 1.050 0.770
Chain 1: 1000 -52085.690 0.974 0.770
Chain 1: 1100 -39457.334 0.906 0.533 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38625.739 0.478 0.378
Chain 1: 1300 -26462.995 0.447 0.378
Chain 1: 1400 -26173.773 0.354 0.320
Chain 1: 1500 -22729.984 0.341 0.320
Chain 1: 1600 -21938.347 0.292 0.294
Chain 1: 1700 -20796.985 0.202 0.294
Chain 1: 1800 -20738.038 0.165 0.152
Chain 1: 1900 -21064.811 0.137 0.055
Chain 1: 2000 -19566.958 0.115 0.055
Chain 1: 2100 -19805.812 0.084 0.036
Chain 1: 2200 -20034.042 0.083 0.036
Chain 1: 2300 -19649.511 0.039 0.020
Chain 1: 2400 -19421.207 0.039 0.020
Chain 1: 2500 -19223.791 0.025 0.016
Chain 1: 2600 -18852.843 0.023 0.016
Chain 1: 2700 -18809.396 0.018 0.012
Chain 1: 2800 -18526.244 0.019 0.015
Chain 1: 2900 -18807.970 0.019 0.015
Chain 1: 3000 -18793.907 0.012 0.012
Chain 1: 3100 -18879.067 0.011 0.012
Chain 1: 3200 -18569.202 0.012 0.015
Chain 1: 3300 -18774.340 0.011 0.012
Chain 1: 3400 -18248.528 0.012 0.015
Chain 1: 3500 -18861.755 0.015 0.015
Chain 1: 3600 -18166.679 0.017 0.015
Chain 1: 3700 -18554.938 0.018 0.017
Chain 1: 3800 -17512.066 0.023 0.021
Chain 1: 3900 -17508.216 0.021 0.021
Chain 1: 4000 -17625.433 0.022 0.021
Chain 1: 4100 -17539.145 0.022 0.021
Chain 1: 4200 -17354.764 0.021 0.021
Chain 1: 4300 -17493.533 0.021 0.021
Chain 1: 4400 -17449.906 0.018 0.011
Chain 1: 4500 -17352.394 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00125 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.5 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49500.959 1.000 1.000
Chain 1: 200 -17505.749 1.414 1.828
Chain 1: 300 -19753.744 0.980 1.000
Chain 1: 400 -22783.912 0.769 1.000
Chain 1: 500 -19392.432 0.650 0.175
Chain 1: 600 -13884.403 0.608 0.397
Chain 1: 700 -14222.695 0.524 0.175
Chain 1: 800 -12255.103 0.479 0.175
Chain 1: 900 -12897.665 0.431 0.161
Chain 1: 1000 -21997.597 0.429 0.175
Chain 1: 1100 -29605.812 0.355 0.175
Chain 1: 1200 -13059.740 0.299 0.175
Chain 1: 1300 -12000.555 0.296 0.175
Chain 1: 1400 -11283.712 0.290 0.175
Chain 1: 1500 -10891.094 0.276 0.161
Chain 1: 1600 -12178.590 0.247 0.106
Chain 1: 1700 -11591.182 0.249 0.106
Chain 1: 1800 -11042.211 0.238 0.088
Chain 1: 1900 -11633.440 0.238 0.088
Chain 1: 2000 -10295.035 0.210 0.088
Chain 1: 2100 -10166.725 0.185 0.064
Chain 1: 2200 -10206.011 0.059 0.051
Chain 1: 2300 -18624.938 0.096 0.051
Chain 1: 2400 -24676.983 0.114 0.051
Chain 1: 2500 -17640.051 0.150 0.106
Chain 1: 2600 -12212.647 0.184 0.130
Chain 1: 2700 -11987.010 0.181 0.130
Chain 1: 2800 -18468.467 0.211 0.245
Chain 1: 2900 -9598.903 0.298 0.351
Chain 1: 3000 -9503.694 0.286 0.351
Chain 1: 3100 -9594.881 0.286 0.351
Chain 1: 3200 -10295.717 0.292 0.351
Chain 1: 3300 -20154.720 0.296 0.351
Chain 1: 3400 -10930.287 0.356 0.399
Chain 1: 3500 -10078.416 0.324 0.351
Chain 1: 3600 -18500.693 0.325 0.351
Chain 1: 3700 -9527.875 0.418 0.455
Chain 1: 3800 -8997.432 0.389 0.455
Chain 1: 3900 -11965.722 0.321 0.248
Chain 1: 4000 -9244.702 0.349 0.294
Chain 1: 4100 -10124.171 0.357 0.294
Chain 1: 4200 -10696.948 0.356 0.294
Chain 1: 4300 -9190.350 0.323 0.248
Chain 1: 4400 -9887.504 0.246 0.164
Chain 1: 4500 -9126.046 0.246 0.164
Chain 1: 4600 -10149.111 0.210 0.101
Chain 1: 4700 -10443.210 0.119 0.087
Chain 1: 4800 -13858.391 0.138 0.101
Chain 1: 4900 -9240.773 0.163 0.101
Chain 1: 5000 -10027.661 0.141 0.087
Chain 1: 5100 -9585.301 0.137 0.083
Chain 1: 5200 -13750.453 0.162 0.101
Chain 1: 5300 -11459.011 0.166 0.101
Chain 1: 5400 -12051.217 0.164 0.101
Chain 1: 5500 -8984.308 0.189 0.200
Chain 1: 5600 -9401.227 0.184 0.200
Chain 1: 5700 -9405.544 0.181 0.200
Chain 1: 5800 -9923.594 0.161 0.078
Chain 1: 5900 -13922.029 0.140 0.078
Chain 1: 6000 -9202.924 0.184 0.200
Chain 1: 6100 -9702.587 0.184 0.200
Chain 1: 6200 -9245.821 0.159 0.052
Chain 1: 6300 -10456.195 0.150 0.052
Chain 1: 6400 -9327.205 0.158 0.116
Chain 1: 6500 -9287.380 0.124 0.052
Chain 1: 6600 -9521.292 0.122 0.052
Chain 1: 6700 -9248.585 0.125 0.052
Chain 1: 6800 -9400.691 0.121 0.051
Chain 1: 6900 -11800.940 0.113 0.051
Chain 1: 7000 -8885.050 0.094 0.051
Chain 1: 7100 -8671.812 0.092 0.049
Chain 1: 7200 -8577.632 0.088 0.029
Chain 1: 7300 -11664.511 0.103 0.029
Chain 1: 7400 -10045.120 0.107 0.029
Chain 1: 7500 -12618.207 0.127 0.161
Chain 1: 7600 -8742.312 0.169 0.203
Chain 1: 7700 -8889.587 0.167 0.203
Chain 1: 7800 -9178.426 0.169 0.203
Chain 1: 7900 -9266.918 0.149 0.161
Chain 1: 8000 -9478.692 0.119 0.031
Chain 1: 8100 -10778.178 0.128 0.121
Chain 1: 8200 -9527.656 0.140 0.131
Chain 1: 8300 -8575.935 0.125 0.121
Chain 1: 8400 -9271.071 0.116 0.111
Chain 1: 8500 -11158.372 0.113 0.111
Chain 1: 8600 -8883.519 0.094 0.111
Chain 1: 8700 -9295.784 0.097 0.111
Chain 1: 8800 -9127.488 0.096 0.111
Chain 1: 8900 -12793.498 0.123 0.121
Chain 1: 9000 -10000.212 0.149 0.131
Chain 1: 9100 -9216.279 0.146 0.131
Chain 1: 9200 -9175.978 0.133 0.111
Chain 1: 9300 -9659.517 0.127 0.085
Chain 1: 9400 -11373.024 0.134 0.151
Chain 1: 9500 -11083.059 0.120 0.085
Chain 1: 9600 -9958.384 0.106 0.085
Chain 1: 9700 -11376.717 0.114 0.113
Chain 1: 9800 -11343.239 0.112 0.113
Chain 1: 9900 -10189.473 0.095 0.113
Chain 1: 10000 -9359.442 0.076 0.089
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001699 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 16.99 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58739.378 1.000 1.000
Chain 1: 200 -18390.720 1.597 2.194
Chain 1: 300 -8964.342 1.415 1.052
Chain 1: 400 -8065.297 1.089 1.052
Chain 1: 500 -9223.008 0.897 1.000
Chain 1: 600 -9898.176 0.758 1.000
Chain 1: 700 -8252.823 0.679 0.199
Chain 1: 800 -8311.868 0.595 0.199
Chain 1: 900 -7881.844 0.535 0.126
Chain 1: 1000 -7818.947 0.482 0.126
Chain 1: 1100 -7910.991 0.383 0.111
Chain 1: 1200 -7758.183 0.166 0.068
Chain 1: 1300 -7793.897 0.061 0.055
Chain 1: 1400 -7873.381 0.051 0.020
Chain 1: 1500 -7502.360 0.043 0.020
Chain 1: 1600 -7724.384 0.039 0.020
Chain 1: 1700 -7601.773 0.021 0.016
Chain 1: 1800 -7549.682 0.021 0.016
Chain 1: 1900 -7564.548 0.016 0.012
Chain 1: 2000 -7616.048 0.016 0.012
Chain 1: 2100 -7510.107 0.016 0.014
Chain 1: 2200 -7749.431 0.017 0.014
Chain 1: 2300 -7501.017 0.020 0.016
Chain 1: 2400 -7555.761 0.020 0.016
Chain 1: 2500 -7618.200 0.015 0.014
Chain 1: 2600 -7507.406 0.014 0.014
Chain 1: 2700 -7495.226 0.013 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003038 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.38 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86793.524 1.000 1.000
Chain 1: 200 -14117.486 3.074 5.148
Chain 1: 300 -10346.804 2.171 1.000
Chain 1: 400 -11987.242 1.662 1.000
Chain 1: 500 -9010.877 1.396 0.364
Chain 1: 600 -8921.900 1.165 0.364
Chain 1: 700 -8752.583 1.001 0.330
Chain 1: 800 -9009.259 0.880 0.330
Chain 1: 900 -9111.854 0.783 0.137
Chain 1: 1000 -8908.708 0.707 0.137
Chain 1: 1100 -9030.488 0.608 0.028 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8680.757 0.098 0.028
Chain 1: 1300 -8972.404 0.065 0.028
Chain 1: 1400 -8743.308 0.053 0.026
Chain 1: 1500 -8823.747 0.021 0.023
Chain 1: 1600 -8928.787 0.022 0.023
Chain 1: 1700 -8979.344 0.020 0.023
Chain 1: 1800 -8524.572 0.023 0.023
Chain 1: 1900 -8634.847 0.023 0.023
Chain 1: 2000 -8639.058 0.021 0.013
Chain 1: 2100 -8753.844 0.021 0.013
Chain 1: 2200 -8530.505 0.019 0.013
Chain 1: 2300 -8669.973 0.017 0.013
Chain 1: 2400 -8536.929 0.016 0.013
Chain 1: 2500 -8608.935 0.016 0.013
Chain 1: 2600 -8519.747 0.016 0.013
Chain 1: 2700 -8552.104 0.016 0.013
Chain 1: 2800 -8503.655 0.011 0.013
Chain 1: 2900 -8615.862 0.011 0.013
Chain 1: 3000 -8546.967 0.012 0.013
Chain 1: 3100 -8495.639 0.011 0.010
Chain 1: 3200 -8468.636 0.009 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003118 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.18 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8411176.148 1.000 1.000
Chain 1: 200 -1587371.164 2.649 4.299
Chain 1: 300 -890798.778 2.027 1.000
Chain 1: 400 -458084.741 1.756 1.000
Chain 1: 500 -358291.119 1.461 0.945
Chain 1: 600 -233362.835 1.307 0.945
Chain 1: 700 -119748.415 1.255 0.945
Chain 1: 800 -86980.815 1.146 0.945
Chain 1: 900 -67360.467 1.051 0.782
Chain 1: 1000 -52192.593 0.975 0.782
Chain 1: 1100 -39692.251 0.906 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38876.724 0.478 0.377
Chain 1: 1300 -26847.063 0.445 0.377
Chain 1: 1400 -26570.415 0.352 0.315
Chain 1: 1500 -23160.473 0.338 0.315
Chain 1: 1600 -22378.529 0.288 0.291
Chain 1: 1700 -21253.376 0.199 0.291
Chain 1: 1800 -21198.173 0.161 0.147
Chain 1: 1900 -21525.069 0.134 0.053
Chain 1: 2000 -20035.086 0.112 0.053
Chain 1: 2100 -20273.752 0.082 0.035
Chain 1: 2200 -20500.575 0.081 0.035
Chain 1: 2300 -20117.208 0.038 0.019
Chain 1: 2400 -19889.020 0.038 0.019
Chain 1: 2500 -19690.869 0.024 0.015
Chain 1: 2600 -19320.605 0.023 0.015
Chain 1: 2700 -19277.311 0.018 0.012
Chain 1: 2800 -18993.814 0.019 0.015
Chain 1: 2900 -19275.332 0.019 0.015
Chain 1: 3000 -19261.553 0.012 0.012
Chain 1: 3100 -19346.677 0.011 0.011
Chain 1: 3200 -19036.916 0.011 0.015
Chain 1: 3300 -19241.922 0.010 0.011
Chain 1: 3400 -18716.021 0.012 0.015
Chain 1: 3500 -19329.160 0.014 0.015
Chain 1: 3600 -18634.095 0.016 0.015
Chain 1: 3700 -19022.215 0.018 0.016
Chain 1: 3800 -17979.251 0.022 0.020
Chain 1: 3900 -17975.265 0.021 0.020
Chain 1: 4000 -18092.619 0.021 0.020
Chain 1: 4100 -18006.290 0.021 0.020
Chain 1: 4200 -17821.867 0.021 0.020
Chain 1: 4300 -17960.766 0.021 0.020
Chain 1: 4400 -17917.118 0.018 0.010
Chain 1: 4500 -17819.501 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001277 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.77 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48972.305 1.000 1.000
Chain 1: 200 -19154.900 1.278 1.557
Chain 1: 300 -12376.877 1.035 1.000
Chain 1: 400 -25108.683 0.903 1.000
Chain 1: 500 -13701.095 0.889 0.833
Chain 1: 600 -18807.353 0.786 0.833
Chain 1: 700 -13202.416 0.734 0.548
Chain 1: 800 -14255.475 0.652 0.548
Chain 1: 900 -17305.334 0.599 0.507
Chain 1: 1000 -11121.620 0.595 0.548
Chain 1: 1100 -14945.395 0.520 0.507 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -12998.198 0.380 0.425
Chain 1: 1300 -12270.068 0.331 0.272
Chain 1: 1400 -11460.693 0.287 0.256
Chain 1: 1500 -27165.094 0.262 0.256
Chain 1: 1600 -10765.533 0.387 0.256
Chain 1: 1700 -10684.672 0.345 0.176
Chain 1: 1800 -9738.929 0.347 0.176
Chain 1: 1900 -28639.594 0.396 0.256
Chain 1: 2000 -10212.423 0.521 0.256 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2100 -9368.309 0.504 0.150 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2200 -9532.348 0.491 0.097
Chain 1: 2300 -11492.789 0.502 0.171 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2400 -18273.357 0.532 0.371 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2500 -9122.990 0.574 0.371 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2600 -10485.257 0.435 0.171
Chain 1: 2700 -11716.222 0.445 0.171
Chain 1: 2800 -9653.466 0.456 0.214
Chain 1: 2900 -9378.483 0.393 0.171
Chain 1: 3000 -17774.242 0.260 0.171
Chain 1: 3100 -9792.000 0.333 0.214
Chain 1: 3200 -9250.494 0.337 0.214
Chain 1: 3300 -9451.014 0.322 0.214
Chain 1: 3400 -9594.231 0.286 0.130
Chain 1: 3500 -9077.032 0.192 0.105
Chain 1: 3600 -10129.398 0.189 0.104
Chain 1: 3700 -8788.518 0.194 0.104
Chain 1: 3800 -15313.331 0.215 0.104
Chain 1: 3900 -12927.950 0.231 0.153
Chain 1: 4000 -11085.985 0.200 0.153
Chain 1: 4100 -9052.492 0.141 0.153
Chain 1: 4200 -9285.852 0.138 0.153
Chain 1: 4300 -9607.769 0.139 0.153
Chain 1: 4400 -10469.663 0.146 0.153
Chain 1: 4500 -8594.128 0.162 0.166
Chain 1: 4600 -8551.912 0.152 0.166
Chain 1: 4700 -8655.830 0.138 0.166
Chain 1: 4800 -8425.844 0.098 0.082
Chain 1: 4900 -11546.119 0.106 0.082
Chain 1: 5000 -9385.645 0.113 0.082
Chain 1: 5100 -8478.384 0.101 0.082
Chain 1: 5200 -8677.994 0.101 0.082
Chain 1: 5300 -12861.197 0.130 0.107
Chain 1: 5400 -9131.865 0.163 0.218
Chain 1: 5500 -8385.997 0.150 0.107
Chain 1: 5600 -8328.633 0.150 0.107
Chain 1: 5700 -13368.990 0.186 0.230
Chain 1: 5800 -8860.738 0.235 0.270
Chain 1: 5900 -11470.406 0.230 0.230
Chain 1: 6000 -10490.242 0.217 0.228
Chain 1: 6100 -8965.831 0.223 0.228
Chain 1: 6200 -8847.705 0.222 0.228
Chain 1: 6300 -8605.224 0.192 0.170
Chain 1: 6400 -10593.663 0.170 0.170
Chain 1: 6500 -9115.533 0.178 0.170
Chain 1: 6600 -8339.308 0.186 0.170
Chain 1: 6700 -8454.188 0.150 0.162
Chain 1: 6800 -8544.316 0.100 0.093
Chain 1: 6900 -9113.781 0.083 0.093
Chain 1: 7000 -8536.364 0.081 0.068
Chain 1: 7100 -9113.229 0.070 0.063
Chain 1: 7200 -10066.684 0.078 0.068
Chain 1: 7300 -8455.046 0.095 0.093
Chain 1: 7400 -8711.313 0.079 0.068
Chain 1: 7500 -8546.673 0.064 0.063
Chain 1: 7600 -8292.937 0.058 0.062
Chain 1: 7700 -9275.044 0.067 0.063
Chain 1: 7800 -11279.092 0.084 0.068
Chain 1: 7900 -8144.758 0.116 0.095
Chain 1: 8000 -8230.436 0.111 0.095
Chain 1: 8100 -8174.059 0.105 0.095
Chain 1: 8200 -10021.554 0.114 0.106
Chain 1: 8300 -10070.095 0.095 0.031
Chain 1: 8400 -8666.080 0.109 0.106
Chain 1: 8500 -9730.903 0.118 0.109
Chain 1: 8600 -8778.092 0.125 0.109
Chain 1: 8700 -8207.193 0.122 0.109
Chain 1: 8800 -8964.242 0.113 0.109
Chain 1: 8900 -8829.819 0.076 0.084
Chain 1: 9000 -9375.137 0.080 0.084
Chain 1: 9100 -8502.058 0.090 0.103
Chain 1: 9200 -8285.665 0.074 0.084
Chain 1: 9300 -9726.619 0.088 0.103
Chain 1: 9400 -8990.367 0.080 0.084
Chain 1: 9500 -11608.627 0.092 0.084
Chain 1: 9600 -8187.853 0.123 0.084
Chain 1: 9700 -8099.426 0.117 0.084
Chain 1: 9800 -11562.330 0.139 0.103
Chain 1: 9900 -10908.750 0.143 0.103
Chain 1: 10000 -7992.230 0.174 0.148
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001368 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.68 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57261.966 1.000 1.000
Chain 1: 200 -17443.439 1.641 2.283
Chain 1: 300 -8722.588 1.428 1.000
Chain 1: 400 -8148.228 1.088 1.000
Chain 1: 500 -8641.201 0.882 1.000
Chain 1: 600 -9144.072 0.744 1.000
Chain 1: 700 -8434.197 0.650 0.084
Chain 1: 800 -8088.302 0.574 0.084
Chain 1: 900 -7937.352 0.512 0.070
Chain 1: 1000 -7782.265 0.463 0.070
Chain 1: 1100 -7752.351 0.363 0.057
Chain 1: 1200 -7605.257 0.137 0.055
Chain 1: 1300 -7716.395 0.039 0.043
Chain 1: 1400 -7789.306 0.032 0.020
Chain 1: 1500 -7610.978 0.029 0.020
Chain 1: 1600 -7762.021 0.026 0.019
Chain 1: 1700 -7508.555 0.021 0.019
Chain 1: 1800 -7566.617 0.017 0.019
Chain 1: 1900 -7577.702 0.015 0.019
Chain 1: 2000 -7607.860 0.014 0.014
Chain 1: 2100 -7600.250 0.013 0.014
Chain 1: 2200 -7684.957 0.013 0.011
Chain 1: 2300 -7598.335 0.012 0.011
Chain 1: 2400 -7644.068 0.012 0.011
Chain 1: 2500 -7548.466 0.011 0.011
Chain 1: 2600 -7514.774 0.009 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002499 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 24.99 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86044.934 1.000 1.000
Chain 1: 200 -13454.475 3.198 5.395
Chain 1: 300 -9793.650 2.256 1.000
Chain 1: 400 -10920.738 1.718 1.000
Chain 1: 500 -8544.764 1.430 0.374
Chain 1: 600 -8224.385 1.198 0.374
Chain 1: 700 -8277.343 1.028 0.278
Chain 1: 800 -8620.934 0.904 0.278
Chain 1: 900 -8540.819 0.805 0.103
Chain 1: 1000 -8548.848 0.725 0.103
Chain 1: 1100 -8458.699 0.626 0.040 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8230.410 0.089 0.039
Chain 1: 1300 -8485.758 0.055 0.030
Chain 1: 1400 -8489.832 0.044 0.028
Chain 1: 1500 -8336.757 0.018 0.018
Chain 1: 1600 -8450.097 0.016 0.013
Chain 1: 1700 -8528.688 0.016 0.013
Chain 1: 1800 -8104.138 0.017 0.013
Chain 1: 1900 -8205.800 0.018 0.013
Chain 1: 2000 -8180.356 0.018 0.013
Chain 1: 2100 -8306.243 0.018 0.015
Chain 1: 2200 -8108.202 0.018 0.015
Chain 1: 2300 -8200.708 0.016 0.013
Chain 1: 2400 -8269.349 0.017 0.013
Chain 1: 2500 -8215.563 0.016 0.012
Chain 1: 2600 -8217.119 0.014 0.011
Chain 1: 2700 -8133.786 0.014 0.011
Chain 1: 2800 -8093.381 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002527 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.27 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8409851.163 1.000 1.000
Chain 1: 200 -1590251.171 2.644 4.288
Chain 1: 300 -892013.696 2.024 1.000
Chain 1: 400 -457953.723 1.755 1.000
Chain 1: 500 -357967.272 1.460 0.948
Chain 1: 600 -232678.487 1.306 0.948
Chain 1: 700 -118993.090 1.256 0.948
Chain 1: 800 -86238.712 1.146 0.948
Chain 1: 900 -66620.187 1.052 0.783
Chain 1: 1000 -51457.979 0.976 0.783
Chain 1: 1100 -38967.237 0.908 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38148.668 0.481 0.380
Chain 1: 1300 -26136.697 0.449 0.380
Chain 1: 1400 -25858.646 0.355 0.321
Chain 1: 1500 -22454.422 0.343 0.321
Chain 1: 1600 -21673.396 0.292 0.295
Chain 1: 1700 -20551.121 0.202 0.294
Chain 1: 1800 -20496.244 0.165 0.152
Chain 1: 1900 -20822.515 0.137 0.055
Chain 1: 2000 -19335.357 0.115 0.055
Chain 1: 2100 -19573.687 0.084 0.036
Chain 1: 2200 -19799.914 0.083 0.036
Chain 1: 2300 -19417.263 0.039 0.020
Chain 1: 2400 -19189.381 0.039 0.020
Chain 1: 2500 -18991.202 0.025 0.016
Chain 1: 2600 -18621.502 0.024 0.016
Chain 1: 2700 -18578.442 0.018 0.012
Chain 1: 2800 -18295.199 0.020 0.015
Chain 1: 2900 -18576.432 0.020 0.015
Chain 1: 3000 -18562.657 0.012 0.012
Chain 1: 3100 -18647.682 0.011 0.012
Chain 1: 3200 -18338.307 0.012 0.015
Chain 1: 3300 -18543.052 0.011 0.012
Chain 1: 3400 -18017.836 0.013 0.015
Chain 1: 3500 -18629.873 0.015 0.015
Chain 1: 3600 -17936.306 0.017 0.015
Chain 1: 3700 -18323.279 0.019 0.017
Chain 1: 3800 -17282.585 0.023 0.021
Chain 1: 3900 -17278.688 0.022 0.021
Chain 1: 4000 -17396.018 0.022 0.021
Chain 1: 4100 -17309.782 0.022 0.021
Chain 1: 4200 -17125.920 0.022 0.021
Chain 1: 4300 -17264.407 0.021 0.021
Chain 1: 4400 -17221.154 0.019 0.011
Chain 1: 4500 -17123.665 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001287 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.87 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49246.968 1.000 1.000
Chain 1: 200 -20359.426 1.209 1.419
Chain 1: 300 -23479.833 0.851 1.000
Chain 1: 400 -14694.260 0.787 1.000
Chain 1: 500 -18461.401 0.671 0.598
Chain 1: 600 -12270.354 0.643 0.598
Chain 1: 700 -22295.018 0.615 0.505
Chain 1: 800 -16199.250 0.586 0.505
Chain 1: 900 -15574.888 0.525 0.450
Chain 1: 1000 -13178.856 0.491 0.450
Chain 1: 1100 -13575.701 0.394 0.376
Chain 1: 1200 -13330.888 0.253 0.204
Chain 1: 1300 -11550.471 0.256 0.204
Chain 1: 1400 -11456.713 0.197 0.182
Chain 1: 1500 -11906.357 0.180 0.154
Chain 1: 1600 -12750.849 0.136 0.066
Chain 1: 1700 -10096.837 0.117 0.066
Chain 1: 1800 -15850.606 0.116 0.066
Chain 1: 1900 -10940.432 0.157 0.154
Chain 1: 2000 -18476.309 0.180 0.154
Chain 1: 2100 -10090.779 0.260 0.263
Chain 1: 2200 -9839.798 0.261 0.263
Chain 1: 2300 -9667.961 0.247 0.263
Chain 1: 2400 -9447.421 0.248 0.263
Chain 1: 2500 -10412.829 0.254 0.263
Chain 1: 2600 -9489.361 0.257 0.263
Chain 1: 2700 -15206.756 0.268 0.363
Chain 1: 2800 -17034.955 0.243 0.107
Chain 1: 2900 -10012.474 0.268 0.107
Chain 1: 3000 -17196.751 0.269 0.107
Chain 1: 3100 -10422.473 0.251 0.107
Chain 1: 3200 -9091.483 0.263 0.146
Chain 1: 3300 -9411.158 0.265 0.146
Chain 1: 3400 -9717.919 0.265 0.146
Chain 1: 3500 -9994.315 0.259 0.146
Chain 1: 3600 -15590.703 0.285 0.359
Chain 1: 3700 -10722.700 0.293 0.359
Chain 1: 3800 -9099.194 0.300 0.359
Chain 1: 3900 -9099.358 0.230 0.178
Chain 1: 4000 -10688.984 0.203 0.149
Chain 1: 4100 -9361.880 0.152 0.146
Chain 1: 4200 -12200.692 0.161 0.149
Chain 1: 4300 -10217.736 0.177 0.178
Chain 1: 4400 -13618.873 0.199 0.194
Chain 1: 4500 -12420.811 0.205 0.194
Chain 1: 4600 -9194.611 0.205 0.194
Chain 1: 4700 -14519.884 0.196 0.194
Chain 1: 4800 -9110.050 0.237 0.233
Chain 1: 4900 -10161.818 0.248 0.233
Chain 1: 5000 -16191.842 0.270 0.250
Chain 1: 5100 -9326.932 0.330 0.351
Chain 1: 5200 -9439.375 0.308 0.351
Chain 1: 5300 -13713.468 0.319 0.351
Chain 1: 5400 -8710.355 0.352 0.367
Chain 1: 5500 -11241.027 0.365 0.367
Chain 1: 5600 -8678.585 0.359 0.367
Chain 1: 5700 -12996.618 0.356 0.332
Chain 1: 5800 -9199.503 0.338 0.332
Chain 1: 5900 -9386.478 0.329 0.332
Chain 1: 6000 -9250.699 0.293 0.312
Chain 1: 6100 -9110.105 0.221 0.295
Chain 1: 6200 -8823.273 0.223 0.295
Chain 1: 6300 -14834.618 0.233 0.295
Chain 1: 6400 -8608.852 0.248 0.295
Chain 1: 6500 -9273.690 0.232 0.295
Chain 1: 6600 -12374.052 0.228 0.251
Chain 1: 6700 -10857.107 0.209 0.140
Chain 1: 6800 -9370.960 0.183 0.140
Chain 1: 6900 -13039.919 0.209 0.159
Chain 1: 7000 -15420.100 0.223 0.159
Chain 1: 7100 -8358.266 0.306 0.251
Chain 1: 7200 -8677.854 0.307 0.251
Chain 1: 7300 -8340.240 0.270 0.159
Chain 1: 7400 -9717.485 0.212 0.154
Chain 1: 7500 -10972.394 0.216 0.154
Chain 1: 7600 -8981.975 0.213 0.154
Chain 1: 7700 -13885.627 0.235 0.159
Chain 1: 7800 -8956.768 0.274 0.222
Chain 1: 7900 -8394.665 0.252 0.154
Chain 1: 8000 -12540.872 0.270 0.222
Chain 1: 8100 -8768.995 0.229 0.222
Chain 1: 8200 -11664.205 0.250 0.248
Chain 1: 8300 -10997.230 0.252 0.248
Chain 1: 8400 -8353.805 0.269 0.316
Chain 1: 8500 -8414.928 0.259 0.316
Chain 1: 8600 -11542.198 0.263 0.316
Chain 1: 8700 -8392.231 0.266 0.316
Chain 1: 8800 -8466.882 0.212 0.271
Chain 1: 8900 -8960.994 0.210 0.271
Chain 1: 9000 -9303.562 0.181 0.248
Chain 1: 9100 -8346.093 0.149 0.115
Chain 1: 9200 -8744.079 0.129 0.061
Chain 1: 9300 -8604.322 0.125 0.055
Chain 1: 9400 -8367.510 0.096 0.046
Chain 1: 9500 -9325.094 0.105 0.055
Chain 1: 9600 -12711.394 0.105 0.055
Chain 1: 9700 -8339.603 0.120 0.055
Chain 1: 9800 -9069.662 0.127 0.080
Chain 1: 9900 -8630.627 0.127 0.080
Chain 1: 10000 -10206.430 0.138 0.103
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001399 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.99 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56910.736 1.000 1.000
Chain 1: 200 -17772.139 1.601 2.202
Chain 1: 300 -8827.817 1.405 1.013
Chain 1: 400 -8201.953 1.073 1.013
Chain 1: 500 -8550.048 0.866 1.000
Chain 1: 600 -8670.350 0.724 1.000
Chain 1: 700 -7950.496 0.634 0.091
Chain 1: 800 -8347.933 0.561 0.091
Chain 1: 900 -8043.576 0.502 0.076
Chain 1: 1000 -7763.057 0.456 0.076
Chain 1: 1100 -7860.268 0.357 0.048
Chain 1: 1200 -7636.626 0.140 0.041
Chain 1: 1300 -7706.943 0.039 0.038
Chain 1: 1400 -7932.426 0.035 0.036
Chain 1: 1500 -7608.761 0.035 0.036
Chain 1: 1600 -7784.202 0.036 0.036
Chain 1: 1700 -7595.834 0.029 0.029
Chain 1: 1800 -7648.799 0.025 0.028
Chain 1: 1900 -7649.293 0.021 0.025
Chain 1: 2000 -7834.282 0.020 0.024
Chain 1: 2100 -7712.957 0.020 0.024
Chain 1: 2200 -8020.616 0.021 0.024
Chain 1: 2300 -7636.846 0.025 0.025
Chain 1: 2400 -7702.058 0.023 0.024
Chain 1: 2500 -7673.504 0.019 0.023
Chain 1: 2600 -7571.056 0.019 0.016
Chain 1: 2700 -7582.057 0.016 0.014
Chain 1: 2800 -7676.935 0.017 0.014
Chain 1: 2900 -7437.592 0.020 0.016
Chain 1: 3000 -7580.335 0.019 0.016
Chain 1: 3100 -7578.836 0.018 0.014
Chain 1: 3200 -7782.045 0.017 0.014
Chain 1: 3300 -7505.207 0.015 0.014
Chain 1: 3400 -7727.990 0.017 0.019
Chain 1: 3500 -7488.040 0.020 0.026
Chain 1: 3600 -7554.588 0.020 0.026
Chain 1: 3700 -7503.779 0.020 0.026
Chain 1: 3800 -7501.003 0.019 0.026
Chain 1: 3900 -7468.960 0.016 0.019
Chain 1: 4000 -7464.336 0.014 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002549 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.49 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86365.906 1.000 1.000
Chain 1: 200 -13811.826 3.127 5.253
Chain 1: 300 -10115.255 2.206 1.000
Chain 1: 400 -11204.983 1.679 1.000
Chain 1: 500 -8987.397 1.392 0.365
Chain 1: 600 -9477.059 1.169 0.365
Chain 1: 700 -8413.070 1.020 0.247
Chain 1: 800 -9283.499 0.904 0.247
Chain 1: 900 -8969.313 0.808 0.126
Chain 1: 1000 -8996.905 0.727 0.126
Chain 1: 1100 -8734.078 0.630 0.097 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8468.268 0.108 0.094
Chain 1: 1300 -8727.173 0.075 0.052
Chain 1: 1400 -8724.406 0.065 0.035
Chain 1: 1500 -8626.264 0.041 0.031
Chain 1: 1600 -8732.172 0.037 0.030
Chain 1: 1700 -8800.033 0.025 0.030
Chain 1: 1800 -8364.621 0.021 0.030
Chain 1: 1900 -8469.081 0.019 0.012
Chain 1: 2000 -8444.858 0.019 0.012
Chain 1: 2100 -8587.581 0.018 0.012
Chain 1: 2200 -8375.697 0.017 0.012
Chain 1: 2300 -8534.862 0.016 0.012
Chain 1: 2400 -8371.232 0.018 0.017
Chain 1: 2500 -8442.682 0.018 0.017
Chain 1: 2600 -8354.946 0.017 0.017
Chain 1: 2700 -8389.113 0.017 0.017
Chain 1: 2800 -8349.025 0.012 0.012
Chain 1: 2900 -8442.467 0.012 0.011
Chain 1: 3000 -8275.575 0.014 0.017
Chain 1: 3100 -8431.696 0.014 0.019
Chain 1: 3200 -8303.633 0.013 0.015
Chain 1: 3300 -8311.309 0.011 0.011
Chain 1: 3400 -8471.525 0.011 0.011
Chain 1: 3500 -8480.047 0.011 0.011
Chain 1: 3600 -8260.126 0.012 0.015
Chain 1: 3700 -8406.190 0.013 0.017
Chain 1: 3800 -8266.632 0.015 0.017
Chain 1: 3900 -8201.140 0.014 0.017
Chain 1: 4000 -8276.931 0.013 0.017
Chain 1: 4100 -8274.544 0.011 0.015
Chain 1: 4200 -8256.517 0.010 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002555 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.55 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8416793.220 1.000 1.000
Chain 1: 200 -1586824.897 2.652 4.304
Chain 1: 300 -890453.199 2.029 1.000
Chain 1: 400 -457853.948 1.758 1.000
Chain 1: 500 -358050.200 1.462 0.945
Chain 1: 600 -233034.095 1.308 0.945
Chain 1: 700 -119375.401 1.257 0.945
Chain 1: 800 -86645.281 1.147 0.945
Chain 1: 900 -67018.595 1.052 0.782
Chain 1: 1000 -51854.096 0.976 0.782
Chain 1: 1100 -39361.524 0.908 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38542.942 0.480 0.378
Chain 1: 1300 -26523.891 0.447 0.378
Chain 1: 1400 -26246.295 0.353 0.317
Chain 1: 1500 -22840.566 0.340 0.317
Chain 1: 1600 -22059.751 0.290 0.293
Chain 1: 1700 -20935.977 0.200 0.292
Chain 1: 1800 -20880.990 0.163 0.149
Chain 1: 1900 -21207.533 0.135 0.054
Chain 1: 2000 -19719.252 0.113 0.054
Chain 1: 2100 -19957.619 0.083 0.035
Chain 1: 2200 -20184.211 0.082 0.035
Chain 1: 2300 -19801.176 0.038 0.019
Chain 1: 2400 -19573.148 0.039 0.019
Chain 1: 2500 -19375.077 0.025 0.015
Chain 1: 2600 -19004.986 0.023 0.015
Chain 1: 2700 -18961.800 0.018 0.012
Chain 1: 2800 -18678.503 0.019 0.015
Chain 1: 2900 -18959.845 0.019 0.015
Chain 1: 3000 -18945.995 0.012 0.012
Chain 1: 3100 -19031.109 0.011 0.012
Chain 1: 3200 -18721.516 0.011 0.015
Chain 1: 3300 -18926.412 0.011 0.012
Chain 1: 3400 -18400.901 0.012 0.015
Chain 1: 3500 -19013.462 0.015 0.015
Chain 1: 3600 -18319.138 0.016 0.015
Chain 1: 3700 -18706.694 0.018 0.017
Chain 1: 3800 -17664.925 0.023 0.021
Chain 1: 3900 -17660.998 0.021 0.021
Chain 1: 4000 -17778.314 0.022 0.021
Chain 1: 4100 -17692.063 0.022 0.021
Chain 1: 4200 -17507.917 0.021 0.021
Chain 1: 4300 -17646.591 0.021 0.021
Chain 1: 4400 -17603.138 0.018 0.011
Chain 1: 4500 -17505.598 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001854 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 18.54 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -50477.776 1.000 1.000
Chain 1: 200 -21995.264 1.147 1.295
Chain 1: 300 -22231.036 0.769 1.000
Chain 1: 400 -19203.248 0.616 1.000
Chain 1: 500 -15023.197 0.548 0.278
Chain 1: 600 -15713.614 0.464 0.278
Chain 1: 700 -23076.836 0.443 0.278
Chain 1: 800 -15956.037 0.444 0.319
Chain 1: 900 -14542.441 0.405 0.278
Chain 1: 1000 -11635.949 0.390 0.278
Chain 1: 1100 -11448.350 0.291 0.250
Chain 1: 1200 -20883.641 0.207 0.250
Chain 1: 1300 -12796.882 0.269 0.278
Chain 1: 1400 -12565.696 0.255 0.278
Chain 1: 1500 -11057.799 0.241 0.250
Chain 1: 1600 -13173.568 0.253 0.250
Chain 1: 1700 -12654.742 0.225 0.161
Chain 1: 1800 -12686.764 0.181 0.136
Chain 1: 1900 -12441.165 0.173 0.136
Chain 1: 2000 -21804.816 0.191 0.136
Chain 1: 2100 -18413.008 0.208 0.161
Chain 1: 2200 -12695.165 0.207 0.161
Chain 1: 2300 -10477.671 0.165 0.161
Chain 1: 2400 -10799.488 0.167 0.161
Chain 1: 2500 -10987.063 0.155 0.161
Chain 1: 2600 -11861.297 0.146 0.074
Chain 1: 2700 -10268.519 0.157 0.155
Chain 1: 2800 -11479.667 0.168 0.155
Chain 1: 2900 -9875.052 0.182 0.162
Chain 1: 3000 -12089.471 0.157 0.162
Chain 1: 3100 -9508.872 0.166 0.162
Chain 1: 3200 -9864.887 0.125 0.155
Chain 1: 3300 -18381.917 0.150 0.155
Chain 1: 3400 -9930.855 0.232 0.162
Chain 1: 3500 -9970.916 0.231 0.162
Chain 1: 3600 -10047.390 0.224 0.162
Chain 1: 3700 -11545.646 0.221 0.162
Chain 1: 3800 -13009.562 0.222 0.162
Chain 1: 3900 -12535.673 0.210 0.130
Chain 1: 4000 -9595.249 0.222 0.130
Chain 1: 4100 -9788.831 0.197 0.113
Chain 1: 4200 -9752.598 0.194 0.113
Chain 1: 4300 -15486.222 0.184 0.113
Chain 1: 4400 -11812.670 0.130 0.113
Chain 1: 4500 -9735.863 0.151 0.130
Chain 1: 4600 -9122.553 0.157 0.130
Chain 1: 4700 -10318.972 0.156 0.116
Chain 1: 4800 -9676.946 0.151 0.116
Chain 1: 4900 -9584.673 0.148 0.116
Chain 1: 5000 -17826.967 0.164 0.116
Chain 1: 5100 -9440.286 0.251 0.213
Chain 1: 5200 -19294.067 0.302 0.311
Chain 1: 5300 -9786.227 0.362 0.311
Chain 1: 5400 -11922.938 0.348 0.213
Chain 1: 5500 -9850.311 0.348 0.210
Chain 1: 5600 -9921.300 0.342 0.210
Chain 1: 5700 -12999.778 0.354 0.237
Chain 1: 5800 -9493.509 0.385 0.369
Chain 1: 5900 -14057.141 0.416 0.369
Chain 1: 6000 -11887.414 0.388 0.325
Chain 1: 6100 -9001.366 0.331 0.321
Chain 1: 6200 -8838.940 0.282 0.237
Chain 1: 6300 -16187.481 0.230 0.237
Chain 1: 6400 -9097.299 0.290 0.321
Chain 1: 6500 -11900.472 0.293 0.321
Chain 1: 6600 -11116.784 0.299 0.321
Chain 1: 6700 -11767.042 0.281 0.321
Chain 1: 6800 -9464.364 0.268 0.243
Chain 1: 6900 -8990.867 0.241 0.236
Chain 1: 7000 -9040.226 0.224 0.236
Chain 1: 7100 -9495.621 0.196 0.070
Chain 1: 7200 -11152.903 0.209 0.149
Chain 1: 7300 -9026.798 0.187 0.149
Chain 1: 7400 -8882.694 0.111 0.070
Chain 1: 7500 -8801.369 0.088 0.055
Chain 1: 7600 -11271.338 0.103 0.055
Chain 1: 7700 -9617.248 0.115 0.149
Chain 1: 7800 -10929.865 0.103 0.120
Chain 1: 7900 -9661.053 0.111 0.131
Chain 1: 8000 -9128.562 0.116 0.131
Chain 1: 8100 -9664.981 0.117 0.131
Chain 1: 8200 -10499.426 0.110 0.120
Chain 1: 8300 -8811.586 0.105 0.120
Chain 1: 8400 -8973.232 0.105 0.120
Chain 1: 8500 -10490.654 0.119 0.131
Chain 1: 8600 -13072.908 0.117 0.131
Chain 1: 8700 -9322.163 0.140 0.131
Chain 1: 8800 -10504.591 0.139 0.131
Chain 1: 8900 -9360.434 0.138 0.122
Chain 1: 9000 -9104.761 0.135 0.122
Chain 1: 9100 -8739.113 0.134 0.122
Chain 1: 9200 -11453.416 0.150 0.145
Chain 1: 9300 -9742.724 0.148 0.145
Chain 1: 9400 -11066.494 0.158 0.145
Chain 1: 9500 -11910.645 0.151 0.122
Chain 1: 9600 -9864.319 0.152 0.122
Chain 1: 9700 -8740.538 0.124 0.122
Chain 1: 9800 -10113.991 0.127 0.129
Chain 1: 9900 -8921.174 0.128 0.134
Chain 1: 10000 -9065.495 0.127 0.134
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001373 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.73 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61465.979 1.000 1.000
Chain 1: 200 -18851.944 1.630 2.260
Chain 1: 300 -9386.523 1.423 1.008
Chain 1: 400 -8501.289 1.093 1.008
Chain 1: 500 -8055.380 0.886 1.000
Chain 1: 600 -9024.196 0.756 1.000
Chain 1: 700 -7613.602 0.674 0.185
Chain 1: 800 -8588.401 0.604 0.185
Chain 1: 900 -8707.811 0.539 0.114
Chain 1: 1000 -7701.785 0.498 0.131
Chain 1: 1100 -8001.588 0.402 0.114
Chain 1: 1200 -7855.322 0.177 0.107
Chain 1: 1300 -7666.497 0.079 0.104
Chain 1: 1400 -8286.277 0.076 0.075
Chain 1: 1500 -7684.532 0.078 0.078
Chain 1: 1600 -7894.659 0.070 0.075
Chain 1: 1700 -7849.562 0.052 0.037
Chain 1: 1800 -7597.946 0.044 0.033
Chain 1: 1900 -7598.554 0.043 0.033
Chain 1: 2000 -7589.263 0.030 0.027
Chain 1: 2100 -7771.685 0.029 0.025
Chain 1: 2200 -7945.544 0.029 0.025
Chain 1: 2300 -7647.133 0.030 0.027
Chain 1: 2400 -7624.441 0.023 0.023
Chain 1: 2500 -7605.223 0.016 0.022
Chain 1: 2600 -7560.245 0.014 0.006 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003208 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.08 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86511.148 1.000 1.000
Chain 1: 200 -14592.103 2.964 4.929
Chain 1: 300 -10740.147 2.096 1.000
Chain 1: 400 -13034.429 1.616 1.000
Chain 1: 500 -10441.738 1.342 0.359
Chain 1: 600 -9415.608 1.137 0.359
Chain 1: 700 -9045.459 0.980 0.248
Chain 1: 800 -9199.454 0.860 0.248
Chain 1: 900 -9362.480 0.766 0.176
Chain 1: 1000 -8981.631 0.694 0.176
Chain 1: 1100 -9419.419 0.598 0.109 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8809.853 0.113 0.069
Chain 1: 1300 -9349.373 0.082 0.058
Chain 1: 1400 -9107.936 0.067 0.046
Chain 1: 1500 -9121.761 0.043 0.042
Chain 1: 1600 -9175.678 0.032 0.041
Chain 1: 1700 -9243.529 0.029 0.027
Chain 1: 1800 -8768.804 0.033 0.042
Chain 1: 1900 -8873.495 0.032 0.042
Chain 1: 2000 -8885.690 0.028 0.027
Chain 1: 2100 -9022.227 0.025 0.015
Chain 1: 2200 -8745.218 0.021 0.015
Chain 1: 2300 -8834.042 0.017 0.012
Chain 1: 2400 -8926.565 0.015 0.010
Chain 1: 2500 -8831.258 0.016 0.011
Chain 1: 2600 -8883.034 0.016 0.011
Chain 1: 2700 -8785.571 0.016 0.011
Chain 1: 2800 -8740.576 0.011 0.011
Chain 1: 2900 -8842.581 0.011 0.011
Chain 1: 3000 -8763.384 0.012 0.011
Chain 1: 3100 -8721.188 0.011 0.010
Chain 1: 3200 -8682.471 0.008 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002548 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.48 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8403398.178 1.000 1.000
Chain 1: 200 -1581584.283 2.657 4.313
Chain 1: 300 -890490.863 2.030 1.000
Chain 1: 400 -458486.997 1.758 1.000
Chain 1: 500 -359303.037 1.462 0.942
Chain 1: 600 -234376.970 1.307 0.942
Chain 1: 700 -120496.781 1.255 0.942
Chain 1: 800 -87742.953 1.145 0.942
Chain 1: 900 -68056.057 1.050 0.776
Chain 1: 1000 -52850.980 0.974 0.776
Chain 1: 1100 -40313.909 0.905 0.533 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39498.120 0.475 0.373
Chain 1: 1300 -27408.314 0.442 0.373
Chain 1: 1400 -27129.668 0.349 0.311
Chain 1: 1500 -23705.282 0.336 0.311
Chain 1: 1600 -22920.865 0.286 0.289
Chain 1: 1700 -21787.025 0.196 0.288
Chain 1: 1800 -21730.648 0.159 0.144
Chain 1: 1900 -22058.050 0.132 0.052
Chain 1: 2000 -20563.426 0.110 0.052
Chain 1: 2100 -20801.926 0.080 0.034
Chain 1: 2200 -21030.039 0.079 0.034
Chain 1: 2300 -20645.526 0.037 0.019
Chain 1: 2400 -20417.095 0.037 0.019
Chain 1: 2500 -20219.483 0.024 0.015
Chain 1: 2600 -19847.940 0.022 0.015
Chain 1: 2700 -19804.447 0.017 0.011
Chain 1: 2800 -19520.874 0.018 0.015
Chain 1: 2900 -19802.815 0.018 0.014
Chain 1: 3000 -19788.762 0.011 0.011
Chain 1: 3100 -19873.981 0.011 0.011
Chain 1: 3200 -19563.710 0.011 0.014
Chain 1: 3300 -19769.220 0.010 0.011
Chain 1: 3400 -19242.568 0.012 0.014
Chain 1: 3500 -19856.895 0.014 0.015
Chain 1: 3600 -19160.391 0.016 0.015
Chain 1: 3700 -19549.578 0.017 0.016
Chain 1: 3800 -18504.442 0.022 0.020
Chain 1: 3900 -18500.522 0.020 0.020
Chain 1: 4000 -18617.773 0.021 0.020
Chain 1: 4100 -18531.312 0.021 0.020
Chain 1: 4200 -18346.514 0.020 0.020
Chain 1: 4300 -18485.602 0.020 0.020
Chain 1: 4400 -18441.514 0.017 0.010
Chain 1: 4500 -18343.965 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001317 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.17 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49133.342 1.000 1.000
Chain 1: 200 -18024.689 1.363 1.726
Chain 1: 300 -21202.412 0.959 1.000
Chain 1: 400 -21484.073 0.722 1.000
Chain 1: 500 -25443.031 0.609 0.156
Chain 1: 600 -15824.190 0.609 0.608
Chain 1: 700 -11531.228 0.575 0.372
Chain 1: 800 -22528.081 0.564 0.488
Chain 1: 900 -11193.825 0.614 0.488
Chain 1: 1000 -11962.598 0.559 0.488
Chain 1: 1100 -21880.544 0.504 0.453 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -10224.550 0.446 0.453
Chain 1: 1300 -10821.599 0.436 0.453
Chain 1: 1400 -11049.577 0.437 0.453
Chain 1: 1500 -10505.281 0.427 0.453
Chain 1: 1600 -22562.238 0.419 0.453
Chain 1: 1700 -20147.094 0.394 0.453
Chain 1: 1800 -12178.114 0.411 0.453
Chain 1: 1900 -10389.600 0.327 0.172
Chain 1: 2000 -13240.887 0.342 0.215
Chain 1: 2100 -10198.053 0.326 0.215
Chain 1: 2200 -16899.537 0.252 0.215
Chain 1: 2300 -10874.559 0.302 0.298
Chain 1: 2400 -9559.557 0.313 0.298
Chain 1: 2500 -9565.964 0.308 0.298
Chain 1: 2600 -10075.699 0.260 0.215
Chain 1: 2700 -10853.289 0.255 0.215
Chain 1: 2800 -19414.316 0.234 0.215
Chain 1: 2900 -9826.846 0.314 0.298
Chain 1: 3000 -9230.675 0.299 0.298
Chain 1: 3100 -9597.760 0.273 0.138
Chain 1: 3200 -13791.076 0.264 0.138
Chain 1: 3300 -15628.936 0.220 0.118
Chain 1: 3400 -9987.437 0.263 0.118
Chain 1: 3500 -9248.035 0.271 0.118
Chain 1: 3600 -15008.978 0.304 0.304
Chain 1: 3700 -13467.078 0.308 0.304
Chain 1: 3800 -15821.783 0.279 0.149
Chain 1: 3900 -9182.261 0.254 0.149
Chain 1: 4000 -9717.788 0.253 0.149
Chain 1: 4100 -9503.657 0.251 0.149
Chain 1: 4200 -8957.936 0.227 0.118
Chain 1: 4300 -10165.530 0.227 0.119
Chain 1: 4400 -8592.642 0.189 0.119
Chain 1: 4500 -10052.118 0.196 0.145
Chain 1: 4600 -12501.745 0.177 0.145
Chain 1: 4700 -9785.209 0.193 0.149
Chain 1: 4800 -8856.114 0.189 0.145
Chain 1: 4900 -12080.321 0.143 0.145
Chain 1: 5000 -12989.080 0.145 0.145
Chain 1: 5100 -8888.324 0.188 0.183
Chain 1: 5200 -9073.411 0.184 0.183
Chain 1: 5300 -12112.181 0.198 0.196
Chain 1: 5400 -9186.780 0.211 0.251
Chain 1: 5500 -9098.866 0.198 0.251
Chain 1: 5600 -8603.265 0.184 0.251
Chain 1: 5700 -9138.906 0.162 0.105
Chain 1: 5800 -8912.336 0.154 0.070
Chain 1: 5900 -12847.073 0.158 0.070
Chain 1: 6000 -9702.833 0.183 0.251
Chain 1: 6100 -9485.654 0.139 0.059
Chain 1: 6200 -9562.909 0.138 0.059
Chain 1: 6300 -13827.831 0.144 0.059
Chain 1: 6400 -13083.870 0.118 0.058
Chain 1: 6500 -12510.888 0.121 0.058
Chain 1: 6600 -8486.826 0.163 0.059
Chain 1: 6700 -11974.768 0.186 0.291
Chain 1: 6800 -12439.818 0.188 0.291
Chain 1: 6900 -10811.536 0.172 0.151
Chain 1: 7000 -8791.019 0.163 0.151
Chain 1: 7100 -8752.433 0.161 0.151
Chain 1: 7200 -11013.241 0.180 0.205
Chain 1: 7300 -8544.452 0.178 0.205
Chain 1: 7400 -8186.729 0.177 0.205
Chain 1: 7500 -9120.923 0.183 0.205
Chain 1: 7600 -8667.053 0.141 0.151
Chain 1: 7700 -8985.378 0.115 0.102
Chain 1: 7800 -9000.584 0.111 0.102
Chain 1: 7900 -8606.890 0.101 0.052
Chain 1: 8000 -12946.161 0.112 0.052
Chain 1: 8100 -8380.000 0.166 0.102
Chain 1: 8200 -8392.351 0.145 0.052
Chain 1: 8300 -8320.339 0.117 0.046
Chain 1: 8400 -9447.397 0.125 0.052
Chain 1: 8500 -8299.710 0.128 0.052
Chain 1: 8600 -11486.795 0.151 0.119
Chain 1: 8700 -8249.607 0.187 0.138
Chain 1: 8800 -9259.352 0.197 0.138
Chain 1: 8900 -12479.861 0.218 0.258
Chain 1: 9000 -9744.688 0.213 0.258
Chain 1: 9100 -11352.660 0.173 0.142
Chain 1: 9200 -10539.713 0.180 0.142
Chain 1: 9300 -8304.678 0.206 0.258
Chain 1: 9400 -11665.714 0.223 0.269
Chain 1: 9500 -8220.729 0.251 0.277
Chain 1: 9600 -9306.859 0.235 0.269
Chain 1: 9700 -10301.231 0.206 0.258
Chain 1: 9800 -8822.139 0.211 0.258
Chain 1: 9900 -9123.766 0.189 0.168
Chain 1: 10000 -8312.208 0.171 0.142
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001483 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.83 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -62129.910 1.000 1.000
Chain 1: 200 -17877.377 1.738 2.475
Chain 1: 300 -8903.148 1.494 1.008
Chain 1: 400 -9573.817 1.138 1.008
Chain 1: 500 -8601.047 0.933 1.000
Chain 1: 600 -8781.790 0.781 1.000
Chain 1: 700 -8195.851 0.680 0.113
Chain 1: 800 -8253.936 0.596 0.113
Chain 1: 900 -7995.204 0.533 0.071
Chain 1: 1000 -7956.952 0.480 0.071
Chain 1: 1100 -7762.975 0.383 0.070
Chain 1: 1200 -7668.312 0.136 0.032
Chain 1: 1300 -7815.976 0.038 0.025
Chain 1: 1400 -7721.773 0.032 0.021
Chain 1: 1500 -7630.744 0.022 0.019
Chain 1: 1600 -7791.728 0.022 0.019
Chain 1: 1700 -7573.458 0.017 0.019
Chain 1: 1800 -7633.781 0.017 0.019
Chain 1: 1900 -7647.575 0.014 0.012
Chain 1: 2000 -7708.136 0.015 0.012
Chain 1: 2100 -7669.672 0.013 0.012
Chain 1: 2200 -7754.049 0.013 0.012
Chain 1: 2300 -7665.118 0.012 0.012
Chain 1: 2400 -7704.147 0.011 0.011
Chain 1: 2500 -7672.349 0.010 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00291 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.1 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85718.352 1.000 1.000
Chain 1: 200 -13529.297 3.168 5.336
Chain 1: 300 -9949.343 2.232 1.000
Chain 1: 400 -10789.861 1.693 1.000
Chain 1: 500 -8883.618 1.398 0.360
Chain 1: 600 -8457.393 1.173 0.360
Chain 1: 700 -8606.295 1.008 0.215
Chain 1: 800 -9218.798 0.890 0.215
Chain 1: 900 -8674.296 0.798 0.078
Chain 1: 1000 -8565.511 0.720 0.078
Chain 1: 1100 -8686.489 0.621 0.066 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8500.838 0.090 0.063
Chain 1: 1300 -8668.033 0.056 0.050
Chain 1: 1400 -8673.471 0.048 0.022
Chain 1: 1500 -8535.552 0.028 0.019
Chain 1: 1600 -8646.551 0.024 0.017
Chain 1: 1700 -8732.771 0.024 0.016
Chain 1: 1800 -8332.466 0.022 0.016
Chain 1: 1900 -8431.404 0.017 0.014
Chain 1: 2000 -8402.766 0.016 0.014
Chain 1: 2100 -8522.601 0.016 0.014
Chain 1: 2200 -8313.579 0.016 0.014
Chain 1: 2300 -8463.313 0.016 0.014
Chain 1: 2400 -8344.015 0.017 0.014
Chain 1: 2500 -8407.030 0.016 0.014
Chain 1: 2600 -8428.630 0.015 0.014
Chain 1: 2700 -8347.600 0.015 0.014
Chain 1: 2800 -8321.478 0.011 0.012
Chain 1: 2900 -8376.850 0.010 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003082 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.82 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8395993.780 1.000 1.000
Chain 1: 200 -1578715.769 2.659 4.318
Chain 1: 300 -889158.497 2.031 1.000
Chain 1: 400 -457146.228 1.760 1.000
Chain 1: 500 -357879.864 1.463 0.945
Chain 1: 600 -232977.432 1.309 0.945
Chain 1: 700 -119243.764 1.258 0.945
Chain 1: 800 -86492.917 1.148 0.945
Chain 1: 900 -66828.786 1.053 0.776
Chain 1: 1000 -51620.944 0.977 0.776
Chain 1: 1100 -39096.147 0.909 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38269.978 0.480 0.379
Chain 1: 1300 -26220.907 0.448 0.379
Chain 1: 1400 -25938.571 0.355 0.320
Chain 1: 1500 -22525.871 0.342 0.320
Chain 1: 1600 -21742.603 0.292 0.295
Chain 1: 1700 -20615.451 0.202 0.294
Chain 1: 1800 -20559.562 0.165 0.152
Chain 1: 1900 -20885.474 0.137 0.055
Chain 1: 2000 -19397.076 0.115 0.055
Chain 1: 2100 -19635.100 0.084 0.036
Chain 1: 2200 -19861.687 0.083 0.036
Chain 1: 2300 -19478.913 0.039 0.020
Chain 1: 2400 -19251.106 0.039 0.020
Chain 1: 2500 -19053.362 0.025 0.016
Chain 1: 2600 -18683.503 0.023 0.016
Chain 1: 2700 -18640.524 0.018 0.012
Chain 1: 2800 -18357.610 0.020 0.015
Chain 1: 2900 -18638.772 0.019 0.015
Chain 1: 3000 -18624.838 0.012 0.012
Chain 1: 3100 -18709.830 0.011 0.012
Chain 1: 3200 -18400.629 0.012 0.015
Chain 1: 3300 -18605.294 0.011 0.012
Chain 1: 3400 -18080.517 0.013 0.015
Chain 1: 3500 -18691.987 0.015 0.015
Chain 1: 3600 -17999.214 0.017 0.015
Chain 1: 3700 -18385.636 0.018 0.017
Chain 1: 3800 -17346.240 0.023 0.021
Chain 1: 3900 -17342.479 0.021 0.021
Chain 1: 4000 -17459.721 0.022 0.021
Chain 1: 4100 -17373.548 0.022 0.021
Chain 1: 4200 -17190.027 0.021 0.021
Chain 1: 4300 -17328.225 0.021 0.021
Chain 1: 4400 -17285.184 0.019 0.011
Chain 1: 4500 -17187.817 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001245 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.45 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49789.326 1.000 1.000
Chain 1: 200 -24268.072 1.026 1.052
Chain 1: 300 -19457.275 0.766 1.000
Chain 1: 400 -17908.056 0.596 1.000
Chain 1: 500 -12832.202 0.556 0.396
Chain 1: 600 -13267.449 0.469 0.396
Chain 1: 700 -15494.713 0.423 0.247
Chain 1: 800 -31252.876 0.433 0.396
Chain 1: 900 -13831.288 0.525 0.396
Chain 1: 1000 -13538.154 0.474 0.396
Chain 1: 1100 -14245.956 0.379 0.247
Chain 1: 1200 -17195.567 0.291 0.172
Chain 1: 1300 -15937.840 0.274 0.144
Chain 1: 1400 -13696.620 0.282 0.164
Chain 1: 1500 -12587.514 0.251 0.144
Chain 1: 1600 -11908.802 0.254 0.144
Chain 1: 1700 -17387.211 0.271 0.164
Chain 1: 1800 -11439.985 0.273 0.164
Chain 1: 1900 -10362.594 0.157 0.104
Chain 1: 2000 -10632.781 0.157 0.104
Chain 1: 2100 -19729.009 0.198 0.164
Chain 1: 2200 -11464.520 0.253 0.164
Chain 1: 2300 -12749.322 0.256 0.164
Chain 1: 2400 -14334.762 0.250 0.111
Chain 1: 2500 -10338.314 0.280 0.315
Chain 1: 2600 -10528.605 0.276 0.315
Chain 1: 2700 -10617.646 0.246 0.111
Chain 1: 2800 -15120.003 0.223 0.111
Chain 1: 2900 -10220.660 0.261 0.298
Chain 1: 3000 -10956.619 0.265 0.298
Chain 1: 3100 -9598.786 0.233 0.141
Chain 1: 3200 -11232.705 0.176 0.141
Chain 1: 3300 -14492.492 0.188 0.145
Chain 1: 3400 -9713.207 0.226 0.225
Chain 1: 3500 -10307.672 0.193 0.145
Chain 1: 3600 -15954.119 0.227 0.225
Chain 1: 3700 -9650.309 0.291 0.298
Chain 1: 3800 -9323.104 0.265 0.225
Chain 1: 3900 -9635.481 0.220 0.145
Chain 1: 4000 -10918.363 0.225 0.145
Chain 1: 4100 -9469.225 0.227 0.153
Chain 1: 4200 -14157.070 0.245 0.225
Chain 1: 4300 -16389.330 0.236 0.153
Chain 1: 4400 -9748.020 0.255 0.153
Chain 1: 4500 -10517.684 0.257 0.153
Chain 1: 4600 -9686.458 0.230 0.136
Chain 1: 4700 -9371.206 0.168 0.117
Chain 1: 4800 -9254.611 0.166 0.117
Chain 1: 4900 -9521.463 0.165 0.117
Chain 1: 5000 -10278.406 0.161 0.086
Chain 1: 5100 -9226.707 0.157 0.086
Chain 1: 5200 -14255.474 0.159 0.086
Chain 1: 5300 -14344.957 0.146 0.074
Chain 1: 5400 -9150.127 0.135 0.074
Chain 1: 5500 -9663.227 0.133 0.074
Chain 1: 5600 -9222.735 0.129 0.053
Chain 1: 5700 -9314.827 0.127 0.053
Chain 1: 5800 -9861.173 0.131 0.055
Chain 1: 5900 -9183.419 0.135 0.074
Chain 1: 6000 -9533.346 0.132 0.055
Chain 1: 6100 -9188.415 0.124 0.053
Chain 1: 6200 -14068.878 0.124 0.053
Chain 1: 6300 -9505.716 0.171 0.055
Chain 1: 6400 -9918.720 0.118 0.053
Chain 1: 6500 -9245.889 0.120 0.055
Chain 1: 6600 -9227.092 0.116 0.055
Chain 1: 6700 -14439.833 0.151 0.073
Chain 1: 6800 -15753.698 0.154 0.074
Chain 1: 6900 -9218.293 0.217 0.083
Chain 1: 7000 -9386.456 0.215 0.083
Chain 1: 7100 -10979.051 0.226 0.145
Chain 1: 7200 -9221.632 0.210 0.145
Chain 1: 7300 -11482.709 0.182 0.145
Chain 1: 7400 -10297.906 0.189 0.145
Chain 1: 7500 -9548.610 0.190 0.145
Chain 1: 7600 -8969.856 0.196 0.145
Chain 1: 7700 -10272.179 0.173 0.127
Chain 1: 7800 -14570.088 0.194 0.145
Chain 1: 7900 -9196.158 0.181 0.145
Chain 1: 8000 -8875.057 0.183 0.145
Chain 1: 8100 -9735.988 0.178 0.127
Chain 1: 8200 -9234.410 0.164 0.115
Chain 1: 8300 -9100.356 0.146 0.088
Chain 1: 8400 -8846.465 0.137 0.078
Chain 1: 8500 -12115.142 0.156 0.088
Chain 1: 8600 -12925.522 0.156 0.088
Chain 1: 8700 -9178.995 0.184 0.088
Chain 1: 8800 -12192.261 0.179 0.088
Chain 1: 8900 -9128.847 0.155 0.088
Chain 1: 9000 -11427.452 0.171 0.201
Chain 1: 9100 -8694.377 0.194 0.247
Chain 1: 9200 -9703.499 0.199 0.247
Chain 1: 9300 -8656.662 0.209 0.247
Chain 1: 9400 -12370.059 0.236 0.270
Chain 1: 9500 -13222.355 0.216 0.247
Chain 1: 9600 -8850.746 0.259 0.300
Chain 1: 9700 -8713.868 0.220 0.247
Chain 1: 9800 -9325.648 0.202 0.201
Chain 1: 9900 -9247.182 0.169 0.121
Chain 1: 10000 -9543.954 0.152 0.104
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001435 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.35 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57927.970 1.000 1.000
Chain 1: 200 -18348.913 1.579 2.157
Chain 1: 300 -9229.027 1.382 1.000
Chain 1: 400 -8385.941 1.061 1.000
Chain 1: 500 -8535.458 0.853 0.988
Chain 1: 600 -9139.661 0.722 0.988
Chain 1: 700 -8467.394 0.630 0.101
Chain 1: 800 -8588.616 0.553 0.101
Chain 1: 900 -8204.268 0.497 0.079
Chain 1: 1000 -8141.751 0.448 0.079
Chain 1: 1100 -8083.418 0.348 0.066
Chain 1: 1200 -7907.664 0.135 0.047
Chain 1: 1300 -7911.886 0.036 0.022
Chain 1: 1400 -8012.158 0.027 0.018
Chain 1: 1500 -7739.973 0.029 0.022
Chain 1: 1600 -7971.377 0.025 0.022
Chain 1: 1700 -7886.338 0.019 0.014
Chain 1: 1800 -7708.369 0.020 0.022
Chain 1: 1900 -7758.585 0.015 0.013
Chain 1: 2000 -7903.600 0.017 0.018
Chain 1: 2100 -7740.818 0.018 0.021
Chain 1: 2200 -7960.659 0.018 0.021
Chain 1: 2300 -7866.357 0.020 0.021
Chain 1: 2400 -7772.082 0.020 0.021
Chain 1: 2500 -7762.917 0.016 0.018
Chain 1: 2600 -7688.323 0.014 0.012
Chain 1: 2700 -7681.342 0.013 0.012
Chain 1: 2800 -7716.578 0.011 0.012
Chain 1: 2900 -7528.791 0.013 0.012
Chain 1: 3000 -7693.778 0.014 0.012
Chain 1: 3100 -7681.384 0.012 0.012
Chain 1: 3200 -7898.814 0.012 0.012
Chain 1: 3300 -7564.272 0.015 0.012
Chain 1: 3400 -7794.701 0.017 0.021
Chain 1: 3500 -7604.561 0.019 0.025
Chain 1: 3600 -7670.368 0.019 0.025
Chain 1: 3700 -7605.454 0.020 0.025
Chain 1: 3800 -7595.441 0.019 0.025
Chain 1: 3900 -7569.299 0.017 0.021
Chain 1: 4000 -7555.553 0.015 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003085 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.85 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86941.601 1.000 1.000
Chain 1: 200 -14420.066 3.015 5.029
Chain 1: 300 -10629.082 2.129 1.000
Chain 1: 400 -12585.705 1.635 1.000
Chain 1: 500 -9074.108 1.386 0.387
Chain 1: 600 -9074.191 1.155 0.387
Chain 1: 700 -8872.568 0.993 0.357
Chain 1: 800 -9467.919 0.877 0.357
Chain 1: 900 -9359.276 0.781 0.155
Chain 1: 1000 -9006.016 0.706 0.155
Chain 1: 1100 -9432.656 0.611 0.063 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -9034.273 0.112 0.045
Chain 1: 1300 -9248.375 0.079 0.044
Chain 1: 1400 -9168.167 0.064 0.039
Chain 1: 1500 -9082.206 0.027 0.023
Chain 1: 1600 -9169.727 0.028 0.023
Chain 1: 1700 -9234.664 0.026 0.023
Chain 1: 1800 -8795.805 0.025 0.023
Chain 1: 1900 -8893.537 0.025 0.023
Chain 1: 2000 -8909.037 0.021 0.011
Chain 1: 2100 -9001.488 0.017 0.010
Chain 1: 2200 -8784.336 0.016 0.010
Chain 1: 2300 -8972.475 0.015 0.010
Chain 1: 2400 -8792.969 0.017 0.011
Chain 1: 2500 -8866.782 0.016 0.011
Chain 1: 2600 -8777.018 0.016 0.011
Chain 1: 2700 -8809.946 0.016 0.011
Chain 1: 2800 -8760.997 0.012 0.010
Chain 1: 2900 -8875.457 0.012 0.010
Chain 1: 3000 -8792.003 0.013 0.010
Chain 1: 3100 -8753.120 0.012 0.010
Chain 1: 3200 -8725.516 0.010 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002572 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.72 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8372462.726 1.000 1.000
Chain 1: 200 -1579355.583 2.651 4.301
Chain 1: 300 -890933.760 2.025 1.000
Chain 1: 400 -458602.206 1.754 1.000
Chain 1: 500 -359613.446 1.458 0.943
Chain 1: 600 -234615.008 1.304 0.943
Chain 1: 700 -120534.962 1.253 0.943
Chain 1: 800 -87700.572 1.143 0.943
Chain 1: 900 -67976.823 1.048 0.773
Chain 1: 1000 -52725.185 0.972 0.773
Chain 1: 1100 -40148.069 0.904 0.533 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39326.315 0.476 0.374
Chain 1: 1300 -27204.591 0.443 0.374
Chain 1: 1400 -26921.305 0.350 0.313
Chain 1: 1500 -23488.138 0.337 0.313
Chain 1: 1600 -22700.459 0.287 0.290
Chain 1: 1700 -21563.297 0.198 0.289
Chain 1: 1800 -21505.645 0.161 0.146
Chain 1: 1900 -21832.672 0.133 0.053
Chain 1: 2000 -20336.879 0.112 0.053
Chain 1: 2100 -20575.557 0.081 0.035
Chain 1: 2200 -20803.621 0.080 0.035
Chain 1: 2300 -20419.173 0.038 0.019
Chain 1: 2400 -20190.847 0.038 0.019
Chain 1: 2500 -19993.356 0.024 0.015
Chain 1: 2600 -19622.237 0.023 0.015
Chain 1: 2700 -19578.816 0.017 0.012
Chain 1: 2800 -19295.519 0.019 0.015
Chain 1: 2900 -19577.266 0.019 0.014
Chain 1: 3000 -19563.254 0.011 0.012
Chain 1: 3100 -19648.418 0.011 0.011
Chain 1: 3200 -19338.441 0.011 0.014
Chain 1: 3300 -19543.702 0.010 0.011
Chain 1: 3400 -19017.637 0.012 0.014
Chain 1: 3500 -19631.180 0.014 0.015
Chain 1: 3600 -18935.691 0.016 0.015
Chain 1: 3700 -19324.180 0.018 0.016
Chain 1: 3800 -18280.666 0.022 0.020
Chain 1: 3900 -18276.795 0.020 0.020
Chain 1: 4000 -18394.020 0.021 0.020
Chain 1: 4100 -18307.680 0.021 0.020
Chain 1: 4200 -18123.199 0.020 0.020
Chain 1: 4300 -18262.044 0.020 0.020
Chain 1: 4400 -18218.282 0.018 0.010
Chain 1: 4500 -18120.766 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00132 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.2 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48388.535 1.000 1.000
Chain 1: 200 -24099.825 1.004 1.008
Chain 1: 300 -12127.427 0.998 1.000
Chain 1: 400 -14678.256 0.792 1.000
Chain 1: 500 -20117.885 0.688 0.987
Chain 1: 600 -23384.258 0.596 0.987
Chain 1: 700 -13175.825 0.622 0.775
Chain 1: 800 -11285.175 0.565 0.775
Chain 1: 900 -18142.180 0.544 0.378
Chain 1: 1000 -13834.885 0.521 0.378
Chain 1: 1100 -14932.093 0.428 0.311
Chain 1: 1200 -16865.045 0.339 0.270
Chain 1: 1300 -11860.541 0.283 0.270
Chain 1: 1400 -22584.084 0.313 0.311
Chain 1: 1500 -13146.507 0.357 0.378
Chain 1: 1600 -10658.998 0.367 0.378
Chain 1: 1700 -12924.374 0.307 0.311
Chain 1: 1800 -9896.991 0.321 0.311
Chain 1: 1900 -14173.234 0.313 0.306
Chain 1: 2000 -22208.253 0.318 0.306
Chain 1: 2100 -9243.070 0.451 0.362
Chain 1: 2200 -17427.601 0.487 0.422
Chain 1: 2300 -10940.330 0.504 0.470 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2400 -8782.990 0.481 0.362
Chain 1: 2500 -13324.922 0.443 0.341
Chain 1: 2600 -12343.095 0.428 0.341
Chain 1: 2700 -11828.377 0.414 0.341
Chain 1: 2800 -10947.498 0.392 0.341
Chain 1: 2900 -16270.908 0.394 0.341
Chain 1: 3000 -12249.708 0.391 0.328
Chain 1: 3100 -12019.931 0.253 0.327
Chain 1: 3200 -8545.499 0.246 0.327
Chain 1: 3300 -8793.080 0.190 0.246
Chain 1: 3400 -8870.226 0.166 0.080
Chain 1: 3500 -8830.766 0.133 0.080
Chain 1: 3600 -9468.587 0.131 0.067
Chain 1: 3700 -9681.611 0.129 0.067
Chain 1: 3800 -11922.253 0.140 0.067
Chain 1: 3900 -8785.524 0.143 0.067
Chain 1: 4000 -9183.493 0.114 0.043
Chain 1: 4100 -9314.529 0.114 0.043
Chain 1: 4200 -9260.613 0.074 0.028
Chain 1: 4300 -14024.932 0.105 0.043
Chain 1: 4400 -12917.946 0.113 0.067
Chain 1: 4500 -8464.053 0.165 0.086
Chain 1: 4600 -11269.197 0.183 0.188
Chain 1: 4700 -8692.870 0.211 0.249
Chain 1: 4800 -8225.819 0.197 0.249
Chain 1: 4900 -8861.605 0.169 0.086
Chain 1: 5000 -9845.516 0.175 0.100
Chain 1: 5100 -8193.947 0.193 0.202
Chain 1: 5200 -8295.118 0.194 0.202
Chain 1: 5300 -9010.028 0.168 0.100
Chain 1: 5400 -8614.753 0.164 0.100
Chain 1: 5500 -12061.618 0.140 0.100
Chain 1: 5600 -14247.962 0.130 0.100
Chain 1: 5700 -8150.757 0.175 0.100
Chain 1: 5800 -8109.416 0.170 0.100
Chain 1: 5900 -8382.155 0.166 0.100
Chain 1: 6000 -9992.801 0.173 0.153
Chain 1: 6100 -9517.229 0.157 0.079
Chain 1: 6200 -8525.178 0.168 0.116
Chain 1: 6300 -9072.087 0.166 0.116
Chain 1: 6400 -8686.502 0.166 0.116
Chain 1: 6500 -10124.472 0.151 0.116
Chain 1: 6600 -10278.031 0.137 0.060
Chain 1: 6700 -8989.993 0.077 0.060
Chain 1: 6800 -8279.123 0.085 0.086
Chain 1: 6900 -8489.572 0.084 0.086
Chain 1: 7000 -11549.370 0.095 0.086
Chain 1: 7100 -11563.540 0.090 0.086
Chain 1: 7200 -10072.234 0.093 0.086
Chain 1: 7300 -12198.467 0.104 0.142
Chain 1: 7400 -8377.737 0.146 0.143
Chain 1: 7500 -8276.626 0.133 0.143
Chain 1: 7600 -8405.580 0.133 0.143
Chain 1: 7700 -7854.370 0.125 0.086
Chain 1: 7800 -13266.516 0.158 0.148
Chain 1: 7900 -7884.426 0.223 0.174
Chain 1: 8000 -8909.053 0.208 0.148
Chain 1: 8100 -7976.433 0.220 0.148
Chain 1: 8200 -7863.023 0.207 0.117
Chain 1: 8300 -7935.432 0.190 0.115
Chain 1: 8400 -8358.406 0.149 0.070
Chain 1: 8500 -10176.412 0.166 0.115
Chain 1: 8600 -8033.292 0.191 0.117
Chain 1: 8700 -10486.227 0.208 0.179
Chain 1: 8800 -7820.733 0.201 0.179
Chain 1: 8900 -8660.141 0.142 0.117
Chain 1: 9000 -7920.995 0.140 0.117
Chain 1: 9100 -7776.375 0.130 0.097
Chain 1: 9200 -7916.200 0.131 0.097
Chain 1: 9300 -10012.677 0.151 0.179
Chain 1: 9400 -7852.115 0.173 0.209
Chain 1: 9500 -10307.991 0.179 0.234
Chain 1: 9600 -8057.251 0.180 0.234
Chain 1: 9700 -11241.811 0.185 0.238
Chain 1: 9800 -8490.583 0.184 0.238
Chain 1: 9900 -9042.603 0.180 0.238
Chain 1: 10000 -8136.668 0.182 0.238
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001391 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.91 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57350.219 1.000 1.000
Chain 1: 200 -17163.792 1.671 2.341
Chain 1: 300 -8400.670 1.461 1.043
Chain 1: 400 -7844.874 1.114 1.043
Chain 1: 500 -8314.177 0.902 1.000
Chain 1: 600 -7898.059 0.761 1.000
Chain 1: 700 -7733.703 0.655 0.071
Chain 1: 800 -7960.962 0.577 0.071
Chain 1: 900 -7802.643 0.515 0.056
Chain 1: 1000 -7646.920 0.465 0.056
Chain 1: 1100 -7548.913 0.367 0.053
Chain 1: 1200 -7475.195 0.134 0.029
Chain 1: 1300 -7534.230 0.030 0.021
Chain 1: 1400 -7759.977 0.026 0.021
Chain 1: 1500 -7491.067 0.024 0.021
Chain 1: 1600 -7405.723 0.020 0.020
Chain 1: 1700 -7389.558 0.018 0.020
Chain 1: 1800 -7468.430 0.016 0.013
Chain 1: 1900 -7485.339 0.014 0.012
Chain 1: 2000 -7496.848 0.012 0.011
Chain 1: 2100 -7450.771 0.012 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003385 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.85 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85729.967 1.000 1.000
Chain 1: 200 -12967.485 3.306 5.611
Chain 1: 300 -9435.260 2.329 1.000
Chain 1: 400 -10272.914 1.767 1.000
Chain 1: 500 -8326.163 1.460 0.374
Chain 1: 600 -8300.073 1.217 0.374
Chain 1: 700 -8232.763 1.045 0.234
Chain 1: 800 -8316.472 0.915 0.234
Chain 1: 900 -8313.130 0.814 0.082
Chain 1: 1000 -8045.475 0.736 0.082
Chain 1: 1100 -8306.660 0.639 0.033 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -7993.130 0.082 0.033
Chain 1: 1300 -8193.552 0.047 0.031
Chain 1: 1400 -8187.102 0.038 0.024
Chain 1: 1500 -8086.471 0.016 0.012
Chain 1: 1600 -8179.997 0.017 0.012
Chain 1: 1700 -8284.131 0.018 0.013
Chain 1: 1800 -7892.573 0.022 0.024
Chain 1: 1900 -7993.723 0.023 0.024
Chain 1: 2000 -7963.628 0.020 0.013
Chain 1: 2100 -8102.659 0.018 0.013
Chain 1: 2200 -7883.983 0.017 0.013
Chain 1: 2300 -8026.258 0.017 0.013
Chain 1: 2400 -7911.399 0.018 0.015
Chain 1: 2500 -7970.232 0.017 0.015
Chain 1: 2600 -7984.469 0.016 0.015
Chain 1: 2700 -7906.748 0.016 0.015
Chain 1: 2800 -7888.214 0.011 0.013
Chain 1: 2900 -7899.567 0.010 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002585 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.85 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8410081.052 1.000 1.000
Chain 1: 200 -1587200.582 2.649 4.299
Chain 1: 300 -891616.081 2.026 1.000
Chain 1: 400 -457176.252 1.757 1.000
Chain 1: 500 -357276.842 1.462 0.950
Chain 1: 600 -232162.825 1.308 0.950
Chain 1: 700 -118539.644 1.258 0.950
Chain 1: 800 -85759.776 1.149 0.950
Chain 1: 900 -66132.180 1.054 0.780
Chain 1: 1000 -50946.635 0.978 0.780
Chain 1: 1100 -38442.377 0.911 0.539 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37616.277 0.483 0.382
Chain 1: 1300 -25605.850 0.452 0.382
Chain 1: 1400 -25324.867 0.358 0.325
Chain 1: 1500 -21920.588 0.346 0.325
Chain 1: 1600 -21138.438 0.296 0.298
Chain 1: 1700 -20017.107 0.205 0.297
Chain 1: 1800 -19961.950 0.167 0.155
Chain 1: 1900 -20287.427 0.139 0.056
Chain 1: 2000 -18802.105 0.117 0.056
Chain 1: 2100 -19040.373 0.086 0.037
Chain 1: 2200 -19265.927 0.085 0.037
Chain 1: 2300 -18884.094 0.040 0.020
Chain 1: 2400 -18656.456 0.040 0.020
Chain 1: 2500 -18458.259 0.026 0.016
Chain 1: 2600 -18089.412 0.024 0.016
Chain 1: 2700 -18046.631 0.019 0.013
Chain 1: 2800 -17763.692 0.020 0.016
Chain 1: 2900 -18044.583 0.020 0.016
Chain 1: 3000 -18030.908 0.012 0.013
Chain 1: 3100 -18115.773 0.011 0.012
Chain 1: 3200 -17806.968 0.012 0.016
Chain 1: 3300 -18011.268 0.011 0.012
Chain 1: 3400 -17487.019 0.013 0.016
Chain 1: 3500 -18097.609 0.015 0.016
Chain 1: 3600 -17405.980 0.017 0.016
Chain 1: 3700 -17791.496 0.019 0.017
Chain 1: 3800 -16753.785 0.024 0.022
Chain 1: 3900 -16749.958 0.022 0.022
Chain 1: 4000 -16867.299 0.023 0.022
Chain 1: 4100 -16781.171 0.023 0.022
Chain 1: 4200 -16597.976 0.022 0.022
Chain 1: 4300 -16736.005 0.022 0.022
Chain 1: 4400 -16693.302 0.019 0.011
Chain 1: 4500 -16595.886 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001249 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.49 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12703.685 1.000 1.000
Chain 1: 200 -9367.866 0.678 1.000
Chain 1: 300 -8272.700 0.496 0.356
Chain 1: 400 -8235.872 0.373 0.356
Chain 1: 500 -8299.165 0.300 0.132
Chain 1: 600 -8187.219 0.252 0.132
Chain 1: 700 -8066.645 0.218 0.015
Chain 1: 800 -8068.309 0.191 0.015
Chain 1: 900 -7998.124 0.171 0.014
Chain 1: 1000 -8115.580 0.155 0.014
Chain 1: 1100 -8408.811 0.059 0.014
Chain 1: 1200 -8106.334 0.027 0.014
Chain 1: 1300 -8023.778 0.015 0.014
Chain 1: 1400 -8041.552 0.014 0.014
Chain 1: 1500 -8155.027 0.015 0.014
Chain 1: 1600 -8082.315 0.015 0.014
Chain 1: 1700 -8040.847 0.014 0.010
Chain 1: 1800 -8015.730 0.014 0.010
Chain 1: 1900 -8038.004 0.013 0.010
Chain 1: 2000 -7976.246 0.013 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001392 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.92 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -62412.119 1.000 1.000
Chain 1: 200 -18328.807 1.703 2.405
Chain 1: 300 -9059.179 1.476 1.023
Chain 1: 400 -9467.731 1.118 1.023
Chain 1: 500 -8316.150 0.922 1.000
Chain 1: 600 -8497.634 0.772 1.000
Chain 1: 700 -7968.650 0.671 0.138
Chain 1: 800 -8157.967 0.590 0.138
Chain 1: 900 -8131.818 0.525 0.066
Chain 1: 1000 -7864.532 0.476 0.066
Chain 1: 1100 -7852.727 0.376 0.043
Chain 1: 1200 -7637.233 0.138 0.034
Chain 1: 1300 -7862.629 0.039 0.029
Chain 1: 1400 -7603.616 0.038 0.029
Chain 1: 1500 -7536.985 0.025 0.028
Chain 1: 1600 -7710.543 0.025 0.028
Chain 1: 1700 -7732.597 0.019 0.023
Chain 1: 1800 -7666.879 0.017 0.023
Chain 1: 1900 -7573.792 0.018 0.023
Chain 1: 2000 -7676.355 0.016 0.013
Chain 1: 2100 -7577.630 0.017 0.013
Chain 1: 2200 -7820.126 0.018 0.013
Chain 1: 2300 -7525.467 0.019 0.013
Chain 1: 2400 -7525.771 0.015 0.013
Chain 1: 2500 -7543.334 0.015 0.013
Chain 1: 2600 -7515.909 0.013 0.012
Chain 1: 2700 -7503.161 0.013 0.012
Chain 1: 2800 -7507.351 0.012 0.012
Chain 1: 2900 -7370.147 0.012 0.013
Chain 1: 3000 -7510.057 0.013 0.013
Chain 1: 3100 -7514.174 0.012 0.004 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002604 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 26.04 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85535.333 1.000 1.000
Chain 1: 200 -13916.073 3.073 5.147
Chain 1: 300 -10193.039 2.171 1.000
Chain 1: 400 -11559.537 1.657 1.000
Chain 1: 500 -9192.276 1.378 0.365
Chain 1: 600 -9059.399 1.150 0.365
Chain 1: 700 -8523.991 0.995 0.258
Chain 1: 800 -8749.743 0.874 0.258
Chain 1: 900 -8975.135 0.780 0.118
Chain 1: 1000 -8832.248 0.703 0.118
Chain 1: 1100 -8947.786 0.604 0.063 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8519.355 0.095 0.050
Chain 1: 1300 -8827.973 0.062 0.035
Chain 1: 1400 -8803.451 0.050 0.026
Chain 1: 1500 -8685.637 0.026 0.025
Chain 1: 1600 -8792.058 0.026 0.025
Chain 1: 1700 -8850.088 0.020 0.016
Chain 1: 1800 -8407.591 0.023 0.016
Chain 1: 1900 -8514.478 0.021 0.014
Chain 1: 2000 -8500.676 0.020 0.013
Chain 1: 2100 -8618.512 0.020 0.014
Chain 1: 2200 -8412.285 0.017 0.014
Chain 1: 2300 -8507.470 0.015 0.013
Chain 1: 2400 -8574.588 0.016 0.013
Chain 1: 2500 -8522.895 0.015 0.012
Chain 1: 2600 -8536.703 0.014 0.011
Chain 1: 2700 -8444.254 0.014 0.011
Chain 1: 2800 -8391.855 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003579 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.79 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8378292.695 1.000 1.000
Chain 1: 200 -1579654.208 2.652 4.304
Chain 1: 300 -890482.827 2.026 1.000
Chain 1: 400 -458006.024 1.756 1.000
Chain 1: 500 -359004.343 1.460 0.944
Chain 1: 600 -234010.046 1.305 0.944
Chain 1: 700 -119972.825 1.255 0.944
Chain 1: 800 -87154.153 1.145 0.944
Chain 1: 900 -67433.252 1.050 0.774
Chain 1: 1000 -52189.878 0.974 0.774
Chain 1: 1100 -39621.244 0.906 0.534 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38795.697 0.478 0.377
Chain 1: 1300 -26687.525 0.446 0.377
Chain 1: 1400 -26403.887 0.352 0.317
Chain 1: 1500 -22974.659 0.340 0.317
Chain 1: 1600 -22187.469 0.290 0.292
Chain 1: 1700 -21052.479 0.200 0.292
Chain 1: 1800 -20995.194 0.163 0.149
Chain 1: 1900 -21321.876 0.135 0.054
Chain 1: 2000 -19827.694 0.114 0.054
Chain 1: 2100 -20066.224 0.083 0.035
Chain 1: 2200 -20293.934 0.082 0.035
Chain 1: 2300 -19909.925 0.039 0.019
Chain 1: 2400 -19681.728 0.039 0.019
Chain 1: 2500 -19484.189 0.025 0.015
Chain 1: 2600 -19113.424 0.023 0.015
Chain 1: 2700 -19070.112 0.018 0.012
Chain 1: 2800 -18786.916 0.019 0.015
Chain 1: 2900 -19068.501 0.019 0.015
Chain 1: 3000 -19054.530 0.012 0.012
Chain 1: 3100 -19139.644 0.011 0.012
Chain 1: 3200 -18829.887 0.011 0.015
Chain 1: 3300 -19034.956 0.011 0.012
Chain 1: 3400 -18509.243 0.012 0.015
Chain 1: 3500 -19122.248 0.014 0.015
Chain 1: 3600 -18427.459 0.016 0.015
Chain 1: 3700 -18815.405 0.018 0.016
Chain 1: 3800 -17772.988 0.022 0.021
Chain 1: 3900 -17769.139 0.021 0.021
Chain 1: 4000 -17886.366 0.022 0.021
Chain 1: 4100 -17800.083 0.022 0.021
Chain 1: 4200 -17615.845 0.021 0.021
Chain 1: 4300 -17754.523 0.021 0.021
Chain 1: 4400 -17710.942 0.018 0.010
Chain 1: 4500 -17613.475 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001364 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.64 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12606.924 1.000 1.000
Chain 1: 200 -9316.038 0.677 1.000
Chain 1: 300 -8169.342 0.498 0.353
Chain 1: 400 -8307.802 0.378 0.353
Chain 1: 500 -7914.374 0.312 0.140
Chain 1: 600 -8044.822 0.263 0.140
Chain 1: 700 -8040.386 0.225 0.050
Chain 1: 800 -8022.433 0.197 0.050
Chain 1: 900 -8074.845 0.176 0.017
Chain 1: 1000 -8023.890 0.159 0.017
Chain 1: 1100 -8097.425 0.060 0.016
Chain 1: 1200 -7999.138 0.026 0.012
Chain 1: 1300 -7914.780 0.013 0.011
Chain 1: 1400 -7944.171 0.012 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001416 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.16 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58872.561 1.000 1.000
Chain 1: 200 -17970.482 1.638 2.276
Chain 1: 300 -8957.543 1.427 1.006
Chain 1: 400 -9534.884 1.086 1.006
Chain 1: 500 -7852.517 0.911 1.000
Chain 1: 600 -8049.037 0.764 1.000
Chain 1: 700 -7965.609 0.656 0.214
Chain 1: 800 -8354.429 0.580 0.214
Chain 1: 900 -7942.291 0.521 0.061
Chain 1: 1000 -7830.127 0.470 0.061
Chain 1: 1100 -7714.979 0.372 0.052
Chain 1: 1200 -7673.109 0.145 0.047
Chain 1: 1300 -7743.158 0.045 0.024
Chain 1: 1400 -7666.772 0.040 0.015
Chain 1: 1500 -7590.140 0.020 0.014
Chain 1: 1600 -7671.385 0.018 0.011
Chain 1: 1700 -7539.112 0.019 0.014
Chain 1: 1800 -7481.598 0.015 0.011
Chain 1: 1900 -7578.205 0.011 0.011
Chain 1: 2000 -7597.893 0.010 0.010
Chain 1: 2100 -7533.980 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002512 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.12 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86823.160 1.000 1.000
Chain 1: 200 -13759.696 3.155 5.310
Chain 1: 300 -10053.378 2.226 1.000
Chain 1: 400 -11475.286 1.701 1.000
Chain 1: 500 -8943.327 1.417 0.369
Chain 1: 600 -8727.692 1.185 0.369
Chain 1: 700 -8770.947 1.016 0.283
Chain 1: 800 -8368.810 0.895 0.283
Chain 1: 900 -8398.160 0.796 0.124
Chain 1: 1000 -8609.417 0.719 0.124
Chain 1: 1100 -8821.245 0.622 0.048 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8397.451 0.096 0.048
Chain 1: 1300 -8728.398 0.063 0.038
Chain 1: 1400 -8663.804 0.051 0.025
Chain 1: 1500 -8566.150 0.024 0.025
Chain 1: 1600 -8667.183 0.022 0.024
Chain 1: 1700 -8732.167 0.023 0.024
Chain 1: 1800 -8294.972 0.023 0.024
Chain 1: 1900 -8399.630 0.024 0.024
Chain 1: 2000 -8376.841 0.022 0.012
Chain 1: 2100 -8516.047 0.021 0.012
Chain 1: 2200 -8305.937 0.019 0.012
Chain 1: 2300 -8464.553 0.017 0.012
Chain 1: 2400 -8301.403 0.018 0.016
Chain 1: 2500 -8374.039 0.018 0.016
Chain 1: 2600 -8285.368 0.017 0.016
Chain 1: 2700 -8319.327 0.017 0.016
Chain 1: 2800 -8278.782 0.012 0.012
Chain 1: 2900 -8373.153 0.012 0.011
Chain 1: 3000 -8209.542 0.014 0.016
Chain 1: 3100 -8361.810 0.014 0.018
Chain 1: 3200 -8233.138 0.013 0.016
Chain 1: 3300 -8243.976 0.011 0.011
Chain 1: 3400 -8411.264 0.011 0.011
Chain 1: 3500 -8421.054 0.011 0.011
Chain 1: 3600 -8189.452 0.012 0.016
Chain 1: 3700 -8336.665 0.014 0.018
Chain 1: 3800 -8195.431 0.015 0.018
Chain 1: 3900 -8129.485 0.015 0.018
Chain 1: 4000 -8208.762 0.014 0.017
Chain 1: 4100 -8201.017 0.012 0.016
Chain 1: 4200 -8185.938 0.011 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003495 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.95 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8444505.097 1.000 1.000
Chain 1: 200 -1592464.390 2.651 4.303
Chain 1: 300 -891880.538 2.029 1.000
Chain 1: 400 -458143.741 1.759 1.000
Chain 1: 500 -357565.279 1.463 0.947
Chain 1: 600 -232364.029 1.309 0.947
Chain 1: 700 -118981.627 1.258 0.947
Chain 1: 800 -86327.327 1.148 0.947
Chain 1: 900 -66766.819 1.053 0.786
Chain 1: 1000 -51656.718 0.977 0.786
Chain 1: 1100 -39217.911 0.909 0.539 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38408.948 0.481 0.378
Chain 1: 1300 -26439.676 0.447 0.378
Chain 1: 1400 -26168.867 0.354 0.317
Chain 1: 1500 -22775.460 0.341 0.317
Chain 1: 1600 -21998.794 0.290 0.293
Chain 1: 1700 -20880.718 0.200 0.293
Chain 1: 1800 -20827.328 0.163 0.149
Chain 1: 1900 -21153.872 0.135 0.054
Chain 1: 2000 -19668.830 0.113 0.054
Chain 1: 2100 -19906.934 0.083 0.035
Chain 1: 2200 -20132.998 0.082 0.035
Chain 1: 2300 -19750.488 0.038 0.019
Chain 1: 2400 -19522.504 0.039 0.019
Chain 1: 2500 -19324.264 0.025 0.015
Chain 1: 2600 -18954.192 0.023 0.015
Chain 1: 2700 -18911.218 0.018 0.012
Chain 1: 2800 -18627.701 0.019 0.015
Chain 1: 2900 -18909.089 0.019 0.015
Chain 1: 3000 -18895.319 0.012 0.012
Chain 1: 3100 -18980.312 0.011 0.012
Chain 1: 3200 -18670.781 0.011 0.015
Chain 1: 3300 -18875.737 0.011 0.012
Chain 1: 3400 -18350.096 0.012 0.015
Chain 1: 3500 -18962.621 0.015 0.015
Chain 1: 3600 -18268.491 0.016 0.015
Chain 1: 3700 -18655.759 0.018 0.017
Chain 1: 3800 -17614.102 0.023 0.021
Chain 1: 3900 -17610.185 0.021 0.021
Chain 1: 4000 -17727.545 0.022 0.021
Chain 1: 4100 -17641.146 0.022 0.021
Chain 1: 4200 -17457.170 0.021 0.021
Chain 1: 4300 -17595.753 0.021 0.021
Chain 1: 4400 -17552.307 0.018 0.011
Chain 1: 4500 -17454.787 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001387 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.87 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -11988.953 1.000 1.000
Chain 1: 200 -8981.702 0.667 1.000
Chain 1: 300 -7723.104 0.499 0.335
Chain 1: 400 -7902.816 0.380 0.335
Chain 1: 500 -7764.941 0.308 0.163
Chain 1: 600 -7670.639 0.258 0.163
Chain 1: 700 -7593.192 0.223 0.023
Chain 1: 800 -7603.805 0.195 0.023
Chain 1: 900 -7512.375 0.175 0.018
Chain 1: 1000 -7694.263 0.160 0.023
Chain 1: 1100 -7732.631 0.060 0.018
Chain 1: 1200 -7630.935 0.028 0.013
Chain 1: 1300 -7567.830 0.013 0.012
Chain 1: 1400 -7588.921 0.011 0.012
Chain 1: 1500 -7674.653 0.010 0.011
Chain 1: 1600 -7644.725 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00139 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.9 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49055.166 1.000 1.000
Chain 1: 200 -15589.861 1.573 2.147
Chain 1: 300 -8542.229 1.324 1.000
Chain 1: 400 -8072.906 1.007 1.000
Chain 1: 500 -8058.661 0.806 0.825
Chain 1: 600 -8812.393 0.686 0.825
Chain 1: 700 -7723.636 0.608 0.141
Chain 1: 800 -7964.509 0.536 0.141
Chain 1: 900 -7900.091 0.477 0.086
Chain 1: 1000 -7801.057 0.431 0.086
Chain 1: 1100 -7762.182 0.331 0.058
Chain 1: 1200 -7801.330 0.117 0.030
Chain 1: 1300 -7765.760 0.035 0.013
Chain 1: 1400 -7877.217 0.031 0.013
Chain 1: 1500 -7618.466 0.034 0.014
Chain 1: 1600 -7526.721 0.027 0.013
Chain 1: 1700 -7515.739 0.013 0.012
Chain 1: 1800 -7544.650 0.010 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002861 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 28.61 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85627.430 1.000 1.000
Chain 1: 200 -13105.957 3.267 5.533
Chain 1: 300 -9506.599 2.304 1.000
Chain 1: 400 -10168.559 1.744 1.000
Chain 1: 500 -8455.985 1.436 0.379
Chain 1: 600 -8357.532 1.199 0.379
Chain 1: 700 -8374.267 1.028 0.203
Chain 1: 800 -8470.785 0.901 0.203
Chain 1: 900 -8420.199 0.801 0.065
Chain 1: 1000 -8057.630 0.726 0.065
Chain 1: 1100 -8417.744 0.630 0.045 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8031.869 0.081 0.045
Chain 1: 1300 -8375.999 0.048 0.043
Chain 1: 1400 -8233.844 0.043 0.041
Chain 1: 1500 -8106.589 0.024 0.017
Chain 1: 1600 -8217.522 0.024 0.017
Chain 1: 1700 -8306.645 0.025 0.017
Chain 1: 1800 -7908.485 0.029 0.041
Chain 1: 1900 -8008.515 0.030 0.041
Chain 1: 2000 -7979.313 0.026 0.017
Chain 1: 2100 -8099.739 0.023 0.016
Chain 1: 2200 -7876.896 0.021 0.016
Chain 1: 2300 -8037.898 0.019 0.016
Chain 1: 2400 -7919.791 0.018 0.015
Chain 1: 2500 -7983.552 0.018 0.015
Chain 1: 2600 -8004.776 0.017 0.015
Chain 1: 2700 -7924.224 0.017 0.015
Chain 1: 2800 -7898.793 0.012 0.012
Chain 1: 2900 -7953.765 0.011 0.010
Chain 1: 3000 -7838.659 0.012 0.015
Chain 1: 3100 -7975.745 0.013 0.015
Chain 1: 3200 -7856.101 0.011 0.015
Chain 1: 3300 -7877.127 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003367 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.67 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8386422.342 1.000 1.000
Chain 1: 200 -1583795.644 2.648 4.295
Chain 1: 300 -892081.571 2.024 1.000
Chain 1: 400 -457851.325 1.755 1.000
Chain 1: 500 -358272.917 1.459 0.948
Chain 1: 600 -233059.581 1.306 0.948
Chain 1: 700 -119109.418 1.256 0.948
Chain 1: 800 -86214.109 1.147 0.948
Chain 1: 900 -66514.298 1.052 0.775
Chain 1: 1000 -51264.854 0.977 0.775
Chain 1: 1100 -38700.051 0.909 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37870.493 0.482 0.382
Chain 1: 1300 -25798.683 0.451 0.382
Chain 1: 1400 -25512.623 0.357 0.325
Chain 1: 1500 -22092.448 0.345 0.325
Chain 1: 1600 -21305.902 0.295 0.297
Chain 1: 1700 -20177.452 0.205 0.296
Chain 1: 1800 -20120.792 0.167 0.155
Chain 1: 1900 -20446.550 0.139 0.056
Chain 1: 2000 -18956.990 0.117 0.056
Chain 1: 2100 -19195.415 0.086 0.037
Chain 1: 2200 -19421.745 0.085 0.037
Chain 1: 2300 -19039.185 0.040 0.020
Chain 1: 2400 -18811.408 0.040 0.020
Chain 1: 2500 -18613.355 0.026 0.016
Chain 1: 2600 -18243.888 0.024 0.016
Chain 1: 2700 -18200.986 0.019 0.012
Chain 1: 2800 -17917.921 0.020 0.016
Chain 1: 2900 -18199.136 0.020 0.015
Chain 1: 3000 -18185.311 0.012 0.012
Chain 1: 3100 -18270.219 0.011 0.012
Chain 1: 3200 -17961.122 0.012 0.015
Chain 1: 3300 -18165.705 0.011 0.012
Chain 1: 3400 -17640.952 0.013 0.015
Chain 1: 3500 -18252.271 0.015 0.016
Chain 1: 3600 -17559.799 0.017 0.016
Chain 1: 3700 -17945.987 0.019 0.017
Chain 1: 3800 -16906.865 0.023 0.022
Chain 1: 3900 -16903.068 0.022 0.022
Chain 1: 4000 -17020.375 0.023 0.022
Chain 1: 4100 -16934.129 0.023 0.022
Chain 1: 4200 -16750.694 0.022 0.022
Chain 1: 4300 -16888.875 0.022 0.022
Chain 1: 4400 -16845.938 0.019 0.011
Chain 1: 4500 -16748.522 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001359 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.59 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49237.783 1.000 1.000
Chain 1: 200 -19933.283 1.235 1.470
Chain 1: 300 -16792.627 0.886 1.000
Chain 1: 400 -12782.570 0.743 1.000
Chain 1: 500 -13344.882 0.603 0.314
Chain 1: 600 -15654.242 0.527 0.314
Chain 1: 700 -17062.154 0.463 0.187
Chain 1: 800 -13515.136 0.438 0.262
Chain 1: 900 -10908.490 0.416 0.239
Chain 1: 1000 -12079.279 0.384 0.239
Chain 1: 1100 -22136.303 0.330 0.239
Chain 1: 1200 -11505.579 0.275 0.239
Chain 1: 1300 -12170.755 0.262 0.239
Chain 1: 1400 -10334.374 0.248 0.178
Chain 1: 1500 -10284.828 0.244 0.178
Chain 1: 1600 -10167.927 0.231 0.178
Chain 1: 1700 -16326.783 0.260 0.239
Chain 1: 1800 -11246.444 0.279 0.239
Chain 1: 1900 -11773.298 0.260 0.178
Chain 1: 2000 -9856.634 0.270 0.194
Chain 1: 2100 -11988.214 0.242 0.178
Chain 1: 2200 -17923.418 0.183 0.178
Chain 1: 2300 -10357.153 0.250 0.194
Chain 1: 2400 -9249.972 0.244 0.194
Chain 1: 2500 -9746.621 0.249 0.194
Chain 1: 2600 -19624.930 0.298 0.331
Chain 1: 2700 -9703.700 0.363 0.331
Chain 1: 2800 -18967.307 0.366 0.331
Chain 1: 2900 -15728.124 0.382 0.331
Chain 1: 3000 -8838.922 0.441 0.488
Chain 1: 3100 -8894.031 0.424 0.488
Chain 1: 3200 -15369.106 0.433 0.488
Chain 1: 3300 -10881.411 0.401 0.421
Chain 1: 3400 -9199.165 0.407 0.421
Chain 1: 3500 -10621.276 0.416 0.421
Chain 1: 3600 -14159.938 0.390 0.412
Chain 1: 3700 -13573.862 0.292 0.250
Chain 1: 3800 -16105.230 0.259 0.206
Chain 1: 3900 -10083.677 0.298 0.250
Chain 1: 4000 -9527.534 0.226 0.183
Chain 1: 4100 -8990.983 0.232 0.183
Chain 1: 4200 -10646.031 0.205 0.157
Chain 1: 4300 -10042.897 0.170 0.155
Chain 1: 4400 -8930.591 0.164 0.134
Chain 1: 4500 -10323.220 0.164 0.135
Chain 1: 4600 -14709.041 0.169 0.135
Chain 1: 4700 -9705.156 0.216 0.155
Chain 1: 4800 -10247.344 0.206 0.135
Chain 1: 4900 -10646.036 0.150 0.125
Chain 1: 5000 -10293.549 0.147 0.125
Chain 1: 5100 -8605.355 0.161 0.135
Chain 1: 5200 -10776.474 0.166 0.135
Chain 1: 5300 -15619.963 0.191 0.196
Chain 1: 5400 -16368.057 0.183 0.196
Chain 1: 5500 -10287.461 0.228 0.201
Chain 1: 5600 -8664.988 0.217 0.196
Chain 1: 5700 -11947.844 0.193 0.196
Chain 1: 5800 -10855.457 0.198 0.196
Chain 1: 5900 -13461.405 0.213 0.196
Chain 1: 6000 -9092.887 0.258 0.201
Chain 1: 6100 -9897.787 0.247 0.201
Chain 1: 6200 -8941.007 0.237 0.194
Chain 1: 6300 -8771.095 0.208 0.187
Chain 1: 6400 -10076.782 0.217 0.187
Chain 1: 6500 -8977.790 0.170 0.130
Chain 1: 6600 -8528.540 0.156 0.122
Chain 1: 6700 -8412.293 0.130 0.107
Chain 1: 6800 -11961.593 0.150 0.122
Chain 1: 6900 -8776.657 0.167 0.122
Chain 1: 7000 -8618.976 0.120 0.107
Chain 1: 7100 -8260.498 0.117 0.107
Chain 1: 7200 -12245.368 0.138 0.122
Chain 1: 7300 -8217.241 0.186 0.130
Chain 1: 7400 -8810.165 0.179 0.122
Chain 1: 7500 -8827.455 0.167 0.067
Chain 1: 7600 -8394.976 0.167 0.067
Chain 1: 7700 -8288.973 0.167 0.067
Chain 1: 7800 -12983.102 0.174 0.067
Chain 1: 7900 -8429.487 0.191 0.067
Chain 1: 8000 -11316.353 0.215 0.255
Chain 1: 8100 -8780.040 0.239 0.289
Chain 1: 8200 -9627.091 0.216 0.255
Chain 1: 8300 -8400.821 0.181 0.146
Chain 1: 8400 -11708.806 0.203 0.255
Chain 1: 8500 -8446.543 0.241 0.283
Chain 1: 8600 -8475.579 0.236 0.283
Chain 1: 8700 -8229.416 0.238 0.283
Chain 1: 8800 -12674.896 0.237 0.283
Chain 1: 8900 -10524.427 0.204 0.255
Chain 1: 9000 -8774.525 0.198 0.204
Chain 1: 9100 -8527.546 0.172 0.199
Chain 1: 9200 -8436.425 0.164 0.199
Chain 1: 9300 -8601.866 0.152 0.199
Chain 1: 9400 -8758.044 0.125 0.030
Chain 1: 9500 -8419.122 0.090 0.030
Chain 1: 9600 -8457.239 0.091 0.030
Chain 1: 9700 -9655.896 0.100 0.040
Chain 1: 9800 -8957.874 0.073 0.040
Chain 1: 9900 -10846.045 0.070 0.040
Chain 1: 10000 -8480.185 0.078 0.040
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001396 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.96 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61901.088 1.000 1.000
Chain 1: 200 -17815.422 1.737 2.475
Chain 1: 300 -8844.198 1.496 1.014
Chain 1: 400 -8135.973 1.144 1.014
Chain 1: 500 -8697.467 0.928 1.000
Chain 1: 600 -8696.652 0.773 1.000
Chain 1: 700 -7508.062 0.686 0.158
Chain 1: 800 -7651.569 0.602 0.158
Chain 1: 900 -7967.609 0.540 0.087
Chain 1: 1000 -7606.339 0.490 0.087
Chain 1: 1100 -7570.298 0.391 0.065
Chain 1: 1200 -7753.706 0.146 0.047
Chain 1: 1300 -7590.124 0.047 0.040
Chain 1: 1400 -7896.217 0.042 0.039
Chain 1: 1500 -7543.346 0.040 0.039
Chain 1: 1600 -7685.582 0.042 0.039
Chain 1: 1700 -7461.047 0.029 0.030
Chain 1: 1800 -7579.052 0.029 0.030
Chain 1: 1900 -7585.752 0.025 0.024
Chain 1: 2000 -7610.374 0.020 0.022
Chain 1: 2100 -7538.692 0.021 0.022
Chain 1: 2200 -7647.821 0.020 0.019
Chain 1: 2300 -7536.092 0.019 0.016
Chain 1: 2400 -7598.246 0.016 0.015
Chain 1: 2500 -7522.810 0.013 0.014
Chain 1: 2600 -7469.709 0.011 0.010
Chain 1: 2700 -7498.161 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002514 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.14 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86628.635 1.000 1.000
Chain 1: 200 -13545.644 3.198 5.395
Chain 1: 300 -9919.933 2.254 1.000
Chain 1: 400 -10768.494 1.710 1.000
Chain 1: 500 -8893.904 1.410 0.365
Chain 1: 600 -8528.374 1.182 0.365
Chain 1: 700 -8559.509 1.014 0.211
Chain 1: 800 -9202.781 0.896 0.211
Chain 1: 900 -8694.333 0.803 0.079
Chain 1: 1000 -8556.278 0.724 0.079
Chain 1: 1100 -8706.101 0.626 0.070 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8332.884 0.091 0.058
Chain 1: 1300 -8630.325 0.058 0.045
Chain 1: 1400 -8641.375 0.050 0.043
Chain 1: 1500 -8484.635 0.031 0.034
Chain 1: 1600 -8603.637 0.028 0.018
Chain 1: 1700 -8688.028 0.028 0.018
Chain 1: 1800 -8277.139 0.026 0.018
Chain 1: 1900 -8372.918 0.022 0.017
Chain 1: 2000 -8345.834 0.020 0.017
Chain 1: 2100 -8467.742 0.020 0.014
Chain 1: 2200 -8292.492 0.018 0.014
Chain 1: 2300 -8369.466 0.015 0.014
Chain 1: 2400 -8437.680 0.016 0.014
Chain 1: 2500 -8383.284 0.015 0.011
Chain 1: 2600 -8381.887 0.013 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002501 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.01 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8442733.177 1.000 1.000
Chain 1: 200 -1591035.525 2.653 4.306
Chain 1: 300 -891517.461 2.030 1.000
Chain 1: 400 -457647.725 1.760 1.000
Chain 1: 500 -357306.950 1.464 0.948
Chain 1: 600 -232120.748 1.310 0.948
Chain 1: 700 -118755.220 1.259 0.948
Chain 1: 800 -86078.796 1.149 0.948
Chain 1: 900 -66517.458 1.054 0.785
Chain 1: 1000 -51396.900 0.978 0.785
Chain 1: 1100 -38954.412 0.910 0.539 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38139.151 0.482 0.380
Chain 1: 1300 -26180.345 0.449 0.380
Chain 1: 1400 -25906.978 0.355 0.319
Chain 1: 1500 -22516.329 0.342 0.319
Chain 1: 1600 -21739.263 0.292 0.294
Chain 1: 1700 -20623.482 0.202 0.294
Chain 1: 1800 -20570.056 0.164 0.151
Chain 1: 1900 -20896.083 0.136 0.054
Chain 1: 2000 -19412.792 0.114 0.054
Chain 1: 2100 -19650.916 0.084 0.036
Chain 1: 2200 -19876.415 0.083 0.036
Chain 1: 2300 -19494.467 0.039 0.020
Chain 1: 2400 -19266.713 0.039 0.020
Chain 1: 2500 -19068.308 0.025 0.016
Chain 1: 2600 -18699.016 0.023 0.016
Chain 1: 2700 -18656.154 0.018 0.012
Chain 1: 2800 -18372.894 0.019 0.015
Chain 1: 2900 -18653.925 0.019 0.015
Chain 1: 3000 -18640.242 0.012 0.012
Chain 1: 3100 -18725.202 0.011 0.012
Chain 1: 3200 -18416.046 0.012 0.015
Chain 1: 3300 -18620.625 0.011 0.012
Chain 1: 3400 -18095.688 0.012 0.015
Chain 1: 3500 -18707.223 0.015 0.015
Chain 1: 3600 -18014.282 0.017 0.015
Chain 1: 3700 -18400.722 0.018 0.017
Chain 1: 3800 -17360.960 0.023 0.021
Chain 1: 3900 -17357.044 0.021 0.021
Chain 1: 4000 -17474.406 0.022 0.021
Chain 1: 4100 -17388.188 0.022 0.021
Chain 1: 4200 -17204.541 0.021 0.021
Chain 1: 4300 -17342.904 0.021 0.021
Chain 1: 4400 -17299.817 0.019 0.011
Chain 1: 4500 -17202.316 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001287 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.87 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48615.707 1.000 1.000
Chain 1: 200 -17905.568 1.358 1.715
Chain 1: 300 -19836.161 0.937 1.000
Chain 1: 400 -12184.513 0.860 1.000
Chain 1: 500 -18416.795 0.756 0.628
Chain 1: 600 -11388.588 0.733 0.628
Chain 1: 700 -13268.105 0.648 0.617
Chain 1: 800 -12151.163 0.579 0.617
Chain 1: 900 -13516.055 0.526 0.338
Chain 1: 1000 -12519.431 0.481 0.338
Chain 1: 1100 -9882.169 0.408 0.267
Chain 1: 1200 -12562.438 0.258 0.213
Chain 1: 1300 -13344.547 0.254 0.213
Chain 1: 1400 -20454.715 0.226 0.213
Chain 1: 1500 -11258.164 0.273 0.213
Chain 1: 1600 -11517.358 0.214 0.142
Chain 1: 1700 -20920.813 0.245 0.213
Chain 1: 1800 -9390.416 0.358 0.267
Chain 1: 1900 -9466.583 0.349 0.267
Chain 1: 2000 -9669.754 0.343 0.267
Chain 1: 2100 -10304.319 0.323 0.213
Chain 1: 2200 -19034.794 0.347 0.348
Chain 1: 2300 -9066.890 0.451 0.449
Chain 1: 2400 -17605.764 0.465 0.459
Chain 1: 2500 -9858.177 0.462 0.459
Chain 1: 2600 -8872.626 0.471 0.459
Chain 1: 2700 -9093.749 0.428 0.459
Chain 1: 2800 -10463.040 0.319 0.131
Chain 1: 2900 -9332.304 0.330 0.131
Chain 1: 3000 -8592.213 0.336 0.131
Chain 1: 3100 -9792.996 0.343 0.131
Chain 1: 3200 -9484.376 0.300 0.123
Chain 1: 3300 -8968.456 0.196 0.121
Chain 1: 3400 -9912.860 0.157 0.111
Chain 1: 3500 -8770.345 0.091 0.111
Chain 1: 3600 -9979.734 0.092 0.121
Chain 1: 3700 -9314.631 0.097 0.121
Chain 1: 3800 -8760.384 0.090 0.095
Chain 1: 3900 -8693.454 0.079 0.086
Chain 1: 4000 -8527.655 0.072 0.071
Chain 1: 4100 -8629.266 0.061 0.063
Chain 1: 4200 -12625.530 0.089 0.071
Chain 1: 4300 -10023.261 0.110 0.095
Chain 1: 4400 -11504.522 0.113 0.121
Chain 1: 4500 -8601.406 0.134 0.121
Chain 1: 4600 -14863.902 0.164 0.129
Chain 1: 4700 -11437.380 0.187 0.260
Chain 1: 4800 -8389.071 0.217 0.300
Chain 1: 4900 -10943.906 0.239 0.300
Chain 1: 5000 -9794.377 0.249 0.300
Chain 1: 5100 -8588.043 0.262 0.300
Chain 1: 5200 -8924.865 0.234 0.260
Chain 1: 5300 -8922.341 0.208 0.233
Chain 1: 5400 -8844.264 0.196 0.233
Chain 1: 5500 -9359.695 0.168 0.140
Chain 1: 5600 -8881.683 0.131 0.117
Chain 1: 5700 -9051.559 0.103 0.055
Chain 1: 5800 -8696.752 0.071 0.054
Chain 1: 5900 -8356.860 0.051 0.041
Chain 1: 6000 -9152.650 0.048 0.041
Chain 1: 6100 -8371.435 0.044 0.041
Chain 1: 6200 -8136.696 0.043 0.041
Chain 1: 6300 -10641.573 0.066 0.054
Chain 1: 6400 -9380.246 0.079 0.055
Chain 1: 6500 -8481.889 0.084 0.087
Chain 1: 6600 -8433.119 0.079 0.087
Chain 1: 6700 -10795.201 0.099 0.093
Chain 1: 6800 -8171.468 0.127 0.106
Chain 1: 6900 -11455.553 0.152 0.134
Chain 1: 7000 -8085.046 0.185 0.219
Chain 1: 7100 -9290.812 0.188 0.219
Chain 1: 7200 -8127.328 0.200 0.219
Chain 1: 7300 -9642.783 0.192 0.157
Chain 1: 7400 -8217.117 0.196 0.173
Chain 1: 7500 -11643.334 0.215 0.219
Chain 1: 7600 -10696.276 0.223 0.219
Chain 1: 7700 -8143.851 0.232 0.287
Chain 1: 7800 -9026.459 0.210 0.173
Chain 1: 7900 -10084.485 0.192 0.157
Chain 1: 8000 -9275.563 0.159 0.143
Chain 1: 8100 -10043.146 0.154 0.143
Chain 1: 8200 -9773.742 0.142 0.105
Chain 1: 8300 -11109.482 0.138 0.105
Chain 1: 8400 -10651.037 0.125 0.098
Chain 1: 8500 -8126.357 0.127 0.098
Chain 1: 8600 -9959.852 0.137 0.105
Chain 1: 8700 -8187.886 0.127 0.105
Chain 1: 8800 -8101.288 0.118 0.105
Chain 1: 8900 -12075.418 0.141 0.120
Chain 1: 9000 -10303.565 0.149 0.172
Chain 1: 9100 -8192.109 0.167 0.184
Chain 1: 9200 -8089.948 0.166 0.184
Chain 1: 9300 -9587.976 0.169 0.184
Chain 1: 9400 -8326.139 0.180 0.184
Chain 1: 9500 -10077.118 0.166 0.174
Chain 1: 9600 -8747.931 0.163 0.172
Chain 1: 9700 -7863.295 0.153 0.156
Chain 1: 9800 -10279.640 0.175 0.172
Chain 1: 9900 -10133.031 0.144 0.156
Chain 1: 10000 -8025.596 0.153 0.156
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001379 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.79 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56590.117 1.000 1.000
Chain 1: 200 -17053.260 1.659 2.318
Chain 1: 300 -8554.424 1.437 1.000
Chain 1: 400 -8742.651 1.083 1.000
Chain 1: 500 -8488.223 0.873 0.994
Chain 1: 600 -8873.676 0.734 0.994
Chain 1: 700 -7669.514 0.652 0.157
Chain 1: 800 -7976.536 0.575 0.157
Chain 1: 900 -7761.051 0.514 0.043
Chain 1: 1000 -7685.799 0.464 0.043
Chain 1: 1100 -7699.569 0.364 0.038
Chain 1: 1200 -7763.421 0.133 0.030
Chain 1: 1300 -7533.871 0.037 0.030
Chain 1: 1400 -7816.065 0.038 0.030
Chain 1: 1500 -7573.582 0.039 0.032
Chain 1: 1600 -7750.642 0.036 0.030
Chain 1: 1700 -7470.483 0.024 0.030
Chain 1: 1800 -7488.863 0.021 0.028
Chain 1: 1900 -7581.269 0.019 0.023
Chain 1: 2000 -7580.884 0.018 0.023
Chain 1: 2100 -7571.195 0.018 0.023
Chain 1: 2200 -7643.498 0.018 0.023
Chain 1: 2300 -7523.901 0.017 0.016
Chain 1: 2400 -7585.758 0.014 0.012
Chain 1: 2500 -7425.368 0.013 0.012
Chain 1: 2600 -7488.807 0.012 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002887 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 28.87 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86338.386 1.000 1.000
Chain 1: 200 -13149.347 3.283 5.566
Chain 1: 300 -9616.687 2.311 1.000
Chain 1: 400 -10581.362 1.756 1.000
Chain 1: 500 -8508.388 1.454 0.367
Chain 1: 600 -8198.608 1.218 0.367
Chain 1: 700 -8241.215 1.044 0.244
Chain 1: 800 -8683.530 0.920 0.244
Chain 1: 900 -8491.333 0.821 0.091
Chain 1: 1000 -8175.892 0.742 0.091
Chain 1: 1100 -8530.114 0.646 0.051 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8197.819 0.094 0.042
Chain 1: 1300 -8196.080 0.057 0.041
Chain 1: 1400 -8197.934 0.048 0.039
Chain 1: 1500 -8231.863 0.024 0.038
Chain 1: 1600 -8237.419 0.020 0.023
Chain 1: 1700 -8171.573 0.021 0.023
Chain 1: 1800 -8052.807 0.017 0.015
Chain 1: 1900 -8169.223 0.016 0.014
Chain 1: 2000 -8129.339 0.013 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002536 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.36 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8417620.222 1.000 1.000
Chain 1: 200 -1587259.561 2.652 4.303
Chain 1: 300 -890977.077 2.028 1.000
Chain 1: 400 -457449.702 1.758 1.000
Chain 1: 500 -357493.578 1.462 0.948
Chain 1: 600 -232389.495 1.308 0.948
Chain 1: 700 -118676.790 1.258 0.948
Chain 1: 800 -85928.819 1.149 0.948
Chain 1: 900 -66292.044 1.054 0.781
Chain 1: 1000 -51107.388 0.978 0.781
Chain 1: 1100 -38611.453 0.911 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37785.899 0.483 0.381
Chain 1: 1300 -25780.710 0.451 0.381
Chain 1: 1400 -25500.964 0.357 0.324
Chain 1: 1500 -22098.785 0.345 0.324
Chain 1: 1600 -21317.619 0.295 0.297
Chain 1: 1700 -20196.542 0.204 0.296
Chain 1: 1800 -20141.552 0.166 0.154
Chain 1: 1900 -20467.069 0.138 0.056
Chain 1: 2000 -18982.146 0.117 0.056
Chain 1: 2100 -19220.312 0.085 0.037
Chain 1: 2200 -19445.892 0.084 0.037
Chain 1: 2300 -19064.005 0.040 0.020
Chain 1: 2400 -18836.353 0.040 0.020
Chain 1: 2500 -18638.260 0.026 0.016
Chain 1: 2600 -18269.230 0.024 0.016
Chain 1: 2700 -18226.427 0.019 0.012
Chain 1: 2800 -17943.515 0.020 0.016
Chain 1: 2900 -18224.431 0.020 0.015
Chain 1: 3000 -18210.685 0.012 0.012
Chain 1: 3100 -18295.574 0.011 0.012
Chain 1: 3200 -17986.702 0.012 0.015
Chain 1: 3300 -18191.080 0.011 0.012
Chain 1: 3400 -17666.783 0.013 0.015
Chain 1: 3500 -18277.429 0.015 0.016
Chain 1: 3600 -17585.675 0.017 0.016
Chain 1: 3700 -17971.288 0.019 0.017
Chain 1: 3800 -16933.412 0.023 0.021
Chain 1: 3900 -16929.584 0.022 0.021
Chain 1: 4000 -17046.909 0.023 0.021
Chain 1: 4100 -16960.776 0.023 0.021
Chain 1: 4200 -16777.549 0.022 0.021
Chain 1: 4300 -16915.584 0.022 0.021
Chain 1: 4400 -16872.834 0.019 0.011
Chain 1: 4500 -16775.425 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001274 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.74 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48424.899 1.000 1.000
Chain 1: 200 -20620.824 1.174 1.348
Chain 1: 300 -19934.810 0.794 1.000
Chain 1: 400 -14393.688 0.692 1.000
Chain 1: 500 -17422.831 0.588 0.385
Chain 1: 600 -14989.833 0.517 0.385
Chain 1: 700 -14002.808 0.453 0.174
Chain 1: 800 -15028.082 0.405 0.174
Chain 1: 900 -21917.509 0.395 0.174
Chain 1: 1000 -10373.016 0.467 0.314
Chain 1: 1100 -11546.093 0.377 0.174
Chain 1: 1200 -10564.303 0.252 0.162
Chain 1: 1300 -19909.326 0.295 0.174
Chain 1: 1400 -10663.937 0.343 0.174
Chain 1: 1500 -10056.813 0.332 0.162
Chain 1: 1600 -9981.085 0.316 0.102
Chain 1: 1700 -12848.773 0.332 0.223
Chain 1: 1800 -10448.872 0.348 0.230
Chain 1: 1900 -22940.705 0.371 0.230
Chain 1: 2000 -9666.027 0.397 0.230
Chain 1: 2100 -9651.765 0.387 0.230
Chain 1: 2200 -10799.225 0.388 0.230
Chain 1: 2300 -10388.144 0.345 0.223
Chain 1: 2400 -19384.766 0.305 0.223
Chain 1: 2500 -19186.969 0.300 0.223
Chain 1: 2600 -9872.101 0.394 0.230
Chain 1: 2700 -10007.784 0.373 0.230
Chain 1: 2800 -9997.665 0.350 0.106
Chain 1: 2900 -9165.084 0.304 0.091
Chain 1: 3000 -9589.490 0.171 0.044
Chain 1: 3100 -8682.452 0.182 0.091
Chain 1: 3200 -9308.668 0.178 0.067
Chain 1: 3300 -9473.891 0.176 0.067
Chain 1: 3400 -15105.837 0.167 0.067
Chain 1: 3500 -9852.311 0.219 0.091
Chain 1: 3600 -9150.719 0.132 0.077
Chain 1: 3700 -8541.370 0.138 0.077
Chain 1: 3800 -10030.222 0.153 0.091
Chain 1: 3900 -9688.799 0.147 0.077
Chain 1: 4000 -8670.403 0.154 0.104
Chain 1: 4100 -10477.229 0.161 0.117
Chain 1: 4200 -10598.745 0.156 0.117
Chain 1: 4300 -11990.082 0.166 0.117
Chain 1: 4400 -8912.909 0.163 0.117
Chain 1: 4500 -8587.155 0.113 0.116
Chain 1: 4600 -8609.801 0.106 0.116
Chain 1: 4700 -9033.212 0.103 0.116
Chain 1: 4800 -12026.382 0.113 0.116
Chain 1: 4900 -14022.129 0.124 0.117
Chain 1: 5000 -13477.026 0.116 0.116
Chain 1: 5100 -8799.850 0.152 0.116
Chain 1: 5200 -9890.686 0.162 0.116
Chain 1: 5300 -9535.816 0.154 0.110
Chain 1: 5400 -12393.127 0.143 0.110
Chain 1: 5500 -10495.391 0.157 0.142
Chain 1: 5600 -11457.812 0.165 0.142
Chain 1: 5700 -8739.847 0.192 0.181
Chain 1: 5800 -8641.529 0.168 0.142
Chain 1: 5900 -13181.588 0.188 0.181
Chain 1: 6000 -8310.889 0.243 0.231
Chain 1: 6100 -8401.378 0.191 0.181
Chain 1: 6200 -9506.137 0.191 0.181
Chain 1: 6300 -9224.000 0.191 0.181
Chain 1: 6400 -10833.763 0.182 0.149
Chain 1: 6500 -8702.597 0.189 0.149
Chain 1: 6600 -8909.592 0.183 0.149
Chain 1: 6700 -9669.607 0.159 0.116
Chain 1: 6800 -12892.714 0.183 0.149
Chain 1: 6900 -10093.147 0.177 0.149
Chain 1: 7000 -11258.883 0.128 0.116
Chain 1: 7100 -10185.586 0.138 0.116
Chain 1: 7200 -8677.960 0.144 0.149
Chain 1: 7300 -11096.870 0.162 0.174
Chain 1: 7400 -8426.902 0.179 0.218
Chain 1: 7500 -10480.518 0.174 0.196
Chain 1: 7600 -8897.657 0.190 0.196
Chain 1: 7700 -8798.383 0.183 0.196
Chain 1: 7800 -10636.595 0.175 0.178
Chain 1: 7900 -9512.783 0.159 0.174
Chain 1: 8000 -8468.436 0.161 0.174
Chain 1: 8100 -8236.919 0.154 0.174
Chain 1: 8200 -8712.697 0.142 0.173
Chain 1: 8300 -8136.690 0.127 0.123
Chain 1: 8400 -8121.584 0.095 0.118
Chain 1: 8500 -8299.698 0.078 0.071
Chain 1: 8600 -8298.029 0.060 0.055
Chain 1: 8700 -8129.553 0.061 0.055
Chain 1: 8800 -10124.700 0.064 0.055
Chain 1: 8900 -8247.955 0.075 0.055
Chain 1: 9000 -9953.884 0.079 0.055
Chain 1: 9100 -8891.882 0.089 0.071
Chain 1: 9200 -9230.830 0.087 0.071
Chain 1: 9300 -9318.695 0.081 0.037
Chain 1: 9400 -11419.288 0.099 0.119
Chain 1: 9500 -8110.748 0.137 0.171
Chain 1: 9600 -8186.442 0.138 0.171
Chain 1: 9700 -11274.894 0.164 0.184
Chain 1: 9800 -8091.969 0.183 0.184
Chain 1: 9900 -10656.259 0.185 0.184
Chain 1: 10000 -8673.057 0.190 0.229
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001453 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.53 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56657.206 1.000 1.000
Chain 1: 200 -17133.501 1.653 2.307
Chain 1: 300 -8580.527 1.435 1.000
Chain 1: 400 -9037.447 1.089 1.000
Chain 1: 500 -8582.572 0.881 0.997
Chain 1: 600 -8447.853 0.737 0.997
Chain 1: 700 -7730.855 0.645 0.093
Chain 1: 800 -8116.651 0.570 0.093
Chain 1: 900 -7875.451 0.510 0.053
Chain 1: 1000 -7842.930 0.460 0.053
Chain 1: 1100 -7715.800 0.361 0.051
Chain 1: 1200 -7592.869 0.132 0.048
Chain 1: 1300 -7745.434 0.035 0.031
Chain 1: 1400 -7821.207 0.031 0.020
Chain 1: 1500 -7604.993 0.028 0.020
Chain 1: 1600 -7518.460 0.028 0.020
Chain 1: 1700 -7519.046 0.018 0.016
Chain 1: 1800 -7580.014 0.014 0.016
Chain 1: 1900 -7484.346 0.013 0.013
Chain 1: 2000 -7582.657 0.014 0.013
Chain 1: 2100 -7628.159 0.013 0.013
Chain 1: 2200 -7672.887 0.011 0.012
Chain 1: 2300 -7574.051 0.011 0.012
Chain 1: 2400 -7609.531 0.010 0.012
Chain 1: 2500 -7624.516 0.008 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003076 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.76 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86082.919 1.000 1.000
Chain 1: 200 -13181.940 3.265 5.530
Chain 1: 300 -9676.118 2.298 1.000
Chain 1: 400 -10487.958 1.743 1.000
Chain 1: 500 -8552.230 1.439 0.362
Chain 1: 600 -8260.206 1.205 0.362
Chain 1: 700 -8576.944 1.038 0.226
Chain 1: 800 -8782.464 0.912 0.226
Chain 1: 900 -8544.251 0.813 0.077
Chain 1: 1000 -8315.125 0.735 0.077
Chain 1: 1100 -8584.405 0.638 0.037 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8334.466 0.088 0.035
Chain 1: 1300 -8291.843 0.052 0.031
Chain 1: 1400 -8298.357 0.044 0.030
Chain 1: 1500 -8316.589 0.022 0.028
Chain 1: 1600 -8315.336 0.019 0.028
Chain 1: 1700 -8255.630 0.016 0.023
Chain 1: 1800 -8134.121 0.015 0.015
Chain 1: 1900 -8248.158 0.013 0.014
Chain 1: 2000 -8209.377 0.011 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003176 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.76 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8397677.258 1.000 1.000
Chain 1: 200 -1582978.683 2.652 4.305
Chain 1: 300 -889770.233 2.028 1.000
Chain 1: 400 -456551.512 1.758 1.000
Chain 1: 500 -357073.442 1.462 0.949
Chain 1: 600 -232314.333 1.308 0.949
Chain 1: 700 -118734.163 1.258 0.949
Chain 1: 800 -85991.496 1.148 0.949
Chain 1: 900 -66363.489 1.054 0.779
Chain 1: 1000 -51172.546 0.978 0.779
Chain 1: 1100 -38665.020 0.910 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37838.008 0.482 0.381
Chain 1: 1300 -25820.301 0.451 0.381
Chain 1: 1400 -25538.519 0.357 0.323
Chain 1: 1500 -22133.253 0.344 0.323
Chain 1: 1600 -21351.028 0.294 0.297
Chain 1: 1700 -20228.435 0.204 0.296
Chain 1: 1800 -20173.051 0.166 0.154
Chain 1: 1900 -20498.477 0.138 0.055
Chain 1: 2000 -19013.005 0.116 0.055
Chain 1: 2100 -19251.143 0.085 0.037
Chain 1: 2200 -19476.816 0.084 0.037
Chain 1: 2300 -19094.911 0.040 0.020
Chain 1: 2400 -18867.312 0.040 0.020
Chain 1: 2500 -18669.286 0.026 0.016
Chain 1: 2600 -18300.337 0.024 0.016
Chain 1: 2700 -18257.557 0.019 0.012
Chain 1: 2800 -17974.755 0.020 0.016
Chain 1: 2900 -18255.613 0.020 0.015
Chain 1: 3000 -18241.857 0.012 0.012
Chain 1: 3100 -18326.725 0.011 0.012
Chain 1: 3200 -18017.938 0.012 0.015
Chain 1: 3300 -18222.259 0.011 0.012
Chain 1: 3400 -17698.129 0.013 0.015
Chain 1: 3500 -18308.558 0.015 0.016
Chain 1: 3600 -17617.126 0.017 0.016
Chain 1: 3700 -18002.526 0.019 0.017
Chain 1: 3800 -16965.152 0.023 0.021
Chain 1: 3900 -16961.369 0.022 0.021
Chain 1: 4000 -17078.666 0.022 0.021
Chain 1: 4100 -16992.568 0.023 0.021
Chain 1: 4200 -16809.464 0.022 0.021
Chain 1: 4300 -16947.409 0.022 0.021
Chain 1: 4400 -16904.754 0.019 0.011
Chain 1: 4500 -16807.378 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001306 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.06 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49395.915 1.000 1.000
Chain 1: 200 -20274.024 1.218 1.436
Chain 1: 300 -13287.166 0.987 1.000
Chain 1: 400 -12499.488 0.756 1.000
Chain 1: 500 -15247.863 0.641 0.526
Chain 1: 600 -17599.658 0.557 0.526
Chain 1: 700 -16234.658 0.489 0.180
Chain 1: 800 -12769.849 0.462 0.271
Chain 1: 900 -20152.297 0.451 0.271
Chain 1: 1000 -13767.997 0.452 0.366
Chain 1: 1100 -12864.279 0.359 0.271
Chain 1: 1200 -16863.439 0.240 0.237
Chain 1: 1300 -12070.604 0.227 0.237
Chain 1: 1400 -15333.788 0.242 0.237
Chain 1: 1500 -12977.962 0.242 0.237
Chain 1: 1600 -26628.975 0.280 0.271
Chain 1: 1700 -10241.808 0.431 0.366
Chain 1: 1800 -10883.936 0.410 0.366
Chain 1: 1900 -11579.589 0.379 0.237
Chain 1: 2000 -10495.026 0.343 0.213
Chain 1: 2100 -9418.918 0.348 0.213
Chain 1: 2200 -10274.790 0.332 0.182
Chain 1: 2300 -16462.281 0.330 0.182
Chain 1: 2400 -9626.803 0.380 0.182
Chain 1: 2500 -9872.893 0.364 0.114
Chain 1: 2600 -10437.149 0.318 0.103
Chain 1: 2700 -11765.220 0.170 0.103
Chain 1: 2800 -9573.070 0.187 0.113
Chain 1: 2900 -10470.395 0.189 0.113
Chain 1: 3000 -11135.867 0.185 0.113
Chain 1: 3100 -9591.162 0.190 0.113
Chain 1: 3200 -9931.057 0.185 0.113
Chain 1: 3300 -10847.847 0.156 0.086
Chain 1: 3400 -15530.641 0.115 0.086
Chain 1: 3500 -11343.406 0.149 0.113
Chain 1: 3600 -16101.379 0.173 0.161
Chain 1: 3700 -9416.021 0.233 0.229
Chain 1: 3800 -11537.729 0.229 0.184
Chain 1: 3900 -9872.269 0.237 0.184
Chain 1: 4000 -9242.855 0.238 0.184
Chain 1: 4100 -9443.162 0.224 0.184
Chain 1: 4200 -14863.882 0.257 0.296
Chain 1: 4300 -10136.470 0.295 0.302
Chain 1: 4400 -10022.625 0.266 0.296
Chain 1: 4500 -10120.551 0.230 0.184
Chain 1: 4600 -14503.113 0.231 0.184
Chain 1: 4700 -11700.619 0.184 0.184
Chain 1: 4800 -9186.307 0.193 0.240
Chain 1: 4900 -9072.819 0.177 0.240
Chain 1: 5000 -13673.022 0.204 0.274
Chain 1: 5100 -9179.103 0.251 0.302
Chain 1: 5200 -9330.902 0.216 0.274
Chain 1: 5300 -9703.441 0.173 0.240
Chain 1: 5400 -12426.581 0.194 0.240
Chain 1: 5500 -12768.047 0.195 0.240
Chain 1: 5600 -8986.806 0.207 0.240
Chain 1: 5700 -9534.792 0.189 0.219
Chain 1: 5800 -9557.798 0.162 0.057
Chain 1: 5900 -14293.857 0.194 0.219
Chain 1: 6000 -9912.904 0.204 0.219
Chain 1: 6100 -9238.491 0.163 0.073
Chain 1: 6200 -8963.321 0.164 0.073
Chain 1: 6300 -14469.661 0.198 0.219
Chain 1: 6400 -8646.491 0.244 0.331
Chain 1: 6500 -9187.206 0.247 0.331
Chain 1: 6600 -8666.380 0.211 0.073
Chain 1: 6700 -9833.279 0.217 0.119
Chain 1: 6800 -8929.124 0.227 0.119
Chain 1: 6900 -11408.767 0.216 0.119
Chain 1: 7000 -8644.195 0.203 0.119
Chain 1: 7100 -8868.040 0.199 0.119
Chain 1: 7200 -8824.608 0.196 0.119
Chain 1: 7300 -8754.037 0.159 0.101
Chain 1: 7400 -12940.700 0.124 0.101
Chain 1: 7500 -10473.648 0.141 0.119
Chain 1: 7600 -8787.640 0.155 0.192
Chain 1: 7700 -9070.444 0.146 0.192
Chain 1: 7800 -9050.178 0.136 0.192
Chain 1: 7900 -9067.893 0.114 0.031
Chain 1: 8000 -12503.827 0.110 0.031
Chain 1: 8100 -9395.890 0.140 0.192
Chain 1: 8200 -8621.378 0.149 0.192
Chain 1: 8300 -9244.725 0.155 0.192
Chain 1: 8400 -8740.157 0.128 0.090
Chain 1: 8500 -15206.912 0.147 0.090
Chain 1: 8600 -8427.482 0.209 0.090
Chain 1: 8700 -9037.301 0.212 0.090
Chain 1: 8800 -8619.722 0.217 0.090
Chain 1: 8900 -8677.731 0.217 0.090
Chain 1: 9000 -9800.846 0.201 0.090
Chain 1: 9100 -12304.721 0.189 0.090
Chain 1: 9200 -8611.506 0.222 0.115
Chain 1: 9300 -8568.222 0.216 0.115
Chain 1: 9400 -8641.745 0.211 0.115
Chain 1: 9500 -9600.202 0.179 0.100
Chain 1: 9600 -10258.747 0.105 0.067
Chain 1: 9700 -9412.182 0.107 0.090
Chain 1: 9800 -8783.695 0.109 0.090
Chain 1: 9900 -8671.524 0.110 0.090
Chain 1: 10000 -8494.029 0.101 0.072
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001365 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.65 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58822.318 1.000 1.000
Chain 1: 200 -18259.306 1.611 2.221
Chain 1: 300 -8923.501 1.423 1.046
Chain 1: 400 -8196.924 1.089 1.046
Chain 1: 500 -8687.293 0.883 1.000
Chain 1: 600 -9041.486 0.742 1.000
Chain 1: 700 -7950.123 0.656 0.137
Chain 1: 800 -8241.835 0.578 0.137
Chain 1: 900 -8105.425 0.516 0.089
Chain 1: 1000 -7930.683 0.466 0.089
Chain 1: 1100 -7676.078 0.370 0.056
Chain 1: 1200 -7740.766 0.148 0.039
Chain 1: 1300 -7767.461 0.044 0.035
Chain 1: 1400 -8086.587 0.039 0.035
Chain 1: 1500 -7645.819 0.039 0.035
Chain 1: 1600 -7804.397 0.037 0.033
Chain 1: 1700 -7760.442 0.024 0.022
Chain 1: 1800 -7772.621 0.021 0.020
Chain 1: 1900 -7649.394 0.021 0.020
Chain 1: 2000 -7736.777 0.020 0.016
Chain 1: 2100 -7637.430 0.018 0.013
Chain 1: 2200 -7899.031 0.020 0.016
Chain 1: 2300 -7600.912 0.024 0.020
Chain 1: 2400 -7589.248 0.020 0.016
Chain 1: 2500 -7599.502 0.014 0.013
Chain 1: 2600 -7581.942 0.013 0.011
Chain 1: 2700 -7501.483 0.013 0.011
Chain 1: 2800 -7567.324 0.014 0.011
Chain 1: 2900 -7429.297 0.014 0.011
Chain 1: 3000 -7588.295 0.015 0.013
Chain 1: 3100 -7576.842 0.014 0.011
Chain 1: 3200 -7789.820 0.013 0.011
Chain 1: 3300 -7499.758 0.013 0.011
Chain 1: 3400 -7744.564 0.016 0.019
Chain 1: 3500 -7488.060 0.019 0.021
Chain 1: 3600 -7555.076 0.020 0.021
Chain 1: 3700 -7505.412 0.020 0.021
Chain 1: 3800 -7506.544 0.019 0.021
Chain 1: 3900 -7464.914 0.018 0.021
Chain 1: 4000 -7457.041 0.016 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002872 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 28.72 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86267.629 1.000 1.000
Chain 1: 200 -13959.818 3.090 5.180
Chain 1: 300 -10281.330 2.179 1.000
Chain 1: 400 -11258.010 1.656 1.000
Chain 1: 500 -9273.973 1.368 0.358
Chain 1: 600 -8740.045 1.150 0.358
Chain 1: 700 -9169.233 0.992 0.214
Chain 1: 800 -9581.517 0.874 0.214
Chain 1: 900 -8975.985 0.784 0.087
Chain 1: 1000 -9078.981 0.707 0.087
Chain 1: 1100 -8913.902 0.609 0.067 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8678.706 0.093 0.061
Chain 1: 1300 -8953.360 0.061 0.047
Chain 1: 1400 -8919.922 0.052 0.043
Chain 1: 1500 -8812.534 0.032 0.031
Chain 1: 1600 -8920.749 0.027 0.027
Chain 1: 1700 -8994.241 0.023 0.019
Chain 1: 1800 -8563.306 0.024 0.019
Chain 1: 1900 -8667.280 0.019 0.012
Chain 1: 2000 -8642.579 0.018 0.012
Chain 1: 2100 -8779.435 0.017 0.012
Chain 1: 2200 -8572.455 0.017 0.012
Chain 1: 2300 -8671.184 0.015 0.012
Chain 1: 2400 -8733.942 0.016 0.012
Chain 1: 2500 -8674.998 0.015 0.012
Chain 1: 2600 -8680.521 0.014 0.011
Chain 1: 2700 -8595.336 0.014 0.011
Chain 1: 2800 -8551.782 0.010 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002758 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 27.58 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8380449.465 1.000 1.000
Chain 1: 200 -1583444.166 2.646 4.293
Chain 1: 300 -892251.552 2.022 1.000
Chain 1: 400 -458906.059 1.753 1.000
Chain 1: 500 -359420.769 1.458 0.944
Chain 1: 600 -234208.839 1.304 0.944
Chain 1: 700 -120091.207 1.253 0.944
Chain 1: 800 -87201.769 1.144 0.944
Chain 1: 900 -67477.758 1.049 0.775
Chain 1: 1000 -52218.471 0.973 0.775
Chain 1: 1100 -39641.968 0.905 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38814.086 0.478 0.377
Chain 1: 1300 -26714.173 0.446 0.377
Chain 1: 1400 -26428.654 0.353 0.317
Chain 1: 1500 -23001.247 0.340 0.317
Chain 1: 1600 -22213.693 0.290 0.292
Chain 1: 1700 -21080.609 0.200 0.292
Chain 1: 1800 -21023.364 0.163 0.149
Chain 1: 1900 -21349.748 0.135 0.054
Chain 1: 2000 -19856.829 0.113 0.054
Chain 1: 2100 -20095.402 0.083 0.035
Chain 1: 2200 -20322.650 0.082 0.035
Chain 1: 2300 -19939.114 0.038 0.019
Chain 1: 2400 -19711.043 0.039 0.019
Chain 1: 2500 -19513.292 0.025 0.015
Chain 1: 2600 -19142.997 0.023 0.015
Chain 1: 2700 -19099.795 0.018 0.012
Chain 1: 2800 -18816.618 0.019 0.015
Chain 1: 2900 -19098.119 0.019 0.015
Chain 1: 3000 -19084.192 0.012 0.012
Chain 1: 3100 -19169.234 0.011 0.012
Chain 1: 3200 -18859.697 0.011 0.015
Chain 1: 3300 -19064.606 0.011 0.012
Chain 1: 3400 -18539.179 0.012 0.015
Chain 1: 3500 -19151.650 0.014 0.015
Chain 1: 3600 -18457.628 0.016 0.015
Chain 1: 3700 -18844.991 0.018 0.016
Chain 1: 3800 -17803.627 0.022 0.021
Chain 1: 3900 -17799.794 0.021 0.021
Chain 1: 4000 -17917.066 0.022 0.021
Chain 1: 4100 -17830.772 0.022 0.021
Chain 1: 4200 -17646.787 0.021 0.021
Chain 1: 4300 -17785.319 0.021 0.021
Chain 1: 4400 -17741.958 0.018 0.010
Chain 1: 4500 -17644.496 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001291 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.91 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12202.380 1.000 1.000
Chain 1: 200 -9044.089 0.675 1.000
Chain 1: 300 -8132.625 0.487 0.349
Chain 1: 400 -8118.005 0.366 0.349
Chain 1: 500 -8020.344 0.295 0.112
Chain 1: 600 -7939.038 0.248 0.112
Chain 1: 700 -7868.338 0.213 0.012
Chain 1: 800 -7874.767 0.187 0.012
Chain 1: 900 -7939.915 0.167 0.010
Chain 1: 1000 -7936.379 0.150 0.010
Chain 1: 1100 -7983.159 0.051 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001434 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.34 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61484.134 1.000 1.000
Chain 1: 200 -17469.741 1.760 2.519
Chain 1: 300 -8704.668 1.509 1.007
Chain 1: 400 -8189.648 1.147 1.007
Chain 1: 500 -8278.152 0.920 1.000
Chain 1: 600 -8006.821 0.772 1.000
Chain 1: 700 -7625.059 0.669 0.063
Chain 1: 800 -8002.884 0.591 0.063
Chain 1: 900 -7834.536 0.528 0.050
Chain 1: 1000 -7586.400 0.479 0.050
Chain 1: 1100 -7566.648 0.379 0.047
Chain 1: 1200 -7514.778 0.128 0.034
Chain 1: 1300 -7730.990 0.030 0.033
Chain 1: 1400 -7615.067 0.025 0.028
Chain 1: 1500 -7546.142 0.025 0.028
Chain 1: 1600 -7518.690 0.022 0.021
Chain 1: 1700 -7458.245 0.017 0.015
Chain 1: 1800 -7523.200 0.014 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003158 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.58 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85952.864 1.000 1.000
Chain 1: 200 -13258.607 3.241 5.483
Chain 1: 300 -9726.019 2.282 1.000
Chain 1: 400 -10572.827 1.732 1.000
Chain 1: 500 -8605.710 1.431 0.363
Chain 1: 600 -8533.487 1.194 0.363
Chain 1: 700 -8519.286 1.024 0.229
Chain 1: 800 -8581.558 0.897 0.229
Chain 1: 900 -8611.382 0.797 0.080
Chain 1: 1000 -8331.943 0.721 0.080
Chain 1: 1100 -8608.464 0.624 0.034 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8257.825 0.080 0.034
Chain 1: 1300 -8315.442 0.044 0.032
Chain 1: 1400 -8313.449 0.036 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00317 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.7 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8398404.782 1.000 1.000
Chain 1: 200 -1581253.847 2.656 4.311
Chain 1: 300 -889627.975 2.030 1.000
Chain 1: 400 -457098.063 1.759 1.000
Chain 1: 500 -357549.695 1.463 0.946
Chain 1: 600 -232619.432 1.308 0.946
Chain 1: 700 -118896.071 1.258 0.946
Chain 1: 800 -86135.874 1.148 0.946
Chain 1: 900 -66485.434 1.054 0.777
Chain 1: 1000 -51281.329 0.978 0.777
Chain 1: 1100 -38768.147 0.910 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37939.450 0.481 0.380
Chain 1: 1300 -25910.251 0.450 0.380
Chain 1: 1400 -25628.313 0.356 0.323
Chain 1: 1500 -22219.963 0.344 0.323
Chain 1: 1600 -21437.296 0.294 0.296
Chain 1: 1700 -20313.101 0.204 0.296
Chain 1: 1800 -20257.473 0.166 0.153
Chain 1: 1900 -20583.010 0.138 0.055
Chain 1: 2000 -19096.600 0.116 0.055
Chain 1: 2100 -19334.676 0.085 0.037
Chain 1: 2200 -19560.635 0.084 0.037
Chain 1: 2300 -19178.447 0.040 0.020
Chain 1: 2400 -18950.788 0.040 0.020
Chain 1: 2500 -18752.874 0.025 0.016
Chain 1: 2600 -18383.617 0.024 0.016
Chain 1: 2700 -18340.800 0.019 0.012
Chain 1: 2800 -18057.995 0.020 0.016
Chain 1: 2900 -18338.882 0.020 0.015
Chain 1: 3000 -18325.125 0.012 0.012
Chain 1: 3100 -18410.029 0.011 0.012
Chain 1: 3200 -18101.119 0.012 0.015
Chain 1: 3300 -18305.525 0.011 0.012
Chain 1: 3400 -17781.230 0.013 0.015
Chain 1: 3500 -18391.951 0.015 0.016
Chain 1: 3600 -17700.107 0.017 0.016
Chain 1: 3700 -18085.800 0.019 0.017
Chain 1: 3800 -17047.860 0.023 0.021
Chain 1: 3900 -17044.084 0.022 0.021
Chain 1: 4000 -17161.354 0.022 0.021
Chain 1: 4100 -17075.247 0.022 0.021
Chain 1: 4200 -16892.031 0.022 0.021
Chain 1: 4300 -17030.044 0.022 0.021
Chain 1: 4400 -16987.275 0.019 0.011
Chain 1: 4500 -16889.906 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00133 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.3 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49502.322 1.000 1.000
Chain 1: 200 -13249.253 1.868 2.736
Chain 1: 300 -16248.636 1.307 1.000
Chain 1: 400 -27614.208 1.083 1.000
Chain 1: 500 -15623.583 1.020 0.767
Chain 1: 600 -20105.171 0.887 0.767
Chain 1: 700 -15142.266 0.807 0.412
Chain 1: 800 -21388.190 0.743 0.412
Chain 1: 900 -12497.242 0.739 0.412
Chain 1: 1000 -13566.300 0.673 0.412
Chain 1: 1100 -10792.592 0.599 0.328 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -12179.130 0.337 0.292
Chain 1: 1300 -13913.137 0.331 0.292
Chain 1: 1400 -12344.201 0.302 0.257
Chain 1: 1500 -9953.503 0.250 0.240
Chain 1: 1600 -10039.255 0.228 0.240
Chain 1: 1700 -17113.230 0.237 0.240
Chain 1: 1800 -9892.839 0.280 0.240
Chain 1: 1900 -10005.563 0.210 0.127
Chain 1: 2000 -17792.647 0.246 0.240
Chain 1: 2100 -10874.046 0.284 0.240
Chain 1: 2200 -9761.697 0.284 0.240
Chain 1: 2300 -11356.252 0.286 0.240
Chain 1: 2400 -9300.730 0.295 0.240
Chain 1: 2500 -12851.609 0.299 0.276
Chain 1: 2600 -18090.087 0.327 0.290
Chain 1: 2700 -9543.490 0.375 0.290
Chain 1: 2800 -11109.755 0.316 0.276
Chain 1: 2900 -9206.975 0.336 0.276
Chain 1: 3000 -9626.057 0.296 0.221
Chain 1: 3100 -11057.766 0.246 0.207
Chain 1: 3200 -9418.954 0.252 0.207
Chain 1: 3300 -10671.033 0.249 0.207
Chain 1: 3400 -9237.265 0.243 0.174
Chain 1: 3500 -9312.722 0.216 0.155
Chain 1: 3600 -9313.440 0.187 0.141
Chain 1: 3700 -9147.894 0.099 0.129
Chain 1: 3800 -8985.850 0.087 0.117
Chain 1: 3900 -8940.873 0.067 0.044
Chain 1: 4000 -10042.124 0.074 0.110
Chain 1: 4100 -11190.246 0.071 0.103
Chain 1: 4200 -10319.274 0.062 0.084
Chain 1: 4300 -10464.355 0.052 0.018
Chain 1: 4400 -13474.336 0.058 0.018
Chain 1: 4500 -12930.673 0.062 0.042
Chain 1: 4600 -8818.936 0.108 0.084
Chain 1: 4700 -12891.535 0.138 0.103
Chain 1: 4800 -9166.954 0.177 0.110
Chain 1: 4900 -9076.411 0.177 0.110
Chain 1: 5000 -9558.308 0.172 0.103
Chain 1: 5100 -10857.903 0.173 0.120
Chain 1: 5200 -10919.897 0.165 0.120
Chain 1: 5300 -10995.678 0.165 0.120
Chain 1: 5400 -9459.501 0.159 0.120
Chain 1: 5500 -8839.526 0.161 0.120
Chain 1: 5600 -9013.846 0.117 0.070
Chain 1: 5700 -13061.108 0.116 0.070
Chain 1: 5800 -8607.552 0.127 0.070
Chain 1: 5900 -15475.439 0.171 0.120
Chain 1: 6000 -8622.138 0.245 0.162
Chain 1: 6100 -9598.853 0.243 0.162
Chain 1: 6200 -8654.260 0.254 0.162
Chain 1: 6300 -14361.774 0.293 0.310
Chain 1: 6400 -8768.312 0.340 0.397
Chain 1: 6500 -10616.077 0.351 0.397
Chain 1: 6600 -11340.644 0.355 0.397
Chain 1: 6700 -9417.162 0.344 0.397
Chain 1: 6800 -13705.299 0.324 0.313
Chain 1: 6900 -11958.826 0.294 0.204
Chain 1: 7000 -13261.230 0.225 0.174
Chain 1: 7100 -8627.393 0.268 0.204
Chain 1: 7200 -8446.318 0.259 0.204
Chain 1: 7300 -8576.031 0.221 0.174
Chain 1: 7400 -8764.494 0.159 0.146
Chain 1: 7500 -8639.638 0.143 0.098
Chain 1: 7600 -9208.331 0.143 0.098
Chain 1: 7700 -9763.242 0.129 0.062
Chain 1: 7800 -9046.136 0.105 0.062
Chain 1: 7900 -9706.706 0.097 0.062
Chain 1: 8000 -8590.670 0.101 0.062
Chain 1: 8100 -8565.791 0.047 0.057
Chain 1: 8200 -9200.765 0.052 0.062
Chain 1: 8300 -8360.856 0.060 0.068
Chain 1: 8400 -11613.729 0.086 0.069
Chain 1: 8500 -8489.794 0.122 0.079
Chain 1: 8600 -11412.748 0.141 0.100
Chain 1: 8700 -10100.108 0.148 0.130
Chain 1: 8800 -9307.060 0.149 0.130
Chain 1: 8900 -8510.605 0.152 0.130
Chain 1: 9000 -9740.693 0.151 0.126
Chain 1: 9100 -8380.253 0.167 0.130
Chain 1: 9200 -9294.101 0.170 0.130
Chain 1: 9300 -8745.949 0.166 0.130
Chain 1: 9400 -9867.935 0.150 0.126
Chain 1: 9500 -12315.487 0.133 0.126
Chain 1: 9600 -8725.649 0.148 0.126
Chain 1: 9700 -9815.599 0.146 0.114
Chain 1: 9800 -8766.094 0.150 0.120
Chain 1: 9900 -11943.587 0.167 0.126
Chain 1: 10000 -8470.715 0.195 0.162
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.0014 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -63617.147 1.000 1.000
Chain 1: 200 -18386.862 1.730 2.460
Chain 1: 300 -8836.055 1.514 1.081
Chain 1: 400 -8573.378 1.143 1.081
Chain 1: 500 -8279.645 0.921 1.000
Chain 1: 600 -9111.119 0.783 1.000
Chain 1: 700 -7705.731 0.697 0.182
Chain 1: 800 -7701.668 0.610 0.182
Chain 1: 900 -7653.648 0.543 0.091
Chain 1: 1000 -7694.010 0.489 0.091
Chain 1: 1100 -7575.563 0.391 0.035
Chain 1: 1200 -7653.589 0.146 0.031
Chain 1: 1300 -7768.849 0.039 0.016
Chain 1: 1400 -7817.643 0.037 0.015
Chain 1: 1500 -7494.695 0.038 0.015
Chain 1: 1600 -7678.064 0.031 0.015
Chain 1: 1700 -7483.582 0.015 0.015
Chain 1: 1800 -7532.005 0.016 0.015
Chain 1: 1900 -7527.510 0.015 0.015
Chain 1: 2000 -7617.794 0.016 0.015
Chain 1: 2100 -7515.359 0.016 0.014
Chain 1: 2200 -7650.654 0.016 0.015
Chain 1: 2300 -7503.332 0.017 0.018
Chain 1: 2400 -7563.347 0.017 0.018
Chain 1: 2500 -7561.383 0.013 0.014
Chain 1: 2600 -7462.721 0.012 0.013
Chain 1: 2700 -7506.989 0.010 0.012 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003281 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.81 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86656.569 1.000 1.000
Chain 1: 200 -13719.355 3.158 5.316
Chain 1: 300 -10050.559 2.227 1.000
Chain 1: 400 -10987.960 1.692 1.000
Chain 1: 500 -9036.536 1.397 0.365
Chain 1: 600 -8497.856 1.174 0.365
Chain 1: 700 -8701.575 1.010 0.216
Chain 1: 800 -9336.779 0.892 0.216
Chain 1: 900 -8827.423 0.799 0.085
Chain 1: 1000 -8658.082 0.721 0.085
Chain 1: 1100 -8695.897 0.622 0.068 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8477.482 0.093 0.063
Chain 1: 1300 -8681.011 0.059 0.058
Chain 1: 1400 -8700.525 0.050 0.026
Chain 1: 1500 -8589.764 0.030 0.023
Chain 1: 1600 -8697.465 0.025 0.023
Chain 1: 1700 -8779.386 0.024 0.020
Chain 1: 1800 -8353.540 0.022 0.020
Chain 1: 1900 -8455.899 0.017 0.013
Chain 1: 2000 -8430.549 0.016 0.012
Chain 1: 2100 -8556.991 0.017 0.013
Chain 1: 2200 -8357.166 0.017 0.013
Chain 1: 2300 -8450.875 0.015 0.012
Chain 1: 2400 -8519.194 0.016 0.012
Chain 1: 2500 -8465.440 0.015 0.012
Chain 1: 2600 -8467.461 0.014 0.011
Chain 1: 2700 -8383.868 0.014 0.011
Chain 1: 2800 -8342.893 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00311 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.1 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8411716.363 1.000 1.000
Chain 1: 200 -1585334.463 2.653 4.306
Chain 1: 300 -890315.817 2.029 1.000
Chain 1: 400 -457166.594 1.759 1.000
Chain 1: 500 -357524.337 1.463 0.947
Chain 1: 600 -232625.474 1.308 0.947
Chain 1: 700 -119168.929 1.257 0.947
Chain 1: 800 -86443.422 1.148 0.947
Chain 1: 900 -66847.706 1.053 0.781
Chain 1: 1000 -51690.938 0.977 0.781
Chain 1: 1100 -39209.791 0.908 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38393.720 0.480 0.379
Chain 1: 1300 -26390.905 0.447 0.379
Chain 1: 1400 -26115.179 0.354 0.318
Chain 1: 1500 -22712.225 0.341 0.318
Chain 1: 1600 -21931.879 0.291 0.293
Chain 1: 1700 -20810.483 0.201 0.293
Chain 1: 1800 -20755.975 0.163 0.150
Chain 1: 1900 -21082.259 0.136 0.054
Chain 1: 2000 -19595.517 0.114 0.054
Chain 1: 2100 -19833.953 0.083 0.036
Chain 1: 2200 -20059.974 0.082 0.036
Chain 1: 2300 -19677.501 0.039 0.019
Chain 1: 2400 -19449.570 0.039 0.019
Chain 1: 2500 -19251.322 0.025 0.015
Chain 1: 2600 -18881.667 0.023 0.015
Chain 1: 2700 -18838.684 0.018 0.012
Chain 1: 2800 -18555.337 0.019 0.015
Chain 1: 2900 -18836.598 0.019 0.015
Chain 1: 3000 -18822.891 0.012 0.012
Chain 1: 3100 -18907.860 0.011 0.012
Chain 1: 3200 -18598.516 0.012 0.015
Chain 1: 3300 -18803.264 0.011 0.012
Chain 1: 3400 -18278.018 0.012 0.015
Chain 1: 3500 -18890.034 0.015 0.015
Chain 1: 3600 -18196.508 0.016 0.015
Chain 1: 3700 -18583.417 0.018 0.017
Chain 1: 3800 -17542.732 0.023 0.021
Chain 1: 3900 -17538.806 0.021 0.021
Chain 1: 4000 -17656.172 0.022 0.021
Chain 1: 4100 -17569.869 0.022 0.021
Chain 1: 4200 -17386.040 0.021 0.021
Chain 1: 4300 -17524.536 0.021 0.021
Chain 1: 4400 -17481.297 0.018 0.011
Chain 1: 4500 -17383.764 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001238 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.38 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13191.040 1.000 1.000
Chain 1: 200 -9821.810 0.672 1.000
Chain 1: 300 -8451.799 0.502 0.343
Chain 1: 400 -8166.970 0.385 0.343
Chain 1: 500 -8035.519 0.311 0.162
Chain 1: 600 -8050.399 0.260 0.162
Chain 1: 700 -7979.198 0.224 0.035
Chain 1: 800 -7947.594 0.196 0.035
Chain 1: 900 -8222.347 0.178 0.033
Chain 1: 1000 -8042.390 0.163 0.033
Chain 1: 1100 -8085.495 0.063 0.022
Chain 1: 1200 -7980.232 0.030 0.016
Chain 1: 1300 -7962.118 0.014 0.013
Chain 1: 1400 -7970.440 0.011 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001437 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.37 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58387.259 1.000 1.000
Chain 1: 200 -17925.249 1.629 2.257
Chain 1: 300 -8812.756 1.430 1.034
Chain 1: 400 -8244.125 1.090 1.034
Chain 1: 500 -8502.110 0.878 1.000
Chain 1: 600 -8778.640 0.737 1.000
Chain 1: 700 -8425.131 0.638 0.069
Chain 1: 800 -8396.302 0.558 0.069
Chain 1: 900 -7937.206 0.503 0.058
Chain 1: 1000 -7887.573 0.453 0.058
Chain 1: 1100 -7737.637 0.355 0.042
Chain 1: 1200 -7722.367 0.130 0.032
Chain 1: 1300 -7914.519 0.029 0.030
Chain 1: 1400 -7964.735 0.022 0.024
Chain 1: 1500 -7669.611 0.023 0.024
Chain 1: 1600 -7822.657 0.022 0.020
Chain 1: 1700 -7620.586 0.020 0.020
Chain 1: 1800 -7657.910 0.021 0.020
Chain 1: 1900 -7677.981 0.015 0.019
Chain 1: 2000 -7640.338 0.015 0.019
Chain 1: 2100 -7660.243 0.013 0.006 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002495 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 24.95 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86254.863 1.000 1.000
Chain 1: 200 -13674.171 3.154 5.308
Chain 1: 300 -10003.764 2.225 1.000
Chain 1: 400 -10964.310 1.691 1.000
Chain 1: 500 -8996.543 1.396 0.367
Chain 1: 600 -8739.750 1.168 0.367
Chain 1: 700 -8488.368 1.006 0.219
Chain 1: 800 -8662.043 0.883 0.219
Chain 1: 900 -8788.008 0.786 0.088
Chain 1: 1000 -8522.553 0.711 0.088
Chain 1: 1100 -8816.268 0.614 0.033 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8410.451 0.088 0.033
Chain 1: 1300 -8692.777 0.054 0.032
Chain 1: 1400 -8695.815 0.046 0.031
Chain 1: 1500 -8538.544 0.026 0.030
Chain 1: 1600 -8651.966 0.024 0.030
Chain 1: 1700 -8728.853 0.022 0.020
Chain 1: 1800 -8300.798 0.025 0.031
Chain 1: 1900 -8403.827 0.025 0.031
Chain 1: 2000 -8378.945 0.022 0.018
Chain 1: 2100 -8506.537 0.020 0.015
Chain 1: 2200 -8304.742 0.018 0.015
Chain 1: 2300 -8399.679 0.016 0.013
Chain 1: 2400 -8467.047 0.017 0.013
Chain 1: 2500 -8413.164 0.015 0.012
Chain 1: 2600 -8415.985 0.014 0.011
Chain 1: 2700 -8332.000 0.014 0.011
Chain 1: 2800 -8290.141 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003229 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.29 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8402999.641 1.000 1.000
Chain 1: 200 -1583826.447 2.653 4.306
Chain 1: 300 -890076.306 2.028 1.000
Chain 1: 400 -457076.275 1.758 1.000
Chain 1: 500 -357585.429 1.462 0.947
Chain 1: 600 -232897.519 1.308 0.947
Chain 1: 700 -119333.654 1.257 0.947
Chain 1: 800 -86546.368 1.147 0.947
Chain 1: 900 -66924.235 1.052 0.779
Chain 1: 1000 -51738.581 0.976 0.779
Chain 1: 1100 -39225.410 0.908 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38406.778 0.480 0.379
Chain 1: 1300 -26375.024 0.447 0.379
Chain 1: 1400 -26096.290 0.354 0.319
Chain 1: 1500 -22685.198 0.341 0.319
Chain 1: 1600 -21902.241 0.291 0.294
Chain 1: 1700 -20777.419 0.201 0.293
Chain 1: 1800 -20722.048 0.164 0.150
Chain 1: 1900 -21048.392 0.136 0.054
Chain 1: 2000 -19559.633 0.114 0.054
Chain 1: 2100 -19798.267 0.083 0.036
Chain 1: 2200 -20024.542 0.082 0.036
Chain 1: 2300 -19641.821 0.039 0.019
Chain 1: 2400 -19413.831 0.039 0.019
Chain 1: 2500 -19215.633 0.025 0.016
Chain 1: 2600 -18845.919 0.023 0.016
Chain 1: 2700 -18802.881 0.018 0.012
Chain 1: 2800 -18519.500 0.019 0.015
Chain 1: 2900 -18800.880 0.019 0.015
Chain 1: 3000 -18787.142 0.012 0.012
Chain 1: 3100 -18872.113 0.011 0.012
Chain 1: 3200 -18562.742 0.012 0.015
Chain 1: 3300 -18767.509 0.011 0.012
Chain 1: 3400 -18242.204 0.012 0.015
Chain 1: 3500 -18854.351 0.015 0.015
Chain 1: 3600 -18160.697 0.016 0.015
Chain 1: 3700 -18547.726 0.018 0.017
Chain 1: 3800 -17506.826 0.023 0.021
Chain 1: 3900 -17502.908 0.021 0.021
Chain 1: 4000 -17620.265 0.022 0.021
Chain 1: 4100 -17533.942 0.022 0.021
Chain 1: 4200 -17350.063 0.021 0.021
Chain 1: 4300 -17488.599 0.021 0.021
Chain 1: 4400 -17445.340 0.018 0.011
Chain 1: 4500 -17347.793 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001401 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.01 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49334.292 1.000 1.000
Chain 1: 200 -16945.215 1.456 1.911
Chain 1: 300 -17518.650 0.981 1.000
Chain 1: 400 -13160.668 0.819 1.000
Chain 1: 500 -16553.242 0.696 0.331
Chain 1: 600 -16101.239 0.585 0.331
Chain 1: 700 -15834.643 0.504 0.205
Chain 1: 800 -15174.014 0.446 0.205
Chain 1: 900 -15006.954 0.398 0.044
Chain 1: 1000 -13612.028 0.368 0.102
Chain 1: 1100 -11110.758 0.291 0.102
Chain 1: 1200 -12755.118 0.112 0.102
Chain 1: 1300 -10481.554 0.131 0.129
Chain 1: 1400 -16257.232 0.133 0.129
Chain 1: 1500 -10250.414 0.171 0.129
Chain 1: 1600 -11712.921 0.181 0.129
Chain 1: 1700 -12246.914 0.184 0.129
Chain 1: 1800 -26084.576 0.232 0.217
Chain 1: 1900 -10000.076 0.392 0.225
Chain 1: 2000 -10115.923 0.383 0.225
Chain 1: 2100 -10264.617 0.362 0.217
Chain 1: 2200 -14239.780 0.377 0.279
Chain 1: 2300 -12652.333 0.368 0.279
Chain 1: 2400 -10271.260 0.356 0.232
Chain 1: 2500 -10624.041 0.300 0.125
Chain 1: 2600 -10258.611 0.291 0.125
Chain 1: 2700 -13726.682 0.312 0.232
Chain 1: 2800 -9667.224 0.301 0.232
Chain 1: 2900 -10489.817 0.148 0.125
Chain 1: 3000 -13084.652 0.167 0.198
Chain 1: 3100 -9109.569 0.209 0.232
Chain 1: 3200 -9839.272 0.189 0.198
Chain 1: 3300 -9932.959 0.177 0.198
Chain 1: 3400 -16646.975 0.194 0.198
Chain 1: 3500 -11573.090 0.235 0.253
Chain 1: 3600 -10502.369 0.241 0.253
Chain 1: 3700 -11127.991 0.222 0.198
Chain 1: 3800 -9266.298 0.200 0.198
Chain 1: 3900 -9170.244 0.193 0.198
Chain 1: 4000 -9936.348 0.181 0.102
Chain 1: 4100 -10508.244 0.143 0.077
Chain 1: 4200 -15732.091 0.168 0.102
Chain 1: 4300 -14164.991 0.179 0.111
Chain 1: 4400 -9017.298 0.195 0.111
Chain 1: 4500 -9195.386 0.153 0.102
Chain 1: 4600 -14939.231 0.182 0.111
Chain 1: 4700 -10125.321 0.224 0.201
Chain 1: 4800 -8845.068 0.218 0.145
Chain 1: 4900 -9251.243 0.221 0.145
Chain 1: 5000 -8989.309 0.217 0.145
Chain 1: 5100 -13343.486 0.244 0.326
Chain 1: 5200 -10153.572 0.242 0.314
Chain 1: 5300 -10749.583 0.236 0.314
Chain 1: 5400 -14703.889 0.206 0.269
Chain 1: 5500 -10851.342 0.240 0.314
Chain 1: 5600 -8670.407 0.226 0.269
Chain 1: 5700 -12619.733 0.210 0.269
Chain 1: 5800 -8597.037 0.243 0.313
Chain 1: 5900 -11813.287 0.265 0.313
Chain 1: 6000 -8973.312 0.294 0.314
Chain 1: 6100 -9054.696 0.262 0.313
Chain 1: 6200 -8603.962 0.236 0.272
Chain 1: 6300 -9526.649 0.240 0.272
Chain 1: 6400 -8600.430 0.224 0.272
Chain 1: 6500 -8912.386 0.192 0.252
Chain 1: 6600 -11047.075 0.186 0.193
Chain 1: 6700 -10087.681 0.165 0.108
Chain 1: 6800 -8637.841 0.135 0.108
Chain 1: 6900 -10364.425 0.124 0.108
Chain 1: 7000 -8567.179 0.113 0.108
Chain 1: 7100 -15572.243 0.157 0.167
Chain 1: 7200 -9499.808 0.216 0.168
Chain 1: 7300 -9456.185 0.207 0.168
Chain 1: 7400 -12750.060 0.222 0.193
Chain 1: 7500 -8483.511 0.269 0.210
Chain 1: 7600 -9311.373 0.258 0.210
Chain 1: 7700 -10905.449 0.263 0.210
Chain 1: 7800 -9087.311 0.267 0.210
Chain 1: 7900 -8612.342 0.256 0.210
Chain 1: 8000 -9484.835 0.244 0.200
Chain 1: 8100 -8653.112 0.208 0.146
Chain 1: 8200 -11113.693 0.167 0.146
Chain 1: 8300 -8463.329 0.197 0.200
Chain 1: 8400 -12399.250 0.203 0.200
Chain 1: 8500 -8713.525 0.195 0.200
Chain 1: 8600 -9212.724 0.192 0.200
Chain 1: 8700 -9728.117 0.183 0.200
Chain 1: 8800 -9380.841 0.166 0.096
Chain 1: 8900 -11216.979 0.177 0.164
Chain 1: 9000 -11554.096 0.171 0.164
Chain 1: 9100 -9104.147 0.188 0.221
Chain 1: 9200 -8416.975 0.174 0.164
Chain 1: 9300 -11091.841 0.167 0.164
Chain 1: 9400 -9308.853 0.154 0.164
Chain 1: 9500 -8460.797 0.122 0.100
Chain 1: 9600 -8600.597 0.118 0.100
Chain 1: 9700 -10118.484 0.128 0.150
Chain 1: 9800 -9117.199 0.135 0.150
Chain 1: 9900 -9848.171 0.126 0.110
Chain 1: 10000 -10425.048 0.129 0.110
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004038 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 40.38 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61920.924 1.000 1.000
Chain 1: 200 -18111.430 1.709 2.419
Chain 1: 300 -9016.006 1.476 1.009
Chain 1: 400 -9739.714 1.125 1.009
Chain 1: 500 -8114.212 0.940 1.000
Chain 1: 600 -8508.400 0.791 1.000
Chain 1: 700 -8186.700 0.684 0.200
Chain 1: 800 -8356.167 0.601 0.200
Chain 1: 900 -8167.744 0.537 0.074
Chain 1: 1000 -7820.673 0.488 0.074
Chain 1: 1100 -7866.706 0.388 0.046
Chain 1: 1200 -7801.771 0.147 0.044
Chain 1: 1300 -7831.078 0.047 0.039
Chain 1: 1400 -7719.184 0.041 0.023
Chain 1: 1500 -7625.538 0.022 0.020
Chain 1: 1600 -7833.666 0.020 0.020
Chain 1: 1700 -7576.947 0.019 0.020
Chain 1: 1800 -7711.798 0.019 0.017
Chain 1: 1900 -7794.442 0.018 0.014
Chain 1: 2000 -7721.020 0.014 0.012
Chain 1: 2100 -7667.263 0.014 0.012
Chain 1: 2200 -7811.679 0.015 0.014
Chain 1: 2300 -7635.255 0.017 0.017
Chain 1: 2400 -7704.038 0.017 0.017
Chain 1: 2500 -7717.317 0.016 0.017
Chain 1: 2600 -7608.445 0.015 0.014
Chain 1: 2700 -7661.772 0.012 0.011
Chain 1: 2800 -7711.997 0.011 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003372 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.72 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86695.928 1.000 1.000
Chain 1: 200 -13776.559 3.147 5.293
Chain 1: 300 -10127.959 2.218 1.000
Chain 1: 400 -11071.482 1.685 1.000
Chain 1: 500 -9108.083 1.391 0.360
Chain 1: 600 -9003.628 1.161 0.360
Chain 1: 700 -8689.129 1.000 0.216
Chain 1: 800 -9131.277 0.881 0.216
Chain 1: 900 -8924.164 0.786 0.085
Chain 1: 1000 -8800.154 0.709 0.085
Chain 1: 1100 -8930.546 0.610 0.048 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8479.012 0.086 0.048
Chain 1: 1300 -8686.519 0.053 0.036
Chain 1: 1400 -8808.951 0.045 0.024
Chain 1: 1500 -8697.360 0.025 0.023
Chain 1: 1600 -8809.237 0.025 0.023
Chain 1: 1700 -8883.770 0.023 0.015
Chain 1: 1800 -8468.947 0.023 0.015
Chain 1: 1900 -8565.392 0.021 0.014
Chain 1: 2000 -8538.917 0.020 0.014
Chain 1: 2100 -8662.472 0.020 0.014
Chain 1: 2200 -8480.048 0.017 0.014
Chain 1: 2300 -8559.870 0.016 0.013
Chain 1: 2400 -8629.554 0.015 0.013
Chain 1: 2500 -8575.366 0.014 0.011
Chain 1: 2600 -8575.319 0.013 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003391 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.91 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8415613.928 1.000 1.000
Chain 1: 200 -1583787.301 2.657 4.314
Chain 1: 300 -890691.394 2.031 1.000
Chain 1: 400 -458183.585 1.759 1.000
Chain 1: 500 -358635.009 1.463 0.944
Chain 1: 600 -233546.133 1.308 0.944
Chain 1: 700 -119618.252 1.257 0.944
Chain 1: 800 -86840.953 1.147 0.944
Chain 1: 900 -67149.896 1.052 0.778
Chain 1: 1000 -51923.364 0.977 0.778
Chain 1: 1100 -39382.669 0.908 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38554.527 0.479 0.377
Chain 1: 1300 -26484.505 0.447 0.377
Chain 1: 1400 -26201.282 0.354 0.318
Chain 1: 1500 -22783.458 0.341 0.318
Chain 1: 1600 -21998.972 0.291 0.293
Chain 1: 1700 -20869.260 0.201 0.293
Chain 1: 1800 -20812.673 0.164 0.150
Chain 1: 1900 -21138.888 0.136 0.054
Chain 1: 2000 -19648.668 0.114 0.054
Chain 1: 2100 -19886.771 0.083 0.036
Chain 1: 2200 -20113.795 0.082 0.036
Chain 1: 2300 -19730.540 0.039 0.019
Chain 1: 2400 -19502.633 0.039 0.019
Chain 1: 2500 -19304.977 0.025 0.015
Chain 1: 2600 -18934.766 0.023 0.015
Chain 1: 2700 -18891.640 0.018 0.012
Chain 1: 2800 -18608.700 0.019 0.015
Chain 1: 2900 -18889.965 0.019 0.015
Chain 1: 3000 -18875.995 0.012 0.012
Chain 1: 3100 -18961.052 0.011 0.012
Chain 1: 3200 -18651.614 0.012 0.015
Chain 1: 3300 -18856.440 0.011 0.012
Chain 1: 3400 -18331.336 0.012 0.015
Chain 1: 3500 -18943.317 0.015 0.015
Chain 1: 3600 -18249.853 0.016 0.015
Chain 1: 3700 -18636.840 0.018 0.017
Chain 1: 3800 -17596.368 0.023 0.021
Chain 1: 3900 -17592.571 0.021 0.021
Chain 1: 4000 -17709.809 0.022 0.021
Chain 1: 4100 -17623.622 0.022 0.021
Chain 1: 4200 -17439.820 0.021 0.021
Chain 1: 4300 -17578.186 0.021 0.021
Chain 1: 4400 -17534.970 0.018 0.011
Chain 1: 4500 -17437.560 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001411 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.11 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48637.477 1.000 1.000
Chain 1: 200 -18338.926 1.326 1.652
Chain 1: 300 -17537.522 0.899 1.000
Chain 1: 400 -15123.837 0.714 1.000
Chain 1: 500 -18405.310 0.607 0.178
Chain 1: 600 -11310.412 0.611 0.627
Chain 1: 700 -12339.166 0.535 0.178
Chain 1: 800 -16566.453 0.500 0.255
Chain 1: 900 -22123.284 0.473 0.251
Chain 1: 1000 -12904.626 0.497 0.255
Chain 1: 1100 -19724.521 0.431 0.255
Chain 1: 1200 -19128.436 0.269 0.251
Chain 1: 1300 -10940.120 0.339 0.255
Chain 1: 1400 -10295.231 0.330 0.255
Chain 1: 1500 -10768.287 0.316 0.255
Chain 1: 1600 -12201.421 0.265 0.251
Chain 1: 1700 -9235.999 0.289 0.255
Chain 1: 1800 -13001.113 0.293 0.290
Chain 1: 1900 -10090.383 0.296 0.290
Chain 1: 2000 -10290.007 0.227 0.288
Chain 1: 2100 -10300.759 0.192 0.117
Chain 1: 2200 -9481.302 0.198 0.117
Chain 1: 2300 -15014.971 0.160 0.117
Chain 1: 2400 -13713.273 0.163 0.117
Chain 1: 2500 -10281.934 0.192 0.288
Chain 1: 2600 -9548.344 0.188 0.288
Chain 1: 2700 -9256.326 0.159 0.095
Chain 1: 2800 -10277.013 0.140 0.095
Chain 1: 2900 -10070.059 0.113 0.086
Chain 1: 3000 -8604.194 0.128 0.095
Chain 1: 3100 -11239.051 0.152 0.099
Chain 1: 3200 -10755.201 0.148 0.099
Chain 1: 3300 -10315.153 0.115 0.095
Chain 1: 3400 -8832.935 0.122 0.099
Chain 1: 3500 -8679.780 0.091 0.077
Chain 1: 3600 -9492.194 0.091 0.086
Chain 1: 3700 -9150.311 0.092 0.086
Chain 1: 3800 -10030.087 0.091 0.086
Chain 1: 3900 -8970.018 0.101 0.088
Chain 1: 4000 -11521.239 0.106 0.088
Chain 1: 4100 -9065.066 0.109 0.088
Chain 1: 4200 -9468.070 0.109 0.088
Chain 1: 4300 -10214.582 0.112 0.088
Chain 1: 4400 -8666.344 0.113 0.088
Chain 1: 4500 -10724.635 0.131 0.118
Chain 1: 4600 -8260.446 0.152 0.179
Chain 1: 4700 -9926.305 0.165 0.179
Chain 1: 4800 -10914.839 0.165 0.179
Chain 1: 4900 -9230.216 0.172 0.183
Chain 1: 5000 -9036.879 0.152 0.179
Chain 1: 5100 -8445.388 0.132 0.168
Chain 1: 5200 -8551.276 0.129 0.168
Chain 1: 5300 -11480.380 0.147 0.179
Chain 1: 5400 -8842.072 0.159 0.183
Chain 1: 5500 -9339.099 0.145 0.168
Chain 1: 5600 -8542.651 0.124 0.093
Chain 1: 5700 -8790.484 0.111 0.091
Chain 1: 5800 -9787.824 0.112 0.093
Chain 1: 5900 -14748.095 0.127 0.093
Chain 1: 6000 -8744.943 0.194 0.102
Chain 1: 6100 -10162.838 0.200 0.140
Chain 1: 6200 -9362.363 0.208 0.140
Chain 1: 6300 -10751.787 0.195 0.129
Chain 1: 6400 -11282.007 0.170 0.102
Chain 1: 6500 -8714.526 0.194 0.129
Chain 1: 6600 -8160.424 0.192 0.129
Chain 1: 6700 -12720.977 0.225 0.140
Chain 1: 6800 -8976.262 0.256 0.295
Chain 1: 6900 -11497.631 0.245 0.219
Chain 1: 7000 -9210.058 0.201 0.219
Chain 1: 7100 -8128.030 0.200 0.219
Chain 1: 7200 -8372.210 0.194 0.219
Chain 1: 7300 -9486.922 0.193 0.219
Chain 1: 7400 -10630.136 0.199 0.219
Chain 1: 7500 -8830.866 0.190 0.204
Chain 1: 7600 -8549.028 0.187 0.204
Chain 1: 7700 -8654.944 0.152 0.133
Chain 1: 7800 -8696.230 0.111 0.117
Chain 1: 7900 -8107.846 0.096 0.108
Chain 1: 8000 -8660.093 0.078 0.073
Chain 1: 8100 -8045.245 0.072 0.073
Chain 1: 8200 -9354.406 0.083 0.076
Chain 1: 8300 -9434.059 0.072 0.073
Chain 1: 8400 -8111.683 0.078 0.073
Chain 1: 8500 -8653.526 0.064 0.064
Chain 1: 8600 -8441.285 0.063 0.064
Chain 1: 8700 -8613.561 0.064 0.064
Chain 1: 8800 -8099.453 0.070 0.064
Chain 1: 8900 -8362.308 0.065 0.063
Chain 1: 9000 -8506.271 0.061 0.063
Chain 1: 9100 -10817.688 0.074 0.063
Chain 1: 9200 -9039.889 0.080 0.063
Chain 1: 9300 -9508.138 0.084 0.063
Chain 1: 9400 -8168.816 0.084 0.063
Chain 1: 9500 -8371.755 0.080 0.049
Chain 1: 9600 -8128.189 0.081 0.049
Chain 1: 9700 -9426.071 0.093 0.063
Chain 1: 9800 -8424.486 0.098 0.119
Chain 1: 9900 -9024.786 0.102 0.119
Chain 1: 10000 -8279.603 0.109 0.119
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001698 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 16.98 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57940.470 1.000 1.000
Chain 1: 200 -17325.044 1.672 2.344
Chain 1: 300 -8520.811 1.459 1.033
Chain 1: 400 -8143.521 1.106 1.033
Chain 1: 500 -8262.905 0.888 1.000
Chain 1: 600 -8653.033 0.747 1.000
Chain 1: 700 -7971.590 0.653 0.085
Chain 1: 800 -8052.201 0.572 0.085
Chain 1: 900 -7954.860 0.510 0.046
Chain 1: 1000 -7732.914 0.462 0.046
Chain 1: 1100 -7634.015 0.363 0.045
Chain 1: 1200 -7571.071 0.130 0.029
Chain 1: 1300 -7736.873 0.028 0.021
Chain 1: 1400 -7815.307 0.025 0.014
Chain 1: 1500 -7600.185 0.026 0.021
Chain 1: 1600 -7523.417 0.023 0.013
Chain 1: 1700 -7512.794 0.014 0.012
Chain 1: 1800 -7566.828 0.014 0.012
Chain 1: 1900 -7584.006 0.013 0.010
Chain 1: 2000 -7607.029 0.011 0.010
Chain 1: 2100 -7588.124 0.009 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003229 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.29 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86258.151 1.000 1.000
Chain 1: 200 -13127.182 3.285 5.571
Chain 1: 300 -9638.011 2.311 1.000
Chain 1: 400 -10526.790 1.754 1.000
Chain 1: 500 -8480.494 1.452 0.362
Chain 1: 600 -8246.617 1.215 0.362
Chain 1: 700 -8557.599 1.046 0.241
Chain 1: 800 -8603.434 0.916 0.241
Chain 1: 900 -8530.205 0.815 0.084
Chain 1: 1000 -8275.698 0.737 0.084
Chain 1: 1100 -8511.097 0.640 0.036 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8249.784 0.086 0.032
Chain 1: 1300 -8413.658 0.051 0.031
Chain 1: 1400 -8335.002 0.044 0.028
Chain 1: 1500 -8290.026 0.020 0.028
Chain 1: 1600 -8290.063 0.017 0.019
Chain 1: 1700 -8232.292 0.015 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.007099 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 70.99 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8424329.354 1.000 1.000
Chain 1: 200 -1588826.691 2.651 4.302
Chain 1: 300 -890592.914 2.029 1.000
Chain 1: 400 -457408.230 1.758 1.000
Chain 1: 500 -357234.896 1.463 0.947
Chain 1: 600 -232151.084 1.309 0.947
Chain 1: 700 -118575.960 1.259 0.947
Chain 1: 800 -85849.539 1.149 0.947
Chain 1: 900 -66228.474 1.054 0.784
Chain 1: 1000 -51049.752 0.979 0.784
Chain 1: 1100 -38564.567 0.911 0.539 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37738.103 0.483 0.381
Chain 1: 1300 -25746.354 0.451 0.381
Chain 1: 1400 -25466.739 0.357 0.324
Chain 1: 1500 -22068.483 0.345 0.324
Chain 1: 1600 -21288.454 0.295 0.297
Chain 1: 1700 -20168.962 0.204 0.296
Chain 1: 1800 -20114.329 0.166 0.154
Chain 1: 1900 -20439.665 0.138 0.056
Chain 1: 2000 -18956.110 0.117 0.056
Chain 1: 2100 -19194.100 0.085 0.037
Chain 1: 2200 -19419.472 0.084 0.037
Chain 1: 2300 -19037.837 0.040 0.020
Chain 1: 2400 -18810.265 0.040 0.020
Chain 1: 2500 -18612.202 0.026 0.016
Chain 1: 2600 -18243.342 0.024 0.016
Chain 1: 2700 -18200.620 0.019 0.012
Chain 1: 2800 -17917.771 0.020 0.016
Chain 1: 2900 -18198.604 0.020 0.015
Chain 1: 3000 -18184.831 0.012 0.012
Chain 1: 3100 -18269.703 0.011 0.012
Chain 1: 3200 -17960.991 0.012 0.015
Chain 1: 3300 -18165.261 0.011 0.012
Chain 1: 3400 -17641.209 0.013 0.015
Chain 1: 3500 -18251.501 0.015 0.016
Chain 1: 3600 -17560.216 0.017 0.016
Chain 1: 3700 -17945.457 0.019 0.017
Chain 1: 3800 -16908.338 0.023 0.021
Chain 1: 3900 -16904.550 0.022 0.021
Chain 1: 4000 -17021.856 0.023 0.021
Chain 1: 4100 -16935.756 0.023 0.021
Chain 1: 4200 -16752.708 0.022 0.021
Chain 1: 4300 -16890.613 0.022 0.021
Chain 1: 4400 -16847.997 0.019 0.011
Chain 1: 4500 -16750.616 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.005875 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 58.75 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -50298.678 1.000 1.000
Chain 1: 200 -15751.723 1.597 2.193
Chain 1: 300 -48127.466 1.289 1.000
Chain 1: 400 -15271.914 1.504 2.151
Chain 1: 500 -14269.216 1.218 1.000
Chain 1: 600 -15701.853 1.030 1.000
Chain 1: 700 -16708.478 0.891 0.673
Chain 1: 800 -15833.661 0.787 0.673
Chain 1: 900 -13100.678 0.723 0.209
Chain 1: 1000 -15407.483 0.665 0.209
Chain 1: 1100 -12207.231 0.591 0.209 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -11359.017 0.380 0.150
Chain 1: 1300 -13574.150 0.329 0.150
Chain 1: 1400 -11519.004 0.131 0.150
Chain 1: 1500 -12320.353 0.131 0.150
Chain 1: 1600 -16014.604 0.145 0.163
Chain 1: 1700 -15494.239 0.142 0.163
Chain 1: 1800 -11444.540 0.172 0.178
Chain 1: 1900 -12413.696 0.159 0.163
Chain 1: 2000 -15159.816 0.162 0.178
Chain 1: 2100 -11351.916 0.169 0.178
Chain 1: 2200 -10642.572 0.169 0.178
Chain 1: 2300 -10936.225 0.155 0.178
Chain 1: 2400 -10736.108 0.139 0.078
Chain 1: 2500 -10588.033 0.134 0.078
Chain 1: 2600 -10385.901 0.113 0.067
Chain 1: 2700 -9903.507 0.114 0.067
Chain 1: 2800 -14372.098 0.110 0.067
Chain 1: 2900 -10452.616 0.140 0.067
Chain 1: 3000 -10035.811 0.126 0.049
Chain 1: 3100 -10610.583 0.098 0.049
Chain 1: 3200 -10540.807 0.092 0.042
Chain 1: 3300 -19823.966 0.136 0.049
Chain 1: 3400 -9726.389 0.238 0.054
Chain 1: 3500 -19861.400 0.287 0.311
Chain 1: 3600 -9948.813 0.385 0.375
Chain 1: 3700 -9891.846 0.381 0.375
Chain 1: 3800 -9385.941 0.355 0.375
Chain 1: 3900 -10439.182 0.328 0.101
Chain 1: 4000 -12158.082 0.338 0.141
Chain 1: 4100 -11700.848 0.336 0.141
Chain 1: 4200 -14002.775 0.352 0.164
Chain 1: 4300 -12165.922 0.320 0.151
Chain 1: 4400 -9671.462 0.242 0.151
Chain 1: 4500 -10746.582 0.201 0.141
Chain 1: 4600 -9764.414 0.111 0.101
Chain 1: 4700 -10007.452 0.113 0.101
Chain 1: 4800 -9201.481 0.117 0.101
Chain 1: 4900 -11348.052 0.126 0.141
Chain 1: 5000 -19211.038 0.152 0.151
Chain 1: 5100 -10195.383 0.237 0.164
Chain 1: 5200 -10590.911 0.224 0.151
Chain 1: 5300 -17072.499 0.247 0.189
Chain 1: 5400 -9193.431 0.307 0.189
Chain 1: 5500 -12302.208 0.322 0.253
Chain 1: 5600 -8982.542 0.349 0.370
Chain 1: 5700 -9253.399 0.350 0.370
Chain 1: 5800 -10070.353 0.349 0.370
Chain 1: 5900 -11075.941 0.339 0.370
Chain 1: 6000 -10404.554 0.305 0.253
Chain 1: 6100 -9818.710 0.222 0.091
Chain 1: 6200 -9882.386 0.219 0.091
Chain 1: 6300 -12539.884 0.202 0.091
Chain 1: 6400 -14611.306 0.131 0.091
Chain 1: 6500 -9022.730 0.167 0.091
Chain 1: 6600 -9381.493 0.134 0.081
Chain 1: 6700 -13223.174 0.160 0.091
Chain 1: 6800 -8793.893 0.203 0.142
Chain 1: 6900 -10186.974 0.207 0.142
Chain 1: 7000 -9516.750 0.208 0.142
Chain 1: 7100 -9930.429 0.206 0.142
Chain 1: 7200 -9017.844 0.216 0.142
Chain 1: 7300 -9279.050 0.197 0.137
Chain 1: 7400 -10998.032 0.199 0.137
Chain 1: 7500 -8934.888 0.160 0.137
Chain 1: 7600 -9212.737 0.159 0.137
Chain 1: 7700 -9721.722 0.135 0.101
Chain 1: 7800 -10810.847 0.095 0.101
Chain 1: 7900 -9042.945 0.101 0.101
Chain 1: 8000 -11008.935 0.112 0.101
Chain 1: 8100 -11149.432 0.109 0.101
Chain 1: 8200 -11547.265 0.102 0.101
Chain 1: 8300 -8918.958 0.129 0.156
Chain 1: 8400 -8918.399 0.113 0.101
Chain 1: 8500 -8835.787 0.091 0.052
Chain 1: 8600 -14360.466 0.126 0.101
Chain 1: 8700 -8813.300 0.184 0.179
Chain 1: 8800 -9972.539 0.186 0.179
Chain 1: 8900 -9857.734 0.167 0.116
Chain 1: 9000 -9370.832 0.155 0.052
Chain 1: 9100 -10104.934 0.161 0.073
Chain 1: 9200 -9570.270 0.163 0.073
Chain 1: 9300 -9136.380 0.138 0.056
Chain 1: 9400 -9521.939 0.142 0.056
Chain 1: 9500 -8523.217 0.153 0.073
Chain 1: 9600 -10765.361 0.135 0.073
Chain 1: 9700 -8699.704 0.096 0.073
Chain 1: 9800 -11309.554 0.107 0.073
Chain 1: 9900 -11354.865 0.107 0.073
Chain 1: 10000 -11228.878 0.103 0.073
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001441 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.41 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -47456.974 1.000 1.000
Chain 1: 200 -16708.229 1.420 1.840
Chain 1: 300 -9358.022 1.209 1.000
Chain 1: 400 -8392.712 0.935 1.000
Chain 1: 500 -8590.598 0.753 0.785
Chain 1: 600 -9288.834 0.640 0.785
Chain 1: 700 -9549.505 0.552 0.115
Chain 1: 800 -8454.916 0.499 0.129
Chain 1: 900 -8401.631 0.445 0.115
Chain 1: 1000 -7769.294 0.408 0.115
Chain 1: 1100 -7924.939 0.310 0.081
Chain 1: 1200 -8125.981 0.129 0.075
Chain 1: 1300 -8243.318 0.052 0.027
Chain 1: 1400 -8490.711 0.043 0.027
Chain 1: 1500 -7710.180 0.051 0.029
Chain 1: 1600 -7905.676 0.046 0.027
Chain 1: 1700 -7731.266 0.045 0.025
Chain 1: 1800 -7702.865 0.033 0.025
Chain 1: 1900 -7607.616 0.033 0.025
Chain 1: 2000 -7745.639 0.027 0.023
Chain 1: 2100 -7645.252 0.026 0.023
Chain 1: 2200 -8018.398 0.029 0.023
Chain 1: 2300 -7623.766 0.032 0.025
Chain 1: 2400 -7628.448 0.029 0.023
Chain 1: 2500 -7665.454 0.020 0.018
Chain 1: 2600 -7636.137 0.018 0.013
Chain 1: 2700 -7450.026 0.018 0.013
Chain 1: 2800 -7581.926 0.019 0.017
Chain 1: 2900 -7411.949 0.020 0.018
Chain 1: 3000 -7571.324 0.021 0.021
Chain 1: 3100 -7571.265 0.019 0.021
Chain 1: 3200 -7785.532 0.017 0.021
Chain 1: 3300 -7491.311 0.016 0.021
Chain 1: 3400 -7822.425 0.020 0.023
Chain 1: 3500 -7456.938 0.025 0.025
Chain 1: 3600 -7519.041 0.025 0.025
Chain 1: 3700 -7478.295 0.023 0.023
Chain 1: 3800 -7490.703 0.022 0.023
Chain 1: 3900 -7453.105 0.020 0.021
Chain 1: 4000 -7427.443 0.018 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003516 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.16 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87958.313 1.000 1.000
Chain 1: 200 -14676.423 2.997 4.993
Chain 1: 300 -10688.996 2.122 1.000
Chain 1: 400 -13565.455 1.645 1.000
Chain 1: 500 -8896.412 1.421 0.525
Chain 1: 600 -9634.930 1.197 0.525
Chain 1: 700 -8991.921 1.036 0.373
Chain 1: 800 -8768.517 0.910 0.373
Chain 1: 900 -8686.350 0.810 0.212
Chain 1: 1000 -9702.421 0.739 0.212
Chain 1: 1100 -9416.646 0.642 0.105 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8651.106 0.152 0.088
Chain 1: 1300 -9077.975 0.119 0.077
Chain 1: 1400 -9131.831 0.098 0.072
Chain 1: 1500 -8971.139 0.048 0.047
Chain 1: 1600 -9078.177 0.041 0.030
Chain 1: 1700 -9111.892 0.034 0.025
Chain 1: 1800 -8639.898 0.037 0.030
Chain 1: 1900 -8738.220 0.038 0.030
Chain 1: 2000 -8744.679 0.027 0.018
Chain 1: 2100 -8929.961 0.026 0.018
Chain 1: 2200 -8574.907 0.022 0.018
Chain 1: 2300 -8639.867 0.018 0.012
Chain 1: 2400 -8742.651 0.018 0.012
Chain 1: 2500 -8635.327 0.018 0.012
Chain 1: 2600 -8704.438 0.017 0.012
Chain 1: 2700 -8609.673 0.018 0.012
Chain 1: 2800 -8583.980 0.013 0.011
Chain 1: 2900 -8663.809 0.013 0.011
Chain 1: 3000 -8617.796 0.013 0.011
Chain 1: 3100 -8563.112 0.012 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003383 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.83 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8412139.060 1.000 1.000
Chain 1: 200 -1586739.357 2.651 4.302
Chain 1: 300 -891539.644 2.027 1.000
Chain 1: 400 -458337.180 1.757 1.000
Chain 1: 500 -358797.664 1.461 0.945
Chain 1: 600 -233990.397 1.306 0.945
Chain 1: 700 -120346.775 1.255 0.944
Chain 1: 800 -87597.599 1.144 0.944
Chain 1: 900 -67981.173 1.049 0.780
Chain 1: 1000 -52824.327 0.973 0.780
Chain 1: 1100 -40321.936 0.904 0.533 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39517.903 0.476 0.374
Chain 1: 1300 -27464.447 0.442 0.374
Chain 1: 1400 -27191.739 0.348 0.310
Chain 1: 1500 -23774.151 0.335 0.310
Chain 1: 1600 -22991.738 0.285 0.289
Chain 1: 1700 -21862.368 0.196 0.287
Chain 1: 1800 -21807.001 0.159 0.144
Chain 1: 1900 -22134.838 0.131 0.052
Chain 1: 2000 -20640.958 0.110 0.052
Chain 1: 2100 -20879.997 0.080 0.034
Chain 1: 2200 -21107.698 0.079 0.034
Chain 1: 2300 -20723.293 0.037 0.019
Chain 1: 2400 -20494.695 0.037 0.019
Chain 1: 2500 -20296.602 0.024 0.015
Chain 1: 2600 -19925.223 0.022 0.015
Chain 1: 2700 -19881.687 0.017 0.011
Chain 1: 2800 -19597.721 0.018 0.014
Chain 1: 2900 -19879.748 0.018 0.014
Chain 1: 3000 -19865.934 0.011 0.011
Chain 1: 3100 -19951.171 0.010 0.011
Chain 1: 3200 -19640.717 0.011 0.014
Chain 1: 3300 -19846.304 0.010 0.011
Chain 1: 3400 -19319.184 0.012 0.014
Chain 1: 3500 -19934.125 0.014 0.014
Chain 1: 3600 -19236.743 0.016 0.014
Chain 1: 3700 -19626.550 0.017 0.016
Chain 1: 3800 -18579.963 0.022 0.020
Chain 1: 3900 -18575.874 0.020 0.020
Chain 1: 4000 -18693.243 0.021 0.020
Chain 1: 4100 -18606.676 0.021 0.020
Chain 1: 4200 -18421.516 0.020 0.020
Chain 1: 4300 -18560.942 0.020 0.020
Chain 1: 4400 -18516.628 0.017 0.010
Chain 1: 4500 -18418.900 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001743 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 17.43 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12884.665 1.000 1.000
Chain 1: 200 -9787.319 0.658 1.000
Chain 1: 300 -8555.351 0.487 0.316
Chain 1: 400 -8649.659 0.368 0.316
Chain 1: 500 -8560.529 0.296 0.144
Chain 1: 600 -8446.231 0.249 0.144
Chain 1: 700 -8368.508 0.215 0.014
Chain 1: 800 -8356.162 0.188 0.014
Chain 1: 900 -8414.402 0.168 0.011
Chain 1: 1000 -8391.781 0.152 0.011
Chain 1: 1100 -8491.034 0.053 0.011
Chain 1: 1200 -8407.306 0.022 0.010
Chain 1: 1300 -8326.703 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001416 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.16 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58549.761 1.000 1.000
Chain 1: 200 -18183.561 1.610 2.220
Chain 1: 300 -8887.831 1.422 1.046
Chain 1: 400 -8109.445 1.090 1.046
Chain 1: 500 -8420.062 0.880 1.000
Chain 1: 600 -7961.185 0.743 1.000
Chain 1: 700 -7767.278 0.640 0.096
Chain 1: 800 -7790.778 0.561 0.096
Chain 1: 900 -7995.857 0.501 0.058
Chain 1: 1000 -7859.207 0.453 0.058
Chain 1: 1100 -7748.840 0.354 0.037
Chain 1: 1200 -7736.392 0.132 0.026
Chain 1: 1300 -7665.079 0.029 0.025
Chain 1: 1400 -7919.062 0.022 0.025
Chain 1: 1500 -7577.050 0.023 0.025
Chain 1: 1600 -7744.363 0.019 0.022
Chain 1: 1700 -7597.875 0.019 0.019
Chain 1: 1800 -7582.883 0.019 0.019
Chain 1: 1900 -7580.609 0.016 0.017
Chain 1: 2000 -7645.392 0.015 0.014
Chain 1: 2100 -7556.073 0.015 0.012
Chain 1: 2200 -7776.670 0.018 0.019
Chain 1: 2300 -7586.832 0.019 0.022
Chain 1: 2400 -7710.306 0.018 0.019
Chain 1: 2500 -7589.065 0.015 0.016
Chain 1: 2600 -7521.062 0.014 0.016
Chain 1: 2700 -7508.613 0.012 0.012
Chain 1: 2800 -7522.459 0.012 0.012
Chain 1: 2900 -7384.502 0.014 0.016
Chain 1: 3000 -7525.094 0.015 0.016
Chain 1: 3100 -7529.760 0.014 0.016
Chain 1: 3200 -7742.253 0.013 0.016
Chain 1: 3300 -7445.873 0.015 0.016
Chain 1: 3400 -7697.715 0.017 0.019
Chain 1: 3500 -7438.254 0.019 0.019
Chain 1: 3600 -7502.137 0.018 0.019
Chain 1: 3700 -7455.136 0.019 0.019
Chain 1: 3800 -7441.904 0.019 0.019
Chain 1: 3900 -7410.984 0.017 0.019
Chain 1: 4000 -7404.447 0.016 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.005065 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 50.65 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87148.029 1.000 1.000
Chain 1: 200 -13987.672 3.115 5.230
Chain 1: 300 -10347.615 2.194 1.000
Chain 1: 400 -11319.580 1.667 1.000
Chain 1: 500 -9200.366 1.380 0.352
Chain 1: 600 -8795.724 1.157 0.352
Chain 1: 700 -8963.528 0.995 0.230
Chain 1: 800 -9533.473 0.878 0.230
Chain 1: 900 -9112.199 0.785 0.086
Chain 1: 1000 -9007.437 0.708 0.086
Chain 1: 1100 -9197.492 0.610 0.060 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8706.517 0.093 0.056
Chain 1: 1300 -9044.901 0.061 0.046
Chain 1: 1400 -9053.899 0.053 0.046
Chain 1: 1500 -8909.319 0.031 0.037
Chain 1: 1600 -9025.642 0.028 0.021
Chain 1: 1700 -9104.372 0.027 0.021
Chain 1: 1800 -8690.769 0.026 0.021
Chain 1: 1900 -8786.597 0.022 0.016
Chain 1: 2000 -8760.428 0.021 0.016
Chain 1: 2100 -8883.384 0.021 0.014
Chain 1: 2200 -8702.960 0.017 0.014
Chain 1: 2300 -8781.653 0.014 0.013
Chain 1: 2400 -8851.422 0.015 0.013
Chain 1: 2500 -8797.066 0.014 0.011
Chain 1: 2600 -8796.702 0.013 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.005043 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 50.43 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8418102.804 1.000 1.000
Chain 1: 200 -1588124.082 2.650 4.301
Chain 1: 300 -891252.504 2.028 1.000
Chain 1: 400 -458541.004 1.757 1.000
Chain 1: 500 -358557.909 1.461 0.944
Chain 1: 600 -233375.639 1.307 0.944
Chain 1: 700 -119615.149 1.256 0.944
Chain 1: 800 -86860.275 1.146 0.944
Chain 1: 900 -67209.916 1.051 0.782
Chain 1: 1000 -52018.591 0.975 0.782
Chain 1: 1100 -39513.960 0.907 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38689.887 0.479 0.377
Chain 1: 1300 -26665.583 0.446 0.377
Chain 1: 1400 -26385.534 0.353 0.316
Chain 1: 1500 -22979.080 0.340 0.316
Chain 1: 1600 -22197.389 0.290 0.292
Chain 1: 1700 -21073.443 0.200 0.292
Chain 1: 1800 -21018.120 0.162 0.148
Chain 1: 1900 -21344.248 0.135 0.053
Chain 1: 2000 -19856.825 0.113 0.053
Chain 1: 2100 -20095.004 0.082 0.035
Chain 1: 2200 -20321.389 0.081 0.035
Chain 1: 2300 -19938.648 0.038 0.019
Chain 1: 2400 -19710.766 0.038 0.019
Chain 1: 2500 -19512.799 0.025 0.015
Chain 1: 2600 -19143.005 0.023 0.015
Chain 1: 2700 -19099.938 0.018 0.012
Chain 1: 2800 -18816.862 0.019 0.015
Chain 1: 2900 -19098.060 0.019 0.015
Chain 1: 3000 -19084.182 0.012 0.012
Chain 1: 3100 -19169.221 0.011 0.012
Chain 1: 3200 -18859.909 0.011 0.015
Chain 1: 3300 -19064.610 0.011 0.012
Chain 1: 3400 -18539.587 0.012 0.015
Chain 1: 3500 -19151.393 0.014 0.015
Chain 1: 3600 -18458.098 0.016 0.015
Chain 1: 3700 -18844.897 0.018 0.016
Chain 1: 3800 -17804.687 0.022 0.021
Chain 1: 3900 -17800.825 0.021 0.021
Chain 1: 4000 -17918.129 0.022 0.021
Chain 1: 4100 -17831.930 0.022 0.021
Chain 1: 4200 -17648.146 0.021 0.021
Chain 1: 4300 -17786.553 0.021 0.021
Chain 1: 4400 -17743.388 0.018 0.010
Chain 1: 4500 -17645.907 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001291 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.91 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12703.227 1.000 1.000
Chain 1: 200 -9495.088 0.669 1.000
Chain 1: 300 -8334.653 0.492 0.338
Chain 1: 400 -8516.717 0.375 0.338
Chain 1: 500 -8513.608 0.300 0.139
Chain 1: 600 -8298.639 0.254 0.139
Chain 1: 700 -8208.831 0.219 0.026
Chain 1: 800 -8236.597 0.192 0.026
Chain 1: 900 -8356.841 0.173 0.021
Chain 1: 1000 -8246.141 0.157 0.021
Chain 1: 1100 -8289.907 0.057 0.014
Chain 1: 1200 -8229.497 0.024 0.013
Chain 1: 1300 -8170.058 0.011 0.011
Chain 1: 1400 -8206.166 0.009 0.007 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003052 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.52 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -62235.597 1.000 1.000
Chain 1: 200 -18300.648 1.700 2.401
Chain 1: 300 -9102.806 1.470 1.010
Chain 1: 400 -9587.637 1.115 1.010
Chain 1: 500 -8700.084 0.913 1.000
Chain 1: 600 -8693.470 0.761 1.000
Chain 1: 700 -8927.907 0.656 0.102
Chain 1: 800 -8367.643 0.582 0.102
Chain 1: 900 -8029.164 0.522 0.067
Chain 1: 1000 -8056.327 0.470 0.067
Chain 1: 1100 -7832.990 0.373 0.051
Chain 1: 1200 -7728.018 0.134 0.042
Chain 1: 1300 -7863.797 0.035 0.029
Chain 1: 1400 -7952.570 0.031 0.026
Chain 1: 1500 -7644.588 0.025 0.026
Chain 1: 1600 -7867.958 0.028 0.028
Chain 1: 1700 -7676.105 0.028 0.028
Chain 1: 1800 -7700.687 0.021 0.025
Chain 1: 1900 -7673.070 0.017 0.017
Chain 1: 2000 -7752.874 0.018 0.017
Chain 1: 2100 -7671.600 0.016 0.014
Chain 1: 2200 -7786.123 0.016 0.015
Chain 1: 2300 -7595.078 0.017 0.015
Chain 1: 2400 -7659.449 0.017 0.015
Chain 1: 2500 -7645.870 0.013 0.011
Chain 1: 2600 -7637.509 0.010 0.010
Chain 1: 2700 -7548.001 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003479 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.79 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85581.572 1.000 1.000
Chain 1: 200 -13842.522 3.091 5.183
Chain 1: 300 -10206.231 2.180 1.000
Chain 1: 400 -11083.744 1.654 1.000
Chain 1: 500 -9187.191 1.365 0.356
Chain 1: 600 -8953.409 1.142 0.356
Chain 1: 700 -9146.104 0.982 0.206
Chain 1: 800 -9546.808 0.864 0.206
Chain 1: 900 -9007.207 0.775 0.079
Chain 1: 1000 -8646.888 0.702 0.079
Chain 1: 1100 -9029.905 0.606 0.060 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8650.593 0.092 0.044
Chain 1: 1300 -8903.066 0.059 0.042
Chain 1: 1400 -8911.666 0.051 0.042
Chain 1: 1500 -8763.175 0.032 0.042
Chain 1: 1600 -8877.250 0.031 0.042
Chain 1: 1700 -8957.423 0.030 0.042
Chain 1: 1800 -8540.361 0.030 0.042
Chain 1: 1900 -8638.268 0.026 0.028
Chain 1: 2000 -8612.145 0.022 0.017
Chain 1: 2100 -8735.786 0.019 0.014
Chain 1: 2200 -8551.786 0.017 0.014
Chain 1: 2300 -8632.897 0.015 0.013
Chain 1: 2400 -8702.539 0.016 0.013
Chain 1: 2500 -8648.428 0.014 0.011
Chain 1: 2600 -8648.546 0.013 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003364 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.64 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8399044.124 1.000 1.000
Chain 1: 200 -1583261.152 2.652 4.305
Chain 1: 300 -891370.587 2.027 1.000
Chain 1: 400 -458445.338 1.756 1.000
Chain 1: 500 -358731.859 1.461 0.944
Chain 1: 600 -233804.011 1.306 0.944
Chain 1: 700 -119791.327 1.256 0.944
Chain 1: 800 -86951.527 1.146 0.944
Chain 1: 900 -67252.558 1.051 0.776
Chain 1: 1000 -52020.881 0.975 0.776
Chain 1: 1100 -39469.107 0.907 0.534 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38642.828 0.479 0.378
Chain 1: 1300 -26567.146 0.447 0.378
Chain 1: 1400 -26284.170 0.353 0.318
Chain 1: 1500 -22863.146 0.340 0.318
Chain 1: 1600 -22077.278 0.291 0.293
Chain 1: 1700 -20947.048 0.201 0.293
Chain 1: 1800 -20890.444 0.163 0.150
Chain 1: 1900 -21216.583 0.135 0.054
Chain 1: 2000 -19725.743 0.114 0.054
Chain 1: 2100 -19964.132 0.083 0.036
Chain 1: 2200 -20190.961 0.082 0.036
Chain 1: 2300 -19807.890 0.039 0.019
Chain 1: 2400 -19579.935 0.039 0.019
Chain 1: 2500 -19382.134 0.025 0.015
Chain 1: 2600 -19012.065 0.023 0.015
Chain 1: 2700 -18969.015 0.018 0.012
Chain 1: 2800 -18685.894 0.019 0.015
Chain 1: 2900 -18967.245 0.019 0.015
Chain 1: 3000 -18953.424 0.012 0.012
Chain 1: 3100 -19038.378 0.011 0.012
Chain 1: 3200 -18729.002 0.011 0.015
Chain 1: 3300 -18933.792 0.011 0.012
Chain 1: 3400 -18408.639 0.012 0.015
Chain 1: 3500 -19020.634 0.014 0.015
Chain 1: 3600 -18327.255 0.016 0.015
Chain 1: 3700 -18714.090 0.018 0.017
Chain 1: 3800 -17673.670 0.023 0.021
Chain 1: 3900 -17669.853 0.021 0.021
Chain 1: 4000 -17787.141 0.022 0.021
Chain 1: 4100 -17700.854 0.022 0.021
Chain 1: 4200 -17517.125 0.021 0.021
Chain 1: 4300 -17655.481 0.021 0.021
Chain 1: 4400 -17612.269 0.018 0.010
Chain 1: 4500 -17514.857 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001296 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.96 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48519.467 1.000 1.000
Chain 1: 200 -20789.350 1.167 1.334
Chain 1: 300 -14252.284 0.931 1.000
Chain 1: 400 -16404.771 0.731 1.000
Chain 1: 500 -16962.116 0.591 0.459
Chain 1: 600 -11273.611 0.577 0.505
Chain 1: 700 -14946.979 0.530 0.459
Chain 1: 800 -22402.233 0.505 0.459
Chain 1: 900 -12808.598 0.532 0.459
Chain 1: 1000 -14545.446 0.491 0.459
Chain 1: 1100 -22193.352 0.425 0.345
Chain 1: 1200 -11841.232 0.379 0.345
Chain 1: 1300 -11413.358 0.337 0.333
Chain 1: 1400 -11391.173 0.324 0.333
Chain 1: 1500 -12008.575 0.326 0.333
Chain 1: 1600 -10090.852 0.295 0.246
Chain 1: 1700 -9495.910 0.276 0.190
Chain 1: 1800 -15707.870 0.283 0.190
Chain 1: 1900 -10588.386 0.256 0.190
Chain 1: 2000 -9415.056 0.257 0.190
Chain 1: 2100 -9899.505 0.227 0.125
Chain 1: 2200 -11440.481 0.153 0.125
Chain 1: 2300 -9276.278 0.173 0.135
Chain 1: 2400 -8874.429 0.177 0.135
Chain 1: 2500 -12485.460 0.201 0.190
Chain 1: 2600 -10045.914 0.206 0.233
Chain 1: 2700 -9045.955 0.211 0.233
Chain 1: 2800 -15253.185 0.212 0.233
Chain 1: 2900 -9116.423 0.231 0.233
Chain 1: 3000 -8935.853 0.221 0.233
Chain 1: 3100 -8671.575 0.219 0.233
Chain 1: 3200 -8481.573 0.207 0.233
Chain 1: 3300 -9164.874 0.192 0.111
Chain 1: 3400 -17790.813 0.236 0.243
Chain 1: 3500 -10416.895 0.277 0.243
Chain 1: 3600 -9198.606 0.266 0.132
Chain 1: 3700 -11695.076 0.277 0.213
Chain 1: 3800 -8510.156 0.273 0.213
Chain 1: 3900 -9663.672 0.218 0.132
Chain 1: 4000 -9142.352 0.222 0.132
Chain 1: 4100 -8482.883 0.226 0.132
Chain 1: 4200 -9780.692 0.237 0.133
Chain 1: 4300 -9618.599 0.232 0.133
Chain 1: 4400 -8838.135 0.192 0.132
Chain 1: 4500 -8953.323 0.123 0.119
Chain 1: 4600 -14423.262 0.147 0.119
Chain 1: 4700 -13319.730 0.134 0.088
Chain 1: 4800 -8313.320 0.157 0.088
Chain 1: 4900 -8960.740 0.152 0.083
Chain 1: 5000 -9477.543 0.152 0.083
Chain 1: 5100 -8416.259 0.157 0.088
Chain 1: 5200 -8709.330 0.147 0.083
Chain 1: 5300 -9201.051 0.151 0.083
Chain 1: 5400 -12796.426 0.170 0.083
Chain 1: 5500 -8100.123 0.227 0.126
Chain 1: 5600 -8319.065 0.191 0.083
Chain 1: 5700 -8293.006 0.183 0.072
Chain 1: 5800 -8650.813 0.127 0.055
Chain 1: 5900 -8060.497 0.127 0.055
Chain 1: 6000 -10964.155 0.148 0.073
Chain 1: 6100 -10839.452 0.137 0.053
Chain 1: 6200 -12731.876 0.148 0.073
Chain 1: 6300 -9132.819 0.182 0.149
Chain 1: 6400 -11767.606 0.177 0.149
Chain 1: 6500 -9727.845 0.140 0.149
Chain 1: 6600 -8468.531 0.152 0.149
Chain 1: 6700 -8468.321 0.152 0.149
Chain 1: 6800 -9502.015 0.158 0.149
Chain 1: 6900 -8610.478 0.161 0.149
Chain 1: 7000 -8145.255 0.141 0.149
Chain 1: 7100 -10504.004 0.162 0.149
Chain 1: 7200 -8589.348 0.169 0.210
Chain 1: 7300 -8755.943 0.132 0.149
Chain 1: 7400 -13984.086 0.147 0.149
Chain 1: 7500 -10992.655 0.153 0.149
Chain 1: 7600 -9118.683 0.159 0.206
Chain 1: 7700 -8246.145 0.169 0.206
Chain 1: 7800 -8067.207 0.161 0.206
Chain 1: 7900 -8231.394 0.152 0.206
Chain 1: 8000 -7971.702 0.150 0.206
Chain 1: 8100 -7979.027 0.127 0.106
Chain 1: 8200 -8282.515 0.109 0.037
Chain 1: 8300 -10324.215 0.127 0.106
Chain 1: 8400 -8974.139 0.104 0.106
Chain 1: 8500 -8023.952 0.089 0.106
Chain 1: 8600 -8429.256 0.073 0.048
Chain 1: 8700 -8129.299 0.066 0.037
Chain 1: 8800 -9993.270 0.083 0.048
Chain 1: 8900 -8635.387 0.097 0.118
Chain 1: 9000 -10931.883 0.114 0.150
Chain 1: 9100 -9413.980 0.130 0.157
Chain 1: 9200 -8646.145 0.136 0.157
Chain 1: 9300 -8658.574 0.116 0.150
Chain 1: 9400 -10726.029 0.120 0.157
Chain 1: 9500 -10534.899 0.110 0.157
Chain 1: 9600 -8122.949 0.135 0.161
Chain 1: 9700 -7992.813 0.133 0.161
Chain 1: 9800 -8492.706 0.120 0.157
Chain 1: 9900 -8829.576 0.108 0.089
Chain 1: 10000 -7939.503 0.098 0.089
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001547 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.47 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56516.998 1.000 1.000
Chain 1: 200 -17064.184 1.656 2.312
Chain 1: 300 -8507.974 1.439 1.006
Chain 1: 400 -8296.875 1.086 1.006
Chain 1: 500 -8143.164 0.872 1.000
Chain 1: 600 -8310.625 0.730 1.000
Chain 1: 700 -7693.621 0.637 0.080
Chain 1: 800 -7953.078 0.562 0.080
Chain 1: 900 -7996.395 0.500 0.033
Chain 1: 1000 -7582.583 0.455 0.055
Chain 1: 1100 -7529.534 0.356 0.033
Chain 1: 1200 -7531.313 0.125 0.025
Chain 1: 1300 -7711.105 0.027 0.023
Chain 1: 1400 -7771.362 0.025 0.020
Chain 1: 1500 -7513.347 0.027 0.023
Chain 1: 1600 -7507.442 0.025 0.023
Chain 1: 1700 -7419.142 0.018 0.012
Chain 1: 1800 -7473.555 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003735 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 37.35 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86416.417 1.000 1.000
Chain 1: 200 -13161.129 3.283 5.566
Chain 1: 300 -9578.160 2.313 1.000
Chain 1: 400 -10575.616 1.759 1.000
Chain 1: 500 -8517.173 1.455 0.374
Chain 1: 600 -8409.997 1.215 0.374
Chain 1: 700 -8216.181 1.045 0.242
Chain 1: 800 -8379.379 0.916 0.242
Chain 1: 900 -8423.743 0.815 0.094
Chain 1: 1000 -8181.609 0.737 0.094
Chain 1: 1100 -8401.229 0.639 0.030 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8082.235 0.087 0.030
Chain 1: 1300 -8296.499 0.052 0.026
Chain 1: 1400 -8279.795 0.043 0.026
Chain 1: 1500 -8181.274 0.020 0.024
Chain 1: 1600 -8282.626 0.020 0.024
Chain 1: 1700 -8370.451 0.018 0.019
Chain 1: 1800 -7974.403 0.021 0.026
Chain 1: 1900 -8075.874 0.022 0.026
Chain 1: 2000 -8046.394 0.019 0.013
Chain 1: 2100 -8168.782 0.018 0.013
Chain 1: 2200 -7949.911 0.017 0.013
Chain 1: 2300 -8104.560 0.016 0.013
Chain 1: 2400 -8118.327 0.016 0.013
Chain 1: 2500 -8087.855 0.016 0.013
Chain 1: 2600 -8090.512 0.014 0.013
Chain 1: 2700 -7996.744 0.015 0.013
Chain 1: 2800 -7967.829 0.010 0.012 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003713 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 37.13 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8395349.285 1.000 1.000
Chain 1: 200 -1583814.029 2.650 4.301
Chain 1: 300 -890974.760 2.026 1.000
Chain 1: 400 -457852.414 1.756 1.000
Chain 1: 500 -358251.395 1.460 0.946
Chain 1: 600 -233011.386 1.307 0.946
Chain 1: 700 -119022.461 1.257 0.946
Chain 1: 800 -86198.588 1.147 0.946
Chain 1: 900 -66503.891 1.053 0.778
Chain 1: 1000 -51270.626 0.977 0.778
Chain 1: 1100 -38726.013 0.910 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37895.792 0.482 0.381
Chain 1: 1300 -25835.226 0.451 0.381
Chain 1: 1400 -25551.020 0.357 0.324
Chain 1: 1500 -22134.571 0.345 0.324
Chain 1: 1600 -21349.559 0.295 0.297
Chain 1: 1700 -20221.564 0.204 0.296
Chain 1: 1800 -20165.079 0.167 0.154
Chain 1: 1900 -20490.887 0.139 0.056
Chain 1: 2000 -19001.838 0.117 0.056
Chain 1: 2100 -19240.163 0.086 0.037
Chain 1: 2200 -19466.615 0.085 0.037
Chain 1: 2300 -19083.877 0.040 0.020
Chain 1: 2400 -18856.076 0.040 0.020
Chain 1: 2500 -18658.243 0.026 0.016
Chain 1: 2600 -18288.691 0.024 0.016
Chain 1: 2700 -18245.691 0.019 0.012
Chain 1: 2800 -17962.809 0.020 0.016
Chain 1: 2900 -18243.894 0.020 0.015
Chain 1: 3000 -18230.044 0.012 0.012
Chain 1: 3100 -18315.023 0.011 0.012
Chain 1: 3200 -18005.908 0.012 0.015
Chain 1: 3300 -18210.456 0.011 0.012
Chain 1: 3400 -17685.841 0.013 0.015
Chain 1: 3500 -18297.113 0.015 0.016
Chain 1: 3600 -17604.543 0.017 0.016
Chain 1: 3700 -17990.831 0.019 0.017
Chain 1: 3800 -16951.782 0.023 0.021
Chain 1: 3900 -16947.970 0.022 0.021
Chain 1: 4000 -17065.237 0.023 0.021
Chain 1: 4100 -16979.115 0.023 0.021
Chain 1: 4200 -16795.588 0.022 0.021
Chain 1: 4300 -16933.795 0.022 0.021
Chain 1: 4400 -16890.845 0.019 0.011
Chain 1: 4500 -16793.424 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001215 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.15 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48412.767 1.000 1.000
Chain 1: 200 -20222.068 1.197 1.394
Chain 1: 300 -19757.398 0.806 1.000
Chain 1: 400 -14414.646 0.697 1.000
Chain 1: 500 -18839.171 0.605 0.371
Chain 1: 600 -16099.631 0.532 0.371
Chain 1: 700 -14983.161 0.467 0.235
Chain 1: 800 -17827.431 0.428 0.235
Chain 1: 900 -12277.586 0.431 0.235
Chain 1: 1000 -10123.171 0.409 0.235
Chain 1: 1100 -13736.335 0.336 0.235
Chain 1: 1200 -10624.581 0.225 0.235
Chain 1: 1300 -11348.402 0.229 0.235
Chain 1: 1400 -25524.436 0.248 0.235
Chain 1: 1500 -10540.087 0.367 0.263
Chain 1: 1600 -9819.304 0.357 0.263
Chain 1: 1700 -9570.529 0.352 0.263
Chain 1: 1800 -11561.627 0.353 0.263
Chain 1: 1900 -9338.211 0.332 0.238
Chain 1: 2000 -11962.499 0.333 0.238
Chain 1: 2100 -11003.366 0.315 0.219
Chain 1: 2200 -12338.316 0.297 0.172
Chain 1: 2300 -11534.627 0.297 0.172
Chain 1: 2400 -13182.632 0.254 0.125
Chain 1: 2500 -8790.939 0.162 0.125
Chain 1: 2600 -8807.819 0.155 0.125
Chain 1: 2700 -10759.748 0.170 0.172
Chain 1: 2800 -11508.452 0.160 0.125
Chain 1: 2900 -9665.138 0.155 0.125
Chain 1: 3000 -13094.920 0.159 0.125
Chain 1: 3100 -14388.081 0.159 0.125
Chain 1: 3200 -8564.020 0.217 0.181
Chain 1: 3300 -14991.562 0.252 0.191
Chain 1: 3400 -16841.012 0.251 0.191
Chain 1: 3500 -9034.609 0.287 0.191
Chain 1: 3600 -14989.188 0.327 0.262
Chain 1: 3700 -9055.344 0.374 0.397
Chain 1: 3800 -9020.590 0.368 0.397
Chain 1: 3900 -8980.766 0.350 0.397
Chain 1: 4000 -8833.497 0.325 0.397
Chain 1: 4100 -8759.338 0.317 0.397
Chain 1: 4200 -8866.895 0.250 0.110
Chain 1: 4300 -8923.066 0.208 0.017
Chain 1: 4400 -8673.330 0.200 0.017
Chain 1: 4500 -8632.061 0.114 0.012
Chain 1: 4600 -8274.482 0.078 0.012
Chain 1: 4700 -8321.853 0.013 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001513 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.13 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -55049.429 1.000 1.000
Chain 1: 200 -16885.728 1.630 2.260
Chain 1: 300 -8522.271 1.414 1.000
Chain 1: 400 -8777.344 1.068 1.000
Chain 1: 500 -8514.325 0.860 0.981
Chain 1: 600 -8435.369 0.718 0.981
Chain 1: 700 -7733.252 0.629 0.091
Chain 1: 800 -7982.596 0.554 0.091
Chain 1: 900 -7741.048 0.496 0.031
Chain 1: 1000 -7761.013 0.447 0.031
Chain 1: 1100 -7577.722 0.349 0.031
Chain 1: 1200 -7565.142 0.123 0.031
Chain 1: 1300 -7618.861 0.026 0.029
Chain 1: 1400 -7775.061 0.025 0.024
Chain 1: 1500 -7552.568 0.025 0.024
Chain 1: 1600 -7533.431 0.024 0.024
Chain 1: 1700 -7488.599 0.016 0.020
Chain 1: 1800 -7549.812 0.013 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003597 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.97 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85148.724 1.000 1.000
Chain 1: 200 -13151.247 3.237 5.475
Chain 1: 300 -9575.312 2.283 1.000
Chain 1: 400 -10521.868 1.734 1.000
Chain 1: 500 -8514.165 1.435 0.373
Chain 1: 600 -8094.270 1.204 0.373
Chain 1: 700 -8213.776 1.034 0.236
Chain 1: 800 -8860.584 0.914 0.236
Chain 1: 900 -8385.244 0.819 0.090
Chain 1: 1000 -8221.535 0.739 0.090
Chain 1: 1100 -8472.962 0.642 0.073 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -7991.976 0.101 0.060
Chain 1: 1300 -8315.136 0.067 0.057
Chain 1: 1400 -8298.363 0.058 0.052
Chain 1: 1500 -8180.508 0.036 0.039
Chain 1: 1600 -8288.322 0.032 0.030
Chain 1: 1700 -8366.923 0.032 0.030
Chain 1: 1800 -7966.553 0.029 0.030
Chain 1: 1900 -8068.110 0.025 0.020
Chain 1: 2000 -8038.914 0.023 0.014
Chain 1: 2100 -8159.395 0.022 0.014
Chain 1: 2200 -7936.011 0.019 0.014
Chain 1: 2300 -8097.336 0.017 0.014
Chain 1: 2400 -7979.126 0.018 0.015
Chain 1: 2500 -8043.296 0.017 0.015
Chain 1: 2600 -8064.175 0.016 0.015
Chain 1: 2700 -7983.768 0.016 0.015
Chain 1: 2800 -7958.532 0.012 0.013
Chain 1: 2900 -8013.209 0.011 0.010
Chain 1: 3000 -7898.507 0.012 0.015
Chain 1: 3100 -8035.396 0.013 0.015
Chain 1: 3200 -7915.893 0.011 0.015
Chain 1: 3300 -7936.809 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.006408 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 64.08 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8379663.153 1.000 1.000
Chain 1: 200 -1580777.035 2.650 4.301
Chain 1: 300 -890968.677 2.025 1.000
Chain 1: 400 -457930.890 1.755 1.000
Chain 1: 500 -358760.387 1.459 0.946
Chain 1: 600 -233590.425 1.306 0.946
Chain 1: 700 -119351.614 1.256 0.946
Chain 1: 800 -86451.029 1.146 0.946
Chain 1: 900 -66696.491 1.052 0.774
Chain 1: 1000 -51414.645 0.976 0.774
Chain 1: 1100 -38823.869 0.909 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37985.724 0.481 0.381
Chain 1: 1300 -25878.862 0.450 0.381
Chain 1: 1400 -25589.571 0.357 0.324
Chain 1: 1500 -22161.183 0.345 0.324
Chain 1: 1600 -21372.351 0.295 0.297
Chain 1: 1700 -20238.973 0.205 0.296
Chain 1: 1800 -20181.132 0.167 0.155
Chain 1: 1900 -20506.951 0.139 0.056
Chain 1: 2000 -19014.920 0.117 0.056
Chain 1: 2100 -19253.369 0.086 0.037
Chain 1: 2200 -19480.332 0.085 0.037
Chain 1: 2300 -19097.161 0.040 0.020
Chain 1: 2400 -18869.291 0.040 0.020
Chain 1: 2500 -18671.638 0.026 0.016
Chain 1: 2600 -18301.911 0.024 0.016
Chain 1: 2700 -18258.809 0.019 0.012
Chain 1: 2800 -17976.013 0.020 0.016
Chain 1: 2900 -18257.145 0.020 0.015
Chain 1: 3000 -18243.276 0.012 0.012
Chain 1: 3100 -18328.267 0.011 0.012
Chain 1: 3200 -18019.109 0.012 0.015
Chain 1: 3300 -18223.660 0.011 0.012
Chain 1: 3400 -17699.018 0.013 0.015
Chain 1: 3500 -18310.443 0.015 0.016
Chain 1: 3600 -17617.689 0.017 0.016
Chain 1: 3700 -18004.147 0.019 0.017
Chain 1: 3800 -16964.889 0.023 0.021
Chain 1: 3900 -16961.102 0.022 0.021
Chain 1: 4000 -17078.335 0.023 0.021
Chain 1: 4100 -16992.242 0.023 0.021
Chain 1: 4200 -16808.640 0.022 0.021
Chain 1: 4300 -16946.880 0.022 0.021
Chain 1: 4400 -16903.884 0.019 0.011
Chain 1: 4500 -16806.488 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001942 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 19.42 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49091.618 1.000 1.000
Chain 1: 200 -16215.345 1.514 2.027
Chain 1: 300 -25009.888 1.126 1.000
Chain 1: 400 -15932.735 0.987 1.000
Chain 1: 500 -17083.458 0.803 0.570
Chain 1: 600 -18691.425 0.684 0.570
Chain 1: 700 -13141.172 0.646 0.422
Chain 1: 800 -13275.871 0.567 0.422
Chain 1: 900 -13406.304 0.505 0.352
Chain 1: 1000 -12774.512 0.459 0.352
Chain 1: 1100 -11436.906 0.371 0.117
Chain 1: 1200 -16349.433 0.198 0.117
Chain 1: 1300 -12524.589 0.194 0.117
Chain 1: 1400 -11412.113 0.147 0.097
Chain 1: 1500 -10821.667 0.145 0.097
Chain 1: 1600 -19544.800 0.181 0.117
Chain 1: 1700 -12613.281 0.194 0.117
Chain 1: 1800 -16206.263 0.215 0.222
Chain 1: 1900 -9725.863 0.281 0.300
Chain 1: 2000 -15444.430 0.313 0.305
Chain 1: 2100 -10268.067 0.352 0.370
Chain 1: 2200 -12122.843 0.337 0.370
Chain 1: 2300 -11719.170 0.310 0.370
Chain 1: 2400 -9682.546 0.321 0.370
Chain 1: 2500 -10459.181 0.323 0.370
Chain 1: 2600 -10461.605 0.278 0.222
Chain 1: 2700 -9795.637 0.230 0.210
Chain 1: 2800 -10123.458 0.211 0.153
Chain 1: 2900 -9612.148 0.150 0.074
Chain 1: 3000 -9125.909 0.118 0.068
Chain 1: 3100 -12996.555 0.098 0.068
Chain 1: 3200 -10104.764 0.111 0.068
Chain 1: 3300 -9360.265 0.116 0.074
Chain 1: 3400 -9565.331 0.097 0.068
Chain 1: 3500 -9940.218 0.093 0.053
Chain 1: 3600 -8875.612 0.105 0.068
Chain 1: 3700 -11017.496 0.118 0.080
Chain 1: 3800 -15466.026 0.143 0.120
Chain 1: 3900 -9352.563 0.203 0.194
Chain 1: 4000 -9527.818 0.200 0.194
Chain 1: 4100 -9358.156 0.172 0.120
Chain 1: 4200 -10353.239 0.153 0.096
Chain 1: 4300 -9659.311 0.152 0.096
Chain 1: 4400 -10155.063 0.155 0.096
Chain 1: 4500 -9391.532 0.159 0.096
Chain 1: 4600 -8794.350 0.154 0.081
Chain 1: 4700 -8665.650 0.136 0.072
Chain 1: 4800 -11716.214 0.133 0.072
Chain 1: 4900 -9064.139 0.097 0.072
Chain 1: 5000 -9465.230 0.099 0.072
Chain 1: 5100 -8836.055 0.105 0.072
Chain 1: 5200 -8851.702 0.095 0.071
Chain 1: 5300 -10260.307 0.102 0.071
Chain 1: 5400 -8834.274 0.113 0.081
Chain 1: 5500 -12366.926 0.134 0.137
Chain 1: 5600 -8504.610 0.172 0.161
Chain 1: 5700 -13571.935 0.208 0.260
Chain 1: 5800 -8622.246 0.239 0.286
Chain 1: 5900 -14661.272 0.251 0.286
Chain 1: 6000 -9724.725 0.298 0.373
Chain 1: 6100 -10101.473 0.294 0.373
Chain 1: 6200 -8468.086 0.314 0.373
Chain 1: 6300 -8747.974 0.303 0.373
Chain 1: 6400 -10574.987 0.304 0.373
Chain 1: 6500 -8447.520 0.301 0.373
Chain 1: 6600 -11857.421 0.284 0.288
Chain 1: 6700 -8545.902 0.286 0.288
Chain 1: 6800 -12858.728 0.262 0.288
Chain 1: 6900 -9633.450 0.254 0.288
Chain 1: 7000 -13178.087 0.230 0.269
Chain 1: 7100 -10882.011 0.247 0.269
Chain 1: 7200 -8840.935 0.251 0.269
Chain 1: 7300 -10677.699 0.265 0.269
Chain 1: 7400 -8972.909 0.267 0.269
Chain 1: 7500 -8695.196 0.245 0.269
Chain 1: 7600 -9829.553 0.228 0.231
Chain 1: 7700 -8789.450 0.201 0.211
Chain 1: 7800 -8731.505 0.168 0.190
Chain 1: 7900 -8518.582 0.137 0.172
Chain 1: 8000 -10833.123 0.131 0.172
Chain 1: 8100 -9095.559 0.129 0.172
Chain 1: 8200 -8658.511 0.111 0.118
Chain 1: 8300 -13003.248 0.128 0.118
Chain 1: 8400 -8559.739 0.161 0.118
Chain 1: 8500 -8523.849 0.158 0.118
Chain 1: 8600 -9700.760 0.158 0.121
Chain 1: 8700 -8641.780 0.159 0.123
Chain 1: 8800 -8349.691 0.162 0.123
Chain 1: 8900 -10226.098 0.177 0.183
Chain 1: 9000 -10576.457 0.159 0.123
Chain 1: 9100 -9096.840 0.157 0.123
Chain 1: 9200 -9457.252 0.155 0.123
Chain 1: 9300 -8299.820 0.136 0.123
Chain 1: 9400 -9906.419 0.100 0.123
Chain 1: 9500 -8819.314 0.112 0.123
Chain 1: 9600 -9446.059 0.107 0.123
Chain 1: 9700 -8642.867 0.104 0.123
Chain 1: 9800 -10117.466 0.115 0.139
Chain 1: 9900 -9358.792 0.104 0.123
Chain 1: 10000 -8935.760 0.106 0.123
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001466 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.66 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -62115.965 1.000 1.000
Chain 1: 200 -17931.434 1.732 2.464
Chain 1: 300 -8943.137 1.490 1.005
Chain 1: 400 -8330.034 1.136 1.005
Chain 1: 500 -8494.893 0.912 1.000
Chain 1: 600 -8897.771 0.768 1.000
Chain 1: 700 -7922.575 0.676 0.123
Chain 1: 800 -8127.423 0.594 0.123
Chain 1: 900 -7956.662 0.531 0.074
Chain 1: 1000 -7885.017 0.479 0.074
Chain 1: 1100 -7874.057 0.379 0.045
Chain 1: 1200 -7675.246 0.135 0.026
Chain 1: 1300 -7809.358 0.036 0.025
Chain 1: 1400 -7922.473 0.030 0.021
Chain 1: 1500 -7672.035 0.032 0.025
Chain 1: 1600 -7871.124 0.030 0.025
Chain 1: 1700 -7584.004 0.021 0.025
Chain 1: 1800 -7713.107 0.020 0.021
Chain 1: 1900 -7633.849 0.019 0.017
Chain 1: 2000 -7690.901 0.019 0.017
Chain 1: 2100 -7673.972 0.019 0.017
Chain 1: 2200 -7782.482 0.018 0.017
Chain 1: 2300 -7659.087 0.018 0.016
Chain 1: 2400 -7723.328 0.017 0.016
Chain 1: 2500 -7645.896 0.015 0.014
Chain 1: 2600 -7590.118 0.013 0.010
Chain 1: 2700 -7636.098 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.005367 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 53.67 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86215.138 1.000 1.000
Chain 1: 200 -13576.767 3.175 5.350
Chain 1: 300 -9995.300 2.236 1.000
Chain 1: 400 -10874.041 1.697 1.000
Chain 1: 500 -8949.912 1.401 0.358
Chain 1: 600 -8545.927 1.175 0.358
Chain 1: 700 -8537.358 1.008 0.215
Chain 1: 800 -9276.343 0.892 0.215
Chain 1: 900 -8814.022 0.798 0.081
Chain 1: 1000 -8580.526 0.721 0.081
Chain 1: 1100 -8742.908 0.623 0.080 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8505.539 0.091 0.052
Chain 1: 1300 -8710.074 0.057 0.047
Chain 1: 1400 -8706.062 0.049 0.028
Chain 1: 1500 -8603.529 0.029 0.027
Chain 1: 1600 -8705.467 0.025 0.023
Chain 1: 1700 -8793.273 0.026 0.023
Chain 1: 1800 -8395.185 0.023 0.023
Chain 1: 1900 -8496.156 0.019 0.019
Chain 1: 2000 -8466.974 0.017 0.012
Chain 1: 2100 -8588.063 0.016 0.012
Chain 1: 2200 -8366.164 0.016 0.012
Chain 1: 2300 -8525.013 0.016 0.012
Chain 1: 2400 -8536.998 0.016 0.012
Chain 1: 2500 -8508.927 0.015 0.012
Chain 1: 2600 -8511.483 0.014 0.012
Chain 1: 2700 -8417.650 0.014 0.012
Chain 1: 2800 -8387.996 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003389 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.89 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8394683.390 1.000 1.000
Chain 1: 200 -1585072.613 2.648 4.296
Chain 1: 300 -891714.861 2.025 1.000
Chain 1: 400 -458251.648 1.755 1.000
Chain 1: 500 -358571.491 1.460 0.946
Chain 1: 600 -233399.353 1.306 0.946
Chain 1: 700 -119444.936 1.255 0.946
Chain 1: 800 -86623.513 1.146 0.946
Chain 1: 900 -66930.059 1.051 0.778
Chain 1: 1000 -51698.792 0.976 0.778
Chain 1: 1100 -39153.966 0.908 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38325.437 0.480 0.379
Chain 1: 1300 -26265.150 0.448 0.379
Chain 1: 1400 -25981.584 0.355 0.320
Chain 1: 1500 -22565.053 0.342 0.320
Chain 1: 1600 -21780.124 0.292 0.295
Chain 1: 1700 -20652.111 0.202 0.294
Chain 1: 1800 -20595.802 0.165 0.151
Chain 1: 1900 -20921.567 0.137 0.055
Chain 1: 2000 -19432.578 0.115 0.055
Chain 1: 2100 -19670.918 0.084 0.036
Chain 1: 2200 -19897.298 0.083 0.036
Chain 1: 2300 -19514.660 0.039 0.020
Chain 1: 2400 -19286.849 0.039 0.020
Chain 1: 2500 -19089.008 0.025 0.016
Chain 1: 2600 -18719.441 0.023 0.016
Chain 1: 2700 -18676.469 0.018 0.012
Chain 1: 2800 -18393.513 0.019 0.015
Chain 1: 2900 -18674.670 0.019 0.015
Chain 1: 3000 -18660.819 0.012 0.012
Chain 1: 3100 -18745.775 0.011 0.012
Chain 1: 3200 -18436.660 0.012 0.015
Chain 1: 3300 -18641.234 0.011 0.012
Chain 1: 3400 -18116.579 0.012 0.015
Chain 1: 3500 -18727.858 0.015 0.015
Chain 1: 3600 -18035.317 0.017 0.015
Chain 1: 3700 -18421.556 0.018 0.017
Chain 1: 3800 -17382.522 0.023 0.021
Chain 1: 3900 -17378.714 0.021 0.021
Chain 1: 4000 -17495.994 0.022 0.021
Chain 1: 4100 -17409.823 0.022 0.021
Chain 1: 4200 -17226.343 0.021 0.021
Chain 1: 4300 -17364.522 0.021 0.021
Chain 1: 4400 -17321.565 0.018 0.011
Chain 1: 4500 -17224.164 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001743 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 17.43 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -11811.467 1.000 1.000
Chain 1: 200 -8766.864 0.674 1.000
Chain 1: 300 -7814.559 0.490 0.347
Chain 1: 400 -7909.965 0.370 0.347
Chain 1: 500 -7790.406 0.299 0.122
Chain 1: 600 -7650.814 0.252 0.122
Chain 1: 700 -7601.324 0.217 0.018
Chain 1: 800 -7608.463 0.190 0.018
Chain 1: 900 -7542.902 0.170 0.015
Chain 1: 1000 -7639.500 0.154 0.015
Chain 1: 1100 -7693.095 0.055 0.013
Chain 1: 1200 -7611.316 0.021 0.012
Chain 1: 1300 -7575.688 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001409 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.09 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56417.151 1.000 1.000
Chain 1: 200 -16777.570 1.681 2.363
Chain 1: 300 -8444.939 1.450 1.000
Chain 1: 400 -8545.829 1.090 1.000
Chain 1: 500 -8049.936 0.885 0.987
Chain 1: 600 -8519.707 0.746 0.987
Chain 1: 700 -7880.297 0.651 0.081
Chain 1: 800 -8030.574 0.572 0.081
Chain 1: 900 -7775.515 0.512 0.062
Chain 1: 1000 -7713.661 0.462 0.062
Chain 1: 1100 -7682.951 0.362 0.055
Chain 1: 1200 -7611.443 0.127 0.033
Chain 1: 1300 -7577.716 0.029 0.019
Chain 1: 1400 -7745.677 0.030 0.022
Chain 1: 1500 -7573.051 0.026 0.022
Chain 1: 1600 -7470.079 0.022 0.019
Chain 1: 1700 -7455.509 0.014 0.014
Chain 1: 1800 -7489.623 0.012 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003611 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 36.11 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85749.051 1.000 1.000
Chain 1: 200 -12865.895 3.332 5.665
Chain 1: 300 -9398.307 2.345 1.000
Chain 1: 400 -10049.508 1.775 1.000
Chain 1: 500 -8264.309 1.463 0.369
Chain 1: 600 -8461.083 1.223 0.369
Chain 1: 700 -8241.465 1.052 0.216
Chain 1: 800 -8488.969 0.924 0.216
Chain 1: 900 -8286.301 0.824 0.065
Chain 1: 1000 -8071.996 0.744 0.065
Chain 1: 1100 -8328.910 0.648 0.031 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8114.898 0.084 0.029
Chain 1: 1300 -8039.303 0.048 0.027
Chain 1: 1400 -8052.853 0.041 0.027
Chain 1: 1500 -8071.752 0.020 0.026
Chain 1: 1600 -8068.939 0.018 0.026
Chain 1: 1700 -8017.895 0.016 0.024
Chain 1: 1800 -7895.764 0.014 0.015
Chain 1: 1900 -8006.436 0.013 0.014
Chain 1: 2000 -7971.231 0.011 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00322 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.2 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8411705.078 1.000 1.000
Chain 1: 200 -1588321.683 2.648 4.296
Chain 1: 300 -891400.148 2.026 1.000
Chain 1: 400 -457343.007 1.757 1.000
Chain 1: 500 -357170.974 1.461 0.949
Chain 1: 600 -232035.206 1.308 0.949
Chain 1: 700 -118380.437 1.258 0.949
Chain 1: 800 -85628.314 1.149 0.949
Chain 1: 900 -66000.268 1.054 0.782
Chain 1: 1000 -50811.187 0.979 0.782
Chain 1: 1100 -38313.771 0.911 0.539 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37484.996 0.484 0.382
Chain 1: 1300 -25484.472 0.453 0.382
Chain 1: 1400 -25203.464 0.359 0.326
Chain 1: 1500 -21802.685 0.346 0.326
Chain 1: 1600 -21021.657 0.296 0.299
Chain 1: 1700 -19901.500 0.206 0.297
Chain 1: 1800 -19846.512 0.168 0.156
Chain 1: 1900 -20171.654 0.140 0.056
Chain 1: 2000 -18687.967 0.118 0.056
Chain 1: 2100 -18925.983 0.086 0.037
Chain 1: 2200 -19151.271 0.085 0.037
Chain 1: 2300 -18769.764 0.040 0.020
Chain 1: 2400 -18542.253 0.040 0.020
Chain 1: 2500 -18344.179 0.026 0.016
Chain 1: 2600 -17975.499 0.024 0.016
Chain 1: 2700 -17932.821 0.019 0.013
Chain 1: 2800 -17650.049 0.020 0.016
Chain 1: 2900 -17930.799 0.020 0.016
Chain 1: 3000 -17917.088 0.012 0.013
Chain 1: 3100 -18001.916 0.012 0.012
Chain 1: 3200 -17693.297 0.012 0.016
Chain 1: 3300 -17897.494 0.011 0.012
Chain 1: 3400 -17373.613 0.013 0.016
Chain 1: 3500 -17983.629 0.015 0.016
Chain 1: 3600 -17292.740 0.017 0.016
Chain 1: 3700 -17677.677 0.019 0.017
Chain 1: 3800 -16641.162 0.024 0.022
Chain 1: 3900 -16637.393 0.022 0.022
Chain 1: 4000 -16754.699 0.023 0.022
Chain 1: 4100 -16668.617 0.023 0.022
Chain 1: 4200 -16485.717 0.022 0.022
Chain 1: 4300 -16623.523 0.022 0.022
Chain 1: 4400 -16581.011 0.019 0.011
Chain 1: 4500 -16483.665 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001523 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.23 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49072.775 1.000 1.000
Chain 1: 200 -13412.072 1.829 2.659
Chain 1: 300 -20130.908 1.331 1.000
Chain 1: 400 -20482.895 1.002 1.000
Chain 1: 500 -18393.970 0.825 0.334
Chain 1: 600 -15802.077 0.715 0.334
Chain 1: 700 -15883.862 0.613 0.164
Chain 1: 800 -14772.318 0.546 0.164
Chain 1: 900 -12763.769 0.503 0.157
Chain 1: 1000 -23041.048 0.497 0.164
Chain 1: 1100 -30424.074 0.421 0.164
Chain 1: 1200 -12734.875 0.294 0.164
Chain 1: 1300 -11663.714 0.270 0.157
Chain 1: 1400 -11672.008 0.269 0.157
Chain 1: 1500 -10328.483 0.270 0.157
Chain 1: 1600 -10052.597 0.257 0.130
Chain 1: 1700 -11394.826 0.268 0.130
Chain 1: 1800 -11022.156 0.264 0.130
Chain 1: 1900 -17083.416 0.283 0.130
Chain 1: 2000 -11599.614 0.286 0.130
Chain 1: 2100 -9495.887 0.284 0.130
Chain 1: 2200 -9951.821 0.150 0.118
Chain 1: 2300 -10626.244 0.147 0.118
Chain 1: 2400 -9160.742 0.163 0.130
Chain 1: 2500 -9732.192 0.156 0.118
Chain 1: 2600 -16446.358 0.194 0.160
Chain 1: 2700 -9313.189 0.259 0.222
Chain 1: 2800 -10671.358 0.268 0.222
Chain 1: 2900 -9485.712 0.245 0.160
Chain 1: 3000 -10009.992 0.203 0.127
Chain 1: 3100 -8785.453 0.195 0.127
Chain 1: 3200 -9506.088 0.198 0.127
Chain 1: 3300 -9756.999 0.194 0.127
Chain 1: 3400 -10127.575 0.182 0.125
Chain 1: 3500 -14094.328 0.204 0.127
Chain 1: 3600 -11045.156 0.191 0.127
Chain 1: 3700 -9226.895 0.134 0.127
Chain 1: 3800 -10790.313 0.135 0.139
Chain 1: 3900 -16329.653 0.157 0.145
Chain 1: 4000 -9824.371 0.218 0.197
Chain 1: 4100 -9562.207 0.207 0.197
Chain 1: 4200 -9396.545 0.201 0.197
Chain 1: 4300 -12251.617 0.222 0.233
Chain 1: 4400 -10238.668 0.238 0.233
Chain 1: 4500 -9319.168 0.219 0.197
Chain 1: 4600 -11255.778 0.209 0.197
Chain 1: 4700 -10870.182 0.193 0.172
Chain 1: 4800 -8735.786 0.203 0.197
Chain 1: 4900 -10141.868 0.183 0.172
Chain 1: 5000 -13494.879 0.141 0.172
Chain 1: 5100 -8681.926 0.194 0.197
Chain 1: 5200 -11874.908 0.219 0.233
Chain 1: 5300 -10087.530 0.213 0.197
Chain 1: 5400 -8620.871 0.211 0.177
Chain 1: 5500 -9270.652 0.208 0.177
Chain 1: 5600 -14256.927 0.226 0.244
Chain 1: 5700 -15335.942 0.229 0.244
Chain 1: 5800 -16617.204 0.212 0.177
Chain 1: 5900 -10606.188 0.255 0.248
Chain 1: 6000 -9002.810 0.248 0.178
Chain 1: 6100 -9023.432 0.193 0.177
Chain 1: 6200 -8549.579 0.172 0.170
Chain 1: 6300 -9242.960 0.161 0.077
Chain 1: 6400 -13289.113 0.175 0.077
Chain 1: 6500 -9554.085 0.207 0.178
Chain 1: 6600 -8578.950 0.183 0.114
Chain 1: 6700 -8672.178 0.177 0.114
Chain 1: 6800 -8777.596 0.171 0.114
Chain 1: 6900 -13534.585 0.149 0.114
Chain 1: 7000 -9608.680 0.172 0.114
Chain 1: 7100 -9936.232 0.176 0.114
Chain 1: 7200 -8404.959 0.188 0.182
Chain 1: 7300 -9032.571 0.188 0.182
Chain 1: 7400 -8363.519 0.165 0.114
Chain 1: 7500 -8412.681 0.127 0.080
Chain 1: 7600 -10719.277 0.137 0.080
Chain 1: 7700 -9550.564 0.148 0.122
Chain 1: 7800 -14084.957 0.179 0.182
Chain 1: 7900 -8430.460 0.211 0.182
Chain 1: 8000 -8606.248 0.172 0.122
Chain 1: 8100 -8507.846 0.170 0.122
Chain 1: 8200 -8983.675 0.157 0.080
Chain 1: 8300 -8418.000 0.157 0.080
Chain 1: 8400 -8364.076 0.149 0.067
Chain 1: 8500 -10611.615 0.170 0.122
Chain 1: 8600 -8725.550 0.170 0.122
Chain 1: 8700 -8403.946 0.162 0.067
Chain 1: 8800 -8515.347 0.131 0.053
Chain 1: 8900 -10608.949 0.084 0.053
Chain 1: 9000 -10107.432 0.086 0.053
Chain 1: 9100 -8893.261 0.099 0.067
Chain 1: 9200 -8333.455 0.100 0.067
Chain 1: 9300 -10011.666 0.110 0.137
Chain 1: 9400 -10305.951 0.113 0.137
Chain 1: 9500 -11142.868 0.099 0.075
Chain 1: 9600 -8483.086 0.109 0.075
Chain 1: 9700 -8539.557 0.106 0.075
Chain 1: 9800 -12277.807 0.135 0.137
Chain 1: 9900 -9112.113 0.150 0.137
Chain 1: 10000 -9837.355 0.152 0.137
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002309 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 23.09 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58388.160 1.000 1.000
Chain 1: 200 -17728.011 1.647 2.294
Chain 1: 300 -8675.487 1.446 1.043
Chain 1: 400 -8170.241 1.100 1.043
Chain 1: 500 -8470.459 0.887 1.000
Chain 1: 600 -8665.174 0.743 1.000
Chain 1: 700 -8619.090 0.637 0.062
Chain 1: 800 -8193.771 0.564 0.062
Chain 1: 900 -7651.272 0.509 0.062
Chain 1: 1000 -7782.214 0.460 0.062
Chain 1: 1100 -7658.341 0.362 0.052
Chain 1: 1200 -7733.061 0.133 0.035
Chain 1: 1300 -7733.255 0.029 0.022
Chain 1: 1400 -7798.120 0.024 0.017
Chain 1: 1500 -7532.861 0.024 0.017
Chain 1: 1600 -7710.721 0.024 0.017
Chain 1: 1700 -7440.524 0.027 0.023
Chain 1: 1800 -7590.717 0.024 0.020
Chain 1: 1900 -7537.038 0.017 0.017
Chain 1: 2000 -7624.643 0.017 0.016
Chain 1: 2100 -7533.816 0.016 0.012
Chain 1: 2200 -7657.780 0.017 0.016
Chain 1: 2300 -7515.875 0.019 0.019
Chain 1: 2400 -7619.977 0.019 0.019
Chain 1: 2500 -7582.842 0.016 0.016
Chain 1: 2600 -7487.231 0.015 0.014
Chain 1: 2700 -7480.347 0.012 0.013
Chain 1: 2800 -7464.836 0.010 0.012
Chain 1: 2900 -7345.950 0.011 0.013
Chain 1: 3000 -7485.918 0.012 0.014
Chain 1: 3100 -7477.306 0.011 0.014
Chain 1: 3200 -7674.381 0.011 0.014
Chain 1: 3300 -7406.888 0.013 0.014
Chain 1: 3400 -7615.326 0.015 0.016
Chain 1: 3500 -7387.834 0.017 0.019
Chain 1: 3600 -7452.178 0.017 0.019
Chain 1: 3700 -7401.373 0.017 0.019
Chain 1: 3800 -7404.765 0.017 0.019
Chain 1: 3900 -7370.370 0.016 0.019
Chain 1: 4000 -7365.259 0.014 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00271 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 27.1 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86521.232 1.000 1.000
Chain 1: 200 -13640.949 3.171 5.343
Chain 1: 300 -10015.576 2.235 1.000
Chain 1: 400 -10836.418 1.695 1.000
Chain 1: 500 -8979.183 1.397 0.362
Chain 1: 600 -8498.118 1.174 0.362
Chain 1: 700 -8502.760 1.006 0.207
Chain 1: 800 -9330.621 0.892 0.207
Chain 1: 900 -8805.269 0.799 0.089
Chain 1: 1000 -8626.459 0.721 0.089
Chain 1: 1100 -8897.341 0.624 0.076 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8628.803 0.093 0.060
Chain 1: 1300 -8725.791 0.058 0.057
Chain 1: 1400 -8748.386 0.051 0.031
Chain 1: 1500 -8576.606 0.032 0.030
Chain 1: 1600 -8696.064 0.028 0.021
Chain 1: 1700 -8780.258 0.029 0.021
Chain 1: 1800 -8370.116 0.025 0.021
Chain 1: 1900 -8465.883 0.020 0.020
Chain 1: 2000 -8438.705 0.018 0.014
Chain 1: 2100 -8560.249 0.017 0.014
Chain 1: 2200 -8407.138 0.015 0.014
Chain 1: 2300 -8464.001 0.015 0.014
Chain 1: 2400 -8530.757 0.015 0.014
Chain 1: 2500 -8476.460 0.014 0.011
Chain 1: 2600 -8474.721 0.013 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003714 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 37.14 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8392881.448 1.000 1.000
Chain 1: 200 -1583906.877 2.649 4.299
Chain 1: 300 -890803.885 2.026 1.000
Chain 1: 400 -457802.494 1.756 1.000
Chain 1: 500 -358264.550 1.460 0.946
Chain 1: 600 -233224.321 1.306 0.946
Chain 1: 700 -119392.448 1.256 0.946
Chain 1: 800 -86598.417 1.146 0.946
Chain 1: 900 -66932.858 1.051 0.778
Chain 1: 1000 -51723.214 0.976 0.778
Chain 1: 1100 -39194.731 0.908 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38370.321 0.480 0.379
Chain 1: 1300 -26322.632 0.448 0.379
Chain 1: 1400 -26041.029 0.354 0.320
Chain 1: 1500 -22627.571 0.342 0.320
Chain 1: 1600 -21843.719 0.292 0.294
Chain 1: 1700 -20717.103 0.202 0.294
Chain 1: 1800 -20661.156 0.164 0.151
Chain 1: 1900 -20987.216 0.136 0.054
Chain 1: 2000 -19498.371 0.115 0.054
Chain 1: 2100 -19736.740 0.084 0.036
Chain 1: 2200 -19963.193 0.083 0.036
Chain 1: 2300 -19580.395 0.039 0.020
Chain 1: 2400 -19352.539 0.039 0.020
Chain 1: 2500 -19154.585 0.025 0.016
Chain 1: 2600 -18784.898 0.023 0.016
Chain 1: 2700 -18741.852 0.018 0.012
Chain 1: 2800 -18458.807 0.019 0.015
Chain 1: 2900 -18739.997 0.019 0.015
Chain 1: 3000 -18726.175 0.012 0.012
Chain 1: 3100 -18811.169 0.011 0.012
Chain 1: 3200 -18501.906 0.012 0.015
Chain 1: 3300 -18706.567 0.011 0.012
Chain 1: 3400 -18181.649 0.012 0.015
Chain 1: 3500 -18793.313 0.015 0.015
Chain 1: 3600 -18100.239 0.017 0.015
Chain 1: 3700 -18486.901 0.018 0.017
Chain 1: 3800 -17446.998 0.023 0.021
Chain 1: 3900 -17443.145 0.021 0.021
Chain 1: 4000 -17560.442 0.022 0.021
Chain 1: 4100 -17474.264 0.022 0.021
Chain 1: 4200 -17290.560 0.021 0.021
Chain 1: 4300 -17428.904 0.021 0.021
Chain 1: 4400 -17385.804 0.018 0.011
Chain 1: 4500 -17288.356 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001471 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.71 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49741.377 1.000 1.000
Chain 1: 200 -16793.721 1.481 1.962
Chain 1: 300 -19214.077 1.029 1.000
Chain 1: 400 -19597.356 0.777 1.000
Chain 1: 500 -12585.765 0.733 0.557
Chain 1: 600 -17150.583 0.655 0.557
Chain 1: 700 -15238.980 0.579 0.266
Chain 1: 800 -15616.976 0.510 0.266
Chain 1: 900 -11199.567 0.497 0.266
Chain 1: 1000 -11125.806 0.448 0.266
Chain 1: 1100 -10937.304 0.350 0.126
Chain 1: 1200 -13723.054 0.174 0.126
Chain 1: 1300 -12152.027 0.174 0.129
Chain 1: 1400 -16150.943 0.197 0.203
Chain 1: 1500 -10838.546 0.190 0.203
Chain 1: 1600 -25997.583 0.222 0.203
Chain 1: 1700 -12261.515 0.322 0.248
Chain 1: 1800 -10553.907 0.335 0.248
Chain 1: 1900 -11256.488 0.302 0.203
Chain 1: 2000 -12187.917 0.309 0.203
Chain 1: 2100 -11117.243 0.317 0.203
Chain 1: 2200 -19736.678 0.340 0.248
Chain 1: 2300 -9519.698 0.435 0.437
Chain 1: 2400 -9392.252 0.411 0.437
Chain 1: 2500 -9590.515 0.364 0.162
Chain 1: 2600 -11138.887 0.320 0.139
Chain 1: 2700 -9674.148 0.223 0.139
Chain 1: 2800 -10615.752 0.216 0.096
Chain 1: 2900 -11190.027 0.215 0.096
Chain 1: 3000 -9445.201 0.226 0.139
Chain 1: 3100 -9958.024 0.221 0.139
Chain 1: 3200 -15458.173 0.213 0.139
Chain 1: 3300 -9911.255 0.162 0.139
Chain 1: 3400 -10380.294 0.165 0.139
Chain 1: 3500 -14010.201 0.189 0.151
Chain 1: 3600 -10823.549 0.204 0.185
Chain 1: 3700 -11633.411 0.196 0.185
Chain 1: 3800 -13682.675 0.202 0.185
Chain 1: 3900 -14647.732 0.204 0.185
Chain 1: 4000 -13738.760 0.192 0.150
Chain 1: 4100 -9198.218 0.236 0.259
Chain 1: 4200 -11637.756 0.221 0.210
Chain 1: 4300 -13011.525 0.176 0.150
Chain 1: 4400 -8970.829 0.216 0.210
Chain 1: 4500 -9912.145 0.200 0.150
Chain 1: 4600 -13091.092 0.195 0.150
Chain 1: 4700 -12288.012 0.194 0.150
Chain 1: 4800 -10173.981 0.200 0.208
Chain 1: 4900 -8659.249 0.211 0.208
Chain 1: 5000 -9750.535 0.216 0.208
Chain 1: 5100 -8835.382 0.177 0.175
Chain 1: 5200 -11182.241 0.177 0.175
Chain 1: 5300 -9216.354 0.187 0.208
Chain 1: 5400 -12275.294 0.167 0.208
Chain 1: 5500 -10100.409 0.179 0.210
Chain 1: 5600 -9165.825 0.165 0.208
Chain 1: 5700 -8578.203 0.166 0.208
Chain 1: 5800 -9139.429 0.151 0.175
Chain 1: 5900 -12077.152 0.158 0.210
Chain 1: 6000 -9764.689 0.170 0.213
Chain 1: 6100 -8899.481 0.170 0.213
Chain 1: 6200 -9593.435 0.156 0.213
Chain 1: 6300 -9472.323 0.136 0.102
Chain 1: 6400 -8456.705 0.123 0.102
Chain 1: 6500 -9482.005 0.112 0.102
Chain 1: 6600 -8775.448 0.110 0.097
Chain 1: 6700 -9234.509 0.108 0.097
Chain 1: 6800 -8756.583 0.108 0.097
Chain 1: 6900 -9505.836 0.091 0.081
Chain 1: 7000 -14782.157 0.103 0.081
Chain 1: 7100 -8517.833 0.167 0.081
Chain 1: 7200 -8956.060 0.165 0.081
Chain 1: 7300 -12015.008 0.189 0.108
Chain 1: 7400 -8407.996 0.220 0.108
Chain 1: 7500 -11689.317 0.237 0.255
Chain 1: 7600 -12820.906 0.238 0.255
Chain 1: 7700 -8657.660 0.281 0.281
Chain 1: 7800 -8582.626 0.276 0.281
Chain 1: 7900 -8668.260 0.269 0.281
Chain 1: 8000 -11454.792 0.258 0.255
Chain 1: 8100 -12103.716 0.190 0.243
Chain 1: 8200 -8481.756 0.228 0.255
Chain 1: 8300 -8627.985 0.204 0.243
Chain 1: 8400 -10778.141 0.181 0.199
Chain 1: 8500 -8434.986 0.181 0.199
Chain 1: 8600 -8454.528 0.172 0.199
Chain 1: 8700 -8924.371 0.129 0.054
Chain 1: 8800 -12204.963 0.155 0.199
Chain 1: 8900 -11640.256 0.159 0.199
Chain 1: 9000 -11156.969 0.139 0.054
Chain 1: 9100 -9501.540 0.151 0.174
Chain 1: 9200 -8736.029 0.117 0.088
Chain 1: 9300 -8599.078 0.117 0.088
Chain 1: 9400 -8629.200 0.097 0.053
Chain 1: 9500 -9418.847 0.078 0.053
Chain 1: 9600 -11019.868 0.092 0.084
Chain 1: 9700 -8402.738 0.118 0.088
Chain 1: 9800 -9027.398 0.098 0.084
Chain 1: 9900 -9397.289 0.097 0.084
Chain 1: 10000 -8195.799 0.108 0.088
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00323 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.3 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57112.925 1.000 1.000
Chain 1: 200 -17852.221 1.600 2.199
Chain 1: 300 -8829.540 1.407 1.022
Chain 1: 400 -7980.290 1.082 1.022
Chain 1: 500 -8719.293 0.882 1.000
Chain 1: 600 -8503.512 0.740 1.000
Chain 1: 700 -7966.465 0.644 0.106
Chain 1: 800 -8135.131 0.566 0.106
Chain 1: 900 -8016.884 0.505 0.085
Chain 1: 1000 -7830.122 0.456 0.085
Chain 1: 1100 -7881.018 0.357 0.067
Chain 1: 1200 -7792.202 0.138 0.025
Chain 1: 1300 -7897.725 0.037 0.024
Chain 1: 1400 -7929.284 0.027 0.021
Chain 1: 1500 -7594.788 0.023 0.021
Chain 1: 1600 -7781.543 0.023 0.021
Chain 1: 1700 -7505.006 0.020 0.021
Chain 1: 1800 -7634.548 0.020 0.017
Chain 1: 1900 -7588.850 0.019 0.017
Chain 1: 2000 -7686.167 0.018 0.013
Chain 1: 2100 -7600.306 0.018 0.013
Chain 1: 2200 -7755.090 0.019 0.017
Chain 1: 2300 -7594.137 0.020 0.020
Chain 1: 2400 -7587.337 0.019 0.020
Chain 1: 2500 -7652.623 0.016 0.017
Chain 1: 2600 -7554.942 0.015 0.013
Chain 1: 2700 -7580.389 0.011 0.013
Chain 1: 2800 -7636.776 0.010 0.011
Chain 1: 2900 -7416.495 0.013 0.013
Chain 1: 3000 -7558.405 0.013 0.013
Chain 1: 3100 -7560.512 0.012 0.013
Chain 1: 3200 -7768.694 0.013 0.013
Chain 1: 3300 -7487.296 0.015 0.013
Chain 1: 3400 -7717.811 0.018 0.019
Chain 1: 3500 -7470.833 0.020 0.027
Chain 1: 3600 -7537.047 0.020 0.027
Chain 1: 3700 -7486.642 0.020 0.027
Chain 1: 3800 -7484.010 0.019 0.027
Chain 1: 3900 -7449.987 0.017 0.019
Chain 1: 4000 -7444.889 0.015 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003241 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.41 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87702.783 1.000 1.000
Chain 1: 200 -13862.910 3.163 5.326
Chain 1: 300 -10149.200 2.231 1.000
Chain 1: 400 -11514.996 1.703 1.000
Chain 1: 500 -9028.927 1.417 0.366
Chain 1: 600 -8844.574 1.185 0.366
Chain 1: 700 -9266.836 1.022 0.275
Chain 1: 800 -8444.513 0.906 0.275
Chain 1: 900 -8535.396 0.807 0.119
Chain 1: 1000 -8704.282 0.728 0.119
Chain 1: 1100 -8971.593 0.631 0.097 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8511.524 0.104 0.054
Chain 1: 1300 -8842.224 0.071 0.046
Chain 1: 1400 -8778.473 0.060 0.037
Chain 1: 1500 -8675.283 0.033 0.030
Chain 1: 1600 -8777.237 0.033 0.030
Chain 1: 1700 -8842.355 0.029 0.019
Chain 1: 1800 -8407.005 0.024 0.019
Chain 1: 1900 -8511.385 0.024 0.019
Chain 1: 2000 -8487.084 0.023 0.012
Chain 1: 2100 -8633.696 0.021 0.012
Chain 1: 2200 -8419.066 0.018 0.012
Chain 1: 2300 -8575.057 0.017 0.012
Chain 1: 2400 -8414.304 0.018 0.017
Chain 1: 2500 -8485.261 0.017 0.017
Chain 1: 2600 -8397.541 0.017 0.017
Chain 1: 2700 -8431.548 0.017 0.017
Chain 1: 2800 -8391.710 0.012 0.012
Chain 1: 2900 -8484.874 0.012 0.011
Chain 1: 3000 -8316.863 0.014 0.017
Chain 1: 3100 -8474.222 0.014 0.018
Chain 1: 3200 -8346.367 0.013 0.015
Chain 1: 3300 -8354.052 0.011 0.011
Chain 1: 3400 -8512.630 0.011 0.011
Chain 1: 3500 -8518.446 0.010 0.011
Chain 1: 3600 -8302.978 0.012 0.015
Chain 1: 3700 -8448.485 0.013 0.017
Chain 1: 3800 -8309.564 0.015 0.017
Chain 1: 3900 -8244.217 0.014 0.017
Chain 1: 4000 -8319.389 0.013 0.017
Chain 1: 4100 -8309.889 0.011 0.015
Chain 1: 4200 -8299.856 0.010 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003594 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.94 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8467646.593 1.000 1.000
Chain 1: 200 -1595688.239 2.653 4.307
Chain 1: 300 -892783.756 2.031 1.000
Chain 1: 400 -458438.477 1.760 1.000
Chain 1: 500 -357650.060 1.465 0.947
Chain 1: 600 -232212.972 1.311 0.947
Chain 1: 700 -118930.433 1.259 0.947
Chain 1: 800 -86288.242 1.149 0.947
Chain 1: 900 -66754.107 1.054 0.787
Chain 1: 1000 -51669.669 0.978 0.787
Chain 1: 1100 -39255.834 0.909 0.540 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38448.436 0.481 0.378
Chain 1: 1300 -26508.523 0.447 0.378
Chain 1: 1400 -26239.419 0.354 0.316
Chain 1: 1500 -22853.945 0.340 0.316
Chain 1: 1600 -22079.269 0.290 0.293
Chain 1: 1700 -20965.123 0.200 0.292
Chain 1: 1800 -20912.332 0.162 0.148
Chain 1: 1900 -21238.894 0.134 0.053
Chain 1: 2000 -19755.747 0.113 0.053
Chain 1: 2100 -19993.750 0.082 0.035
Chain 1: 2200 -20219.543 0.081 0.035
Chain 1: 2300 -19837.263 0.038 0.019
Chain 1: 2400 -19609.360 0.038 0.019
Chain 1: 2500 -19410.916 0.025 0.015
Chain 1: 2600 -19041.114 0.023 0.015
Chain 1: 2700 -18998.133 0.018 0.012
Chain 1: 2800 -18714.648 0.019 0.015
Chain 1: 2900 -18995.911 0.019 0.015
Chain 1: 3000 -18982.168 0.012 0.012
Chain 1: 3100 -19067.190 0.011 0.012
Chain 1: 3200 -18757.700 0.011 0.015
Chain 1: 3300 -18962.574 0.011 0.012
Chain 1: 3400 -18437.009 0.012 0.015
Chain 1: 3500 -19049.389 0.014 0.015
Chain 1: 3600 -18355.391 0.016 0.015
Chain 1: 3700 -18742.608 0.018 0.016
Chain 1: 3800 -17701.109 0.023 0.021
Chain 1: 3900 -17697.151 0.021 0.021
Chain 1: 4000 -17814.546 0.022 0.021
Chain 1: 4100 -17728.219 0.022 0.021
Chain 1: 4200 -17544.198 0.021 0.021
Chain 1: 4300 -17682.817 0.021 0.021
Chain 1: 4400 -17639.424 0.018 0.010
Chain 1: 4500 -17541.870 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001495 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.95 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12932.383 1.000 1.000
Chain 1: 200 -9787.186 0.661 1.000
Chain 1: 300 -8084.968 0.511 0.321
Chain 1: 400 -8372.487 0.392 0.321
Chain 1: 500 -8109.920 0.320 0.211
Chain 1: 600 -8077.881 0.267 0.211
Chain 1: 700 -7957.043 0.231 0.034
Chain 1: 800 -8134.298 0.205 0.034
Chain 1: 900 -7866.492 0.186 0.034
Chain 1: 1000 -8118.245 0.170 0.034
Chain 1: 1100 -8361.493 0.073 0.032
Chain 1: 1200 -7989.349 0.046 0.032
Chain 1: 1300 -8106.592 0.026 0.031
Chain 1: 1400 -7943.472 0.025 0.029
Chain 1: 1500 -8163.146 0.024 0.027
Chain 1: 1600 -7983.908 0.026 0.027
Chain 1: 1700 -7920.700 0.025 0.027
Chain 1: 1800 -7888.503 0.024 0.027
Chain 1: 1900 -7903.177 0.020 0.022
Chain 1: 2000 -7848.657 0.018 0.021
Chain 1: 2100 -7854.629 0.015 0.014
Chain 1: 2200 -7832.490 0.011 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001698 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 16.98 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58697.294 1.000 1.000
Chain 1: 200 -18292.353 1.604 2.209
Chain 1: 300 -8953.916 1.417 1.043
Chain 1: 400 -8029.156 1.092 1.043
Chain 1: 500 -8827.527 0.891 1.000
Chain 1: 600 -8502.041 0.749 1.000
Chain 1: 700 -7786.098 0.655 0.115
Chain 1: 800 -8344.866 0.582 0.115
Chain 1: 900 -8013.870 0.522 0.092
Chain 1: 1000 -7975.046 0.470 0.092
Chain 1: 1100 -7525.922 0.376 0.090
Chain 1: 1200 -7816.663 0.159 0.067
Chain 1: 1300 -7711.855 0.056 0.060
Chain 1: 1400 -7889.659 0.047 0.041
Chain 1: 1500 -7516.454 0.043 0.041
Chain 1: 1600 -7789.399 0.042 0.041
Chain 1: 1700 -7614.401 0.035 0.037
Chain 1: 1800 -7618.276 0.029 0.035
Chain 1: 1900 -7595.170 0.025 0.023
Chain 1: 2000 -7734.507 0.026 0.023
Chain 1: 2100 -7610.976 0.022 0.023
Chain 1: 2200 -7768.843 0.020 0.020
Chain 1: 2300 -7606.079 0.021 0.021
Chain 1: 2400 -7567.047 0.019 0.020
Chain 1: 2500 -7406.864 0.016 0.020
Chain 1: 2600 -7527.001 0.015 0.018
Chain 1: 2700 -7519.328 0.012 0.016
Chain 1: 2800 -7525.431 0.012 0.016
Chain 1: 2900 -7371.221 0.014 0.018
Chain 1: 3000 -7519.754 0.014 0.020
Chain 1: 3100 -7513.980 0.013 0.020
Chain 1: 3200 -7765.540 0.014 0.020
Chain 1: 3300 -7436.166 0.016 0.020
Chain 1: 3400 -7705.043 0.019 0.021
Chain 1: 3500 -7433.722 0.021 0.021
Chain 1: 3600 -7487.933 0.020 0.021
Chain 1: 3700 -7447.188 0.020 0.021
Chain 1: 3800 -7415.134 0.021 0.021
Chain 1: 3900 -7401.018 0.019 0.020
Chain 1: 4000 -7396.042 0.017 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003979 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 39.79 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86832.080 1.000 1.000
Chain 1: 200 -14054.472 3.089 5.178
Chain 1: 300 -10202.368 2.185 1.000
Chain 1: 400 -12455.010 1.684 1.000
Chain 1: 500 -8571.654 1.438 0.453
Chain 1: 600 -8476.723 1.200 0.453
Chain 1: 700 -8824.002 1.034 0.378
Chain 1: 800 -9127.693 0.909 0.378
Chain 1: 900 -9006.542 0.810 0.181
Chain 1: 1000 -8982.722 0.729 0.181
Chain 1: 1100 -8883.766 0.630 0.039 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8369.037 0.118 0.039
Chain 1: 1300 -8876.910 0.086 0.039
Chain 1: 1400 -8584.964 0.072 0.034
Chain 1: 1500 -8656.664 0.027 0.033
Chain 1: 1600 -8727.129 0.027 0.033
Chain 1: 1700 -8782.323 0.024 0.013
Chain 1: 1800 -8311.809 0.026 0.013
Chain 1: 1900 -8426.027 0.026 0.014
Chain 1: 2000 -8444.422 0.026 0.014
Chain 1: 2100 -8550.343 0.026 0.014
Chain 1: 2200 -8303.245 0.023 0.014
Chain 1: 2300 -8455.744 0.019 0.014
Chain 1: 2400 -8336.706 0.017 0.014
Chain 1: 2500 -8395.212 0.017 0.014
Chain 1: 2600 -8300.946 0.017 0.014
Chain 1: 2700 -8336.786 0.017 0.014
Chain 1: 2800 -8299.777 0.012 0.012
Chain 1: 2900 -8404.093 0.012 0.012
Chain 1: 3000 -8312.362 0.012 0.012
Chain 1: 3100 -8278.963 0.012 0.011
Chain 1: 3200 -8247.187 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003504 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.04 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8399446.300 1.000 1.000
Chain 1: 200 -1585337.316 2.649 4.298
Chain 1: 300 -890684.197 2.026 1.000
Chain 1: 400 -458063.572 1.756 1.000
Chain 1: 500 -358379.882 1.460 0.944
Chain 1: 600 -233439.629 1.306 0.944
Chain 1: 700 -119772.209 1.255 0.944
Chain 1: 800 -86977.777 1.145 0.944
Chain 1: 900 -67349.256 1.050 0.780
Chain 1: 1000 -52179.339 0.974 0.780
Chain 1: 1100 -39668.019 0.906 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38857.028 0.478 0.377
Chain 1: 1300 -26810.925 0.445 0.377
Chain 1: 1400 -26534.855 0.352 0.315
Chain 1: 1500 -23119.426 0.339 0.315
Chain 1: 1600 -22336.541 0.289 0.291
Chain 1: 1700 -21208.960 0.199 0.291
Chain 1: 1800 -21153.570 0.162 0.148
Chain 1: 1900 -21480.765 0.134 0.053
Chain 1: 2000 -19988.919 0.112 0.053
Chain 1: 2100 -20227.799 0.082 0.035
Chain 1: 2200 -20454.918 0.081 0.035
Chain 1: 2300 -20071.210 0.038 0.019
Chain 1: 2400 -19842.892 0.038 0.019
Chain 1: 2500 -19644.745 0.024 0.015
Chain 1: 2600 -19274.087 0.023 0.015
Chain 1: 2700 -19230.763 0.018 0.012
Chain 1: 2800 -18947.034 0.019 0.015
Chain 1: 2900 -19228.771 0.019 0.015
Chain 1: 3000 -19215.009 0.012 0.012
Chain 1: 3100 -19300.150 0.011 0.012
Chain 1: 3200 -18990.169 0.011 0.015
Chain 1: 3300 -19195.385 0.010 0.012
Chain 1: 3400 -18669.051 0.012 0.015
Chain 1: 3500 -19282.806 0.014 0.015
Chain 1: 3600 -18587.019 0.016 0.015
Chain 1: 3700 -18975.641 0.018 0.016
Chain 1: 3800 -17931.492 0.022 0.020
Chain 1: 3900 -17927.491 0.021 0.020
Chain 1: 4000 -18044.849 0.021 0.020
Chain 1: 4100 -17958.399 0.021 0.020
Chain 1: 4200 -17773.789 0.021 0.020
Chain 1: 4300 -17912.833 0.021 0.020
Chain 1: 4400 -17868.980 0.018 0.010
Chain 1: 4500 -17771.338 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001368 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.68 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12533.623 1.000 1.000
Chain 1: 200 -9335.083 0.671 1.000
Chain 1: 300 -8021.874 0.502 0.343
Chain 1: 400 -8102.862 0.379 0.343
Chain 1: 500 -8070.428 0.304 0.164
Chain 1: 600 -7950.552 0.256 0.164
Chain 1: 700 -8018.872 0.221 0.015
Chain 1: 800 -7859.086 0.196 0.020
Chain 1: 900 -7772.497 0.175 0.015
Chain 1: 1000 -7965.430 0.160 0.020
Chain 1: 1100 -7986.497 0.060 0.015
Chain 1: 1200 -7874.572 0.027 0.014
Chain 1: 1300 -7832.006 0.012 0.011
Chain 1: 1400 -7854.461 0.011 0.011
Chain 1: 1500 -7944.117 0.012 0.011
Chain 1: 1600 -7885.855 0.011 0.011
Chain 1: 1700 -7831.836 0.011 0.011
Chain 1: 1800 -7809.239 0.009 0.007 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001446 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.46 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -62034.949 1.000 1.000
Chain 1: 200 -17870.818 1.736 2.471
Chain 1: 300 -8862.153 1.496 1.017
Chain 1: 400 -8298.250 1.139 1.017
Chain 1: 500 -8525.210 0.916 1.000
Chain 1: 600 -8157.251 0.771 1.000
Chain 1: 700 -7801.394 0.668 0.068
Chain 1: 800 -8154.409 0.590 0.068
Chain 1: 900 -7934.184 0.527 0.046
Chain 1: 1000 -7604.795 0.479 0.046
Chain 1: 1100 -7633.466 0.379 0.045
Chain 1: 1200 -7665.281 0.132 0.043
Chain 1: 1300 -7752.244 0.032 0.043
Chain 1: 1400 -7815.293 0.026 0.028
Chain 1: 1500 -7550.364 0.027 0.035
Chain 1: 1600 -7629.373 0.023 0.028
Chain 1: 1700 -7536.964 0.020 0.012
Chain 1: 1800 -7556.191 0.016 0.011
Chain 1: 1900 -7580.518 0.013 0.010
Chain 1: 2000 -7565.597 0.009 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003246 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.46 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85870.437 1.000 1.000
Chain 1: 200 -13495.619 3.181 5.363
Chain 1: 300 -9858.262 2.244 1.000
Chain 1: 400 -10642.706 1.701 1.000
Chain 1: 500 -8809.756 1.403 0.369
Chain 1: 600 -8343.178 1.178 0.369
Chain 1: 700 -8204.651 1.012 0.208
Chain 1: 800 -8916.192 0.896 0.208
Chain 1: 900 -8693.136 0.799 0.080
Chain 1: 1000 -8516.795 0.721 0.080
Chain 1: 1100 -8641.732 0.623 0.074 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8198.215 0.092 0.056
Chain 1: 1300 -8418.217 0.058 0.054
Chain 1: 1400 -8576.419 0.052 0.026
Chain 1: 1500 -8408.382 0.033 0.026
Chain 1: 1600 -8522.675 0.029 0.021
Chain 1: 1700 -8598.011 0.028 0.021
Chain 1: 1800 -8180.462 0.025 0.021
Chain 1: 1900 -8278.397 0.024 0.020
Chain 1: 2000 -8252.239 0.022 0.018
Chain 1: 2100 -8376.440 0.022 0.018
Chain 1: 2200 -8189.063 0.019 0.018
Chain 1: 2300 -8272.941 0.017 0.015
Chain 1: 2400 -8342.391 0.016 0.013
Chain 1: 2500 -8288.359 0.015 0.012
Chain 1: 2600 -8288.752 0.014 0.010
Chain 1: 2700 -8205.901 0.014 0.010
Chain 1: 2800 -8167.350 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003747 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 37.47 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8388325.418 1.000 1.000
Chain 1: 200 -1580634.647 2.653 4.307
Chain 1: 300 -890297.842 2.027 1.000
Chain 1: 400 -457478.430 1.757 1.000
Chain 1: 500 -358340.811 1.461 0.946
Chain 1: 600 -233325.895 1.307 0.946
Chain 1: 700 -119417.641 1.256 0.946
Chain 1: 800 -86597.472 1.147 0.946
Chain 1: 900 -66900.679 1.052 0.775
Chain 1: 1000 -51671.853 0.976 0.775
Chain 1: 1100 -39119.711 0.908 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38293.607 0.480 0.379
Chain 1: 1300 -26217.443 0.448 0.379
Chain 1: 1400 -25933.950 0.355 0.321
Chain 1: 1500 -22512.826 0.342 0.321
Chain 1: 1600 -21727.021 0.292 0.295
Chain 1: 1700 -20596.676 0.203 0.294
Chain 1: 1800 -20540.003 0.165 0.152
Chain 1: 1900 -20866.218 0.137 0.055
Chain 1: 2000 -19375.195 0.115 0.055
Chain 1: 2100 -19613.638 0.084 0.036
Chain 1: 2200 -19840.522 0.083 0.036
Chain 1: 2300 -19457.358 0.039 0.020
Chain 1: 2400 -19229.374 0.039 0.020
Chain 1: 2500 -19031.560 0.025 0.016
Chain 1: 2600 -18661.569 0.024 0.016
Chain 1: 2700 -18618.463 0.018 0.012
Chain 1: 2800 -18335.376 0.020 0.015
Chain 1: 2900 -18616.707 0.020 0.015
Chain 1: 3000 -18602.828 0.012 0.012
Chain 1: 3100 -18687.846 0.011 0.012
Chain 1: 3200 -18378.469 0.012 0.015
Chain 1: 3300 -18583.233 0.011 0.012
Chain 1: 3400 -18058.125 0.013 0.015
Chain 1: 3500 -18670.139 0.015 0.015
Chain 1: 3600 -17976.642 0.017 0.015
Chain 1: 3700 -18363.625 0.019 0.017
Chain 1: 3800 -17323.111 0.023 0.021
Chain 1: 3900 -17319.277 0.021 0.021
Chain 1: 4000 -17436.552 0.022 0.021
Chain 1: 4100 -17350.333 0.022 0.021
Chain 1: 4200 -17166.517 0.022 0.021
Chain 1: 4300 -17304.941 0.021 0.021
Chain 1: 4400 -17261.715 0.019 0.011
Chain 1: 4500 -17164.263 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001704 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 17.04 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13273.355 1.000 1.000
Chain 1: 200 -10051.473 0.660 1.000
Chain 1: 300 -8709.911 0.492 0.321
Chain 1: 400 -8924.675 0.375 0.321
Chain 1: 500 -8809.992 0.302 0.154
Chain 1: 600 -8648.912 0.255 0.154
Chain 1: 700 -8545.233 0.220 0.024
Chain 1: 800 -8507.592 0.193 0.024
Chain 1: 900 -8516.643 0.172 0.019
Chain 1: 1000 -8668.151 0.157 0.019
Chain 1: 1100 -8646.281 0.057 0.017
Chain 1: 1200 -8575.134 0.026 0.013
Chain 1: 1300 -8488.431 0.011 0.012
Chain 1: 1400 -8500.942 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002315 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 23.15 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -62557.668 1.000 1.000
Chain 1: 200 -18894.808 1.655 2.311
Chain 1: 300 -9362.830 1.443 1.018
Chain 1: 400 -10208.127 1.103 1.018
Chain 1: 500 -8497.046 0.923 1.000
Chain 1: 600 -9062.837 0.779 1.000
Chain 1: 700 -9116.902 0.669 0.201
Chain 1: 800 -8320.039 0.597 0.201
Chain 1: 900 -8122.048 0.534 0.096
Chain 1: 1000 -7921.513 0.483 0.096
Chain 1: 1100 -7701.907 0.386 0.083
Chain 1: 1200 -7889.331 0.157 0.062
Chain 1: 1300 -7834.225 0.056 0.029
Chain 1: 1400 -7820.476 0.048 0.025
Chain 1: 1500 -7623.519 0.030 0.025
Chain 1: 1600 -7745.466 0.025 0.024
Chain 1: 1700 -7675.540 0.026 0.024
Chain 1: 1800 -7770.401 0.017 0.024
Chain 1: 1900 -7659.772 0.016 0.016
Chain 1: 2000 -7768.297 0.015 0.014
Chain 1: 2100 -7578.412 0.015 0.014
Chain 1: 2200 -7885.005 0.016 0.014
Chain 1: 2300 -7749.436 0.017 0.016
Chain 1: 2400 -7577.772 0.020 0.017
Chain 1: 2500 -7662.207 0.018 0.016
Chain 1: 2600 -7588.356 0.017 0.014
Chain 1: 2700 -7455.878 0.018 0.017
Chain 1: 2800 -7625.243 0.019 0.018
Chain 1: 2900 -7413.084 0.021 0.022
Chain 1: 3000 -7570.836 0.021 0.022
Chain 1: 3100 -7550.424 0.019 0.021
Chain 1: 3200 -7805.024 0.019 0.021
Chain 1: 3300 -7462.447 0.021 0.022
Chain 1: 3400 -7579.661 0.021 0.021
Chain 1: 3500 -7488.383 0.021 0.021
Chain 1: 3600 -7520.829 0.020 0.021
Chain 1: 3700 -7468.436 0.019 0.021
Chain 1: 3800 -7430.746 0.017 0.015
Chain 1: 3900 -7434.079 0.015 0.012
Chain 1: 4000 -7426.970 0.013 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003213 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.13 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86982.366 1.000 1.000
Chain 1: 200 -14436.394 3.013 5.025
Chain 1: 300 -10693.843 2.125 1.000
Chain 1: 400 -12330.374 1.627 1.000
Chain 1: 500 -9271.622 1.368 0.350
Chain 1: 600 -9294.110 1.140 0.350
Chain 1: 700 -9114.162 0.980 0.330
Chain 1: 800 -9342.108 0.861 0.330
Chain 1: 900 -9485.542 0.767 0.133
Chain 1: 1000 -9126.063 0.694 0.133
Chain 1: 1100 -9465.339 0.597 0.039 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -9012.215 0.100 0.039
Chain 1: 1300 -9322.218 0.068 0.036
Chain 1: 1400 -9158.584 0.057 0.033
Chain 1: 1500 -9159.744 0.024 0.024
Chain 1: 1600 -9269.117 0.025 0.024
Chain 1: 1700 -9322.907 0.023 0.024
Chain 1: 1800 -8872.842 0.026 0.033
Chain 1: 1900 -8981.459 0.026 0.033
Chain 1: 2000 -8974.509 0.022 0.018
Chain 1: 2100 -9151.403 0.020 0.018
Chain 1: 2200 -8875.839 0.018 0.018
Chain 1: 2300 -9063.024 0.017 0.018
Chain 1: 2400 -8876.975 0.017 0.019
Chain 1: 2500 -8954.348 0.018 0.019
Chain 1: 2600 -8868.475 0.018 0.019
Chain 1: 2700 -8899.801 0.018 0.019
Chain 1: 2800 -8851.810 0.013 0.012
Chain 1: 2900 -8960.800 0.013 0.012
Chain 1: 3000 -8910.149 0.014 0.012
Chain 1: 3100 -8842.894 0.013 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003779 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 37.79 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8408023.458 1.000 1.000
Chain 1: 200 -1583636.191 2.655 4.309
Chain 1: 300 -891085.257 2.029 1.000
Chain 1: 400 -458102.615 1.758 1.000
Chain 1: 500 -358532.501 1.462 0.945
Chain 1: 600 -233784.384 1.307 0.945
Chain 1: 700 -120133.546 1.256 0.945
Chain 1: 800 -87364.289 1.146 0.945
Chain 1: 900 -67732.667 1.050 0.777
Chain 1: 1000 -52550.158 0.974 0.777
Chain 1: 1100 -40036.590 0.906 0.534 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39220.673 0.477 0.375
Chain 1: 1300 -27172.209 0.443 0.375
Chain 1: 1400 -26894.375 0.350 0.313
Chain 1: 1500 -23479.685 0.337 0.313
Chain 1: 1600 -22696.678 0.287 0.290
Chain 1: 1700 -21569.089 0.197 0.289
Chain 1: 1800 -21513.464 0.160 0.145
Chain 1: 1900 -21840.245 0.133 0.052
Chain 1: 2000 -20349.600 0.111 0.052
Chain 1: 2100 -20588.082 0.081 0.034
Chain 1: 2200 -20815.048 0.080 0.034
Chain 1: 2300 -20431.681 0.037 0.019
Chain 1: 2400 -20203.547 0.038 0.019
Chain 1: 2500 -20005.534 0.024 0.015
Chain 1: 2600 -19635.024 0.022 0.015
Chain 1: 2700 -19591.866 0.017 0.012
Chain 1: 2800 -19308.404 0.019 0.015
Chain 1: 2900 -19589.962 0.019 0.014
Chain 1: 3000 -19576.172 0.011 0.012
Chain 1: 3100 -19661.200 0.011 0.011
Chain 1: 3200 -19351.447 0.011 0.014
Chain 1: 3300 -19556.550 0.010 0.011
Chain 1: 3400 -19030.642 0.012 0.014
Chain 1: 3500 -19643.683 0.014 0.015
Chain 1: 3600 -18948.918 0.016 0.015
Chain 1: 3700 -19336.750 0.018 0.016
Chain 1: 3800 -18294.135 0.022 0.020
Chain 1: 3900 -18290.234 0.020 0.020
Chain 1: 4000 -18407.555 0.021 0.020
Chain 1: 4100 -18321.133 0.021 0.020
Chain 1: 4200 -18136.946 0.020 0.020
Chain 1: 4300 -18275.672 0.020 0.020
Chain 1: 4400 -18232.070 0.018 0.010
Chain 1: 4500 -18134.544 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001892 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 18.92 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48504.762 1.000 1.000
Chain 1: 200 -17434.642 1.391 1.782
Chain 1: 300 -17567.739 0.930 1.000
Chain 1: 400 -12674.476 0.794 1.000
Chain 1: 500 -14003.411 0.654 0.386
Chain 1: 600 -12054.739 0.572 0.386
Chain 1: 700 -14755.757 0.516 0.183
Chain 1: 800 -11698.449 0.485 0.261
Chain 1: 900 -17372.795 0.467 0.261
Chain 1: 1000 -11498.627 0.471 0.327
Chain 1: 1100 -30352.918 0.434 0.327
Chain 1: 1200 -12607.178 0.396 0.327
Chain 1: 1300 -12069.822 0.400 0.327
Chain 1: 1400 -14990.395 0.381 0.261
Chain 1: 1500 -10507.684 0.414 0.327
Chain 1: 1600 -13427.390 0.419 0.327
Chain 1: 1700 -11780.842 0.415 0.327
Chain 1: 1800 -20286.313 0.431 0.419
Chain 1: 1900 -9501.128 0.512 0.427 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2000 -10400.913 0.469 0.419
Chain 1: 2100 -10079.316 0.410 0.217
Chain 1: 2200 -9529.524 0.275 0.195
Chain 1: 2300 -16808.226 0.314 0.217
Chain 1: 2400 -8734.485 0.387 0.419
Chain 1: 2500 -12013.114 0.372 0.273
Chain 1: 2600 -9760.814 0.373 0.273
Chain 1: 2700 -9754.984 0.359 0.273
Chain 1: 2800 -8880.453 0.327 0.231
Chain 1: 2900 -9158.058 0.217 0.098
Chain 1: 3000 -14969.018 0.247 0.231
Chain 1: 3100 -9734.273 0.297 0.273
Chain 1: 3200 -15386.109 0.328 0.367
Chain 1: 3300 -8828.815 0.359 0.367
Chain 1: 3400 -9053.181 0.269 0.273
Chain 1: 3500 -15466.979 0.284 0.367
Chain 1: 3600 -17248.916 0.271 0.367
Chain 1: 3700 -9267.967 0.357 0.388
Chain 1: 3800 -9547.805 0.350 0.388
Chain 1: 3900 -13459.615 0.376 0.388
Chain 1: 4000 -8969.645 0.387 0.415
Chain 1: 4100 -11340.815 0.354 0.367
Chain 1: 4200 -12716.310 0.328 0.291
Chain 1: 4300 -9936.191 0.282 0.280
Chain 1: 4400 -14966.344 0.313 0.291
Chain 1: 4500 -9117.567 0.336 0.291
Chain 1: 4600 -13079.159 0.356 0.303
Chain 1: 4700 -13173.926 0.271 0.291
Chain 1: 4800 -11767.018 0.280 0.291
Chain 1: 4900 -8368.119 0.291 0.303
Chain 1: 5000 -10681.933 0.263 0.280
Chain 1: 5100 -8580.892 0.266 0.280
Chain 1: 5200 -14455.958 0.296 0.303
Chain 1: 5300 -8304.663 0.342 0.336
Chain 1: 5400 -11828.552 0.338 0.303
Chain 1: 5500 -8532.259 0.313 0.303
Chain 1: 5600 -9929.204 0.297 0.298
Chain 1: 5700 -8549.184 0.312 0.298
Chain 1: 5800 -10376.755 0.318 0.298
Chain 1: 5900 -8551.526 0.298 0.245
Chain 1: 6000 -12844.859 0.310 0.298
Chain 1: 6100 -8374.094 0.339 0.334
Chain 1: 6200 -9643.563 0.312 0.298
Chain 1: 6300 -13507.782 0.266 0.286
Chain 1: 6400 -8050.385 0.304 0.286
Chain 1: 6500 -8715.033 0.273 0.213
Chain 1: 6600 -9746.827 0.270 0.213
Chain 1: 6700 -12614.881 0.276 0.227
Chain 1: 6800 -8994.598 0.299 0.286
Chain 1: 6900 -9242.846 0.280 0.286
Chain 1: 7000 -8205.703 0.259 0.227
Chain 1: 7100 -8226.492 0.206 0.132
Chain 1: 7200 -8686.304 0.198 0.126
Chain 1: 7300 -9125.623 0.175 0.106
Chain 1: 7400 -8141.004 0.119 0.106
Chain 1: 7500 -8021.362 0.113 0.106
Chain 1: 7600 -10209.961 0.124 0.121
Chain 1: 7700 -8030.147 0.128 0.121
Chain 1: 7800 -9167.850 0.100 0.121
Chain 1: 7900 -8673.664 0.103 0.121
Chain 1: 8000 -9437.233 0.099 0.081
Chain 1: 8100 -8476.525 0.110 0.113
Chain 1: 8200 -8567.366 0.106 0.113
Chain 1: 8300 -8055.891 0.107 0.113
Chain 1: 8400 -11801.837 0.127 0.113
Chain 1: 8500 -10548.601 0.137 0.119
Chain 1: 8600 -8278.289 0.143 0.119
Chain 1: 8700 -9071.935 0.125 0.113
Chain 1: 8800 -8385.747 0.121 0.087
Chain 1: 8900 -8386.488 0.115 0.087
Chain 1: 9000 -10405.618 0.126 0.113
Chain 1: 9100 -8367.600 0.139 0.119
Chain 1: 9200 -12492.358 0.171 0.194
Chain 1: 9300 -8308.138 0.215 0.244
Chain 1: 9400 -8041.232 0.187 0.194
Chain 1: 9500 -10756.762 0.200 0.244
Chain 1: 9600 -9917.777 0.181 0.194
Chain 1: 9700 -8244.835 0.193 0.203
Chain 1: 9800 -8791.975 0.191 0.203
Chain 1: 9900 -9816.520 0.201 0.203
Chain 1: 10000 -8860.350 0.193 0.203
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00168 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 16.8 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61295.868 1.000 1.000
Chain 1: 200 -17606.988 1.741 2.481
Chain 1: 300 -8683.204 1.503 1.028
Chain 1: 400 -9084.816 1.138 1.028
Chain 1: 500 -7647.838 0.948 1.000
Chain 1: 600 -7955.078 0.797 1.000
Chain 1: 700 -7634.525 0.689 0.188
Chain 1: 800 -8068.436 0.609 0.188
Chain 1: 900 -7720.901 0.547 0.054
Chain 1: 1000 -7633.345 0.493 0.054
Chain 1: 1100 -7439.959 0.396 0.045
Chain 1: 1200 -7561.677 0.149 0.044
Chain 1: 1300 -7570.706 0.047 0.042
Chain 1: 1400 -7565.553 0.042 0.039
Chain 1: 1500 -7459.959 0.025 0.026
Chain 1: 1600 -7411.548 0.022 0.016
Chain 1: 1700 -7426.740 0.018 0.014
Chain 1: 1800 -7491.892 0.013 0.011
Chain 1: 1900 -7458.238 0.009 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00298 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.8 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86091.125 1.000 1.000
Chain 1: 200 -13303.702 3.236 5.471
Chain 1: 300 -9674.927 2.282 1.000
Chain 1: 400 -10510.030 1.731 1.000
Chain 1: 500 -8611.633 1.429 0.375
Chain 1: 600 -8181.299 1.200 0.375
Chain 1: 700 -8455.097 1.033 0.220
Chain 1: 800 -8779.198 0.909 0.220
Chain 1: 900 -8493.230 0.811 0.079
Chain 1: 1000 -8336.795 0.732 0.079
Chain 1: 1100 -8541.384 0.634 0.053 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8039.294 0.094 0.053
Chain 1: 1300 -8385.598 0.060 0.041
Chain 1: 1400 -8383.127 0.052 0.037
Chain 1: 1500 -8257.607 0.032 0.034
Chain 1: 1600 -8363.217 0.028 0.032
Chain 1: 1700 -8449.071 0.026 0.024
Chain 1: 1800 -8040.197 0.027 0.024
Chain 1: 1900 -8136.326 0.025 0.019
Chain 1: 2000 -8108.779 0.023 0.015
Chain 1: 2100 -8230.127 0.022 0.015
Chain 1: 2200 -8057.769 0.018 0.015
Chain 1: 2300 -8135.746 0.015 0.013
Chain 1: 2400 -8200.850 0.016 0.013
Chain 1: 2500 -8146.423 0.015 0.012
Chain 1: 2600 -8145.001 0.014 0.010
Chain 1: 2700 -8061.850 0.014 0.010
Chain 1: 2800 -8027.467 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003237 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.37 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8393651.989 1.000 1.000
Chain 1: 200 -1580359.246 2.656 4.311
Chain 1: 300 -890829.289 2.028 1.000
Chain 1: 400 -457847.539 1.758 1.000
Chain 1: 500 -358648.557 1.462 0.946
Chain 1: 600 -233434.602 1.307 0.946
Chain 1: 700 -119368.379 1.257 0.946
Chain 1: 800 -86493.278 1.147 0.946
Chain 1: 900 -66766.763 1.053 0.774
Chain 1: 1000 -51508.540 0.977 0.774
Chain 1: 1100 -38938.009 0.909 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38106.003 0.480 0.380
Chain 1: 1300 -26015.711 0.450 0.380
Chain 1: 1400 -25728.410 0.356 0.323
Chain 1: 1500 -22304.254 0.344 0.323
Chain 1: 1600 -21517.253 0.294 0.296
Chain 1: 1700 -20385.714 0.204 0.295
Chain 1: 1800 -20328.435 0.166 0.154
Chain 1: 1900 -20654.489 0.138 0.056
Chain 1: 2000 -19163.159 0.116 0.056
Chain 1: 2100 -19401.532 0.085 0.037
Chain 1: 2200 -19628.454 0.084 0.037
Chain 1: 2300 -19245.273 0.040 0.020
Chain 1: 2400 -19017.397 0.040 0.020
Chain 1: 2500 -18819.611 0.025 0.016
Chain 1: 2600 -18449.701 0.024 0.016
Chain 1: 2700 -18406.612 0.019 0.012
Chain 1: 2800 -18123.689 0.020 0.016
Chain 1: 2900 -18404.901 0.020 0.015
Chain 1: 3000 -18390.977 0.012 0.012
Chain 1: 3100 -18476.023 0.011 0.012
Chain 1: 3200 -18166.689 0.012 0.015
Chain 1: 3300 -18371.412 0.011 0.012
Chain 1: 3400 -17846.476 0.013 0.015
Chain 1: 3500 -18458.219 0.015 0.016
Chain 1: 3600 -17765.050 0.017 0.016
Chain 1: 3700 -18151.846 0.019 0.017
Chain 1: 3800 -17111.832 0.023 0.021
Chain 1: 3900 -17108.031 0.022 0.021
Chain 1: 4000 -17225.272 0.022 0.021
Chain 1: 4100 -17139.107 0.022 0.021
Chain 1: 4200 -16955.408 0.022 0.021
Chain 1: 4300 -17093.737 0.021 0.021
Chain 1: 4400 -17050.633 0.019 0.011
Chain 1: 4500 -16953.198 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001534 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.34 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48393.313 1.000 1.000
Chain 1: 200 -19286.448 1.255 1.509
Chain 1: 300 -20700.908 0.859 1.000
Chain 1: 400 -14197.697 0.759 1.000
Chain 1: 500 -12699.212 0.631 0.458
Chain 1: 600 -13581.981 0.536 0.458
Chain 1: 700 -12204.722 0.476 0.118
Chain 1: 800 -13477.119 0.428 0.118
Chain 1: 900 -11205.496 0.403 0.118
Chain 1: 1000 -10008.223 0.375 0.120
Chain 1: 1100 -11842.369 0.290 0.120
Chain 1: 1200 -10754.112 0.150 0.118
Chain 1: 1300 -11268.624 0.147 0.118
Chain 1: 1400 -15867.554 0.130 0.118
Chain 1: 1500 -9409.132 0.187 0.120
Chain 1: 1600 -9599.174 0.183 0.120
Chain 1: 1700 -11992.903 0.191 0.155
Chain 1: 1800 -9481.964 0.208 0.200
Chain 1: 1900 -10833.785 0.201 0.155
Chain 1: 2000 -10254.560 0.194 0.155
Chain 1: 2100 -9363.769 0.188 0.125
Chain 1: 2200 -9097.106 0.181 0.125
Chain 1: 2300 -11726.578 0.199 0.200
Chain 1: 2400 -8628.598 0.206 0.200
Chain 1: 2500 -9626.409 0.148 0.125
Chain 1: 2600 -9217.734 0.150 0.125
Chain 1: 2700 -14078.699 0.165 0.125
Chain 1: 2800 -8692.180 0.200 0.125
Chain 1: 2900 -10245.284 0.203 0.152
Chain 1: 3000 -16230.600 0.234 0.224
Chain 1: 3100 -9075.471 0.303 0.345
Chain 1: 3200 -8764.970 0.304 0.345
Chain 1: 3300 -10893.055 0.301 0.345
Chain 1: 3400 -8449.632 0.294 0.289
Chain 1: 3500 -13175.262 0.320 0.345
Chain 1: 3600 -9147.462 0.359 0.359
Chain 1: 3700 -8700.828 0.330 0.359
Chain 1: 3800 -8219.267 0.274 0.289
Chain 1: 3900 -11664.660 0.288 0.295
Chain 1: 4000 -8448.578 0.289 0.295
Chain 1: 4100 -8309.214 0.212 0.289
Chain 1: 4200 -12455.116 0.242 0.295
Chain 1: 4300 -11450.442 0.231 0.295
Chain 1: 4400 -9504.665 0.223 0.295
Chain 1: 4500 -8639.297 0.197 0.205
Chain 1: 4600 -11603.374 0.178 0.205
Chain 1: 4700 -14641.983 0.194 0.208
Chain 1: 4800 -8447.632 0.261 0.255
Chain 1: 4900 -8308.917 0.234 0.208
Chain 1: 5000 -8311.830 0.196 0.205
Chain 1: 5100 -16580.333 0.244 0.208
Chain 1: 5200 -9449.866 0.286 0.208
Chain 1: 5300 -10055.496 0.283 0.208
Chain 1: 5400 -8614.126 0.279 0.208
Chain 1: 5500 -8500.643 0.271 0.208
Chain 1: 5600 -12238.319 0.276 0.208
Chain 1: 5700 -8980.569 0.291 0.305
Chain 1: 5800 -8760.408 0.220 0.167
Chain 1: 5900 -8892.762 0.220 0.167
Chain 1: 6000 -8412.944 0.226 0.167
Chain 1: 6100 -10382.631 0.195 0.167
Chain 1: 6200 -9943.855 0.124 0.060
Chain 1: 6300 -8146.390 0.140 0.167
Chain 1: 6400 -12240.982 0.157 0.190
Chain 1: 6500 -9744.219 0.181 0.221
Chain 1: 6600 -7984.977 0.173 0.220
Chain 1: 6700 -7996.578 0.136 0.190
Chain 1: 6800 -12848.557 0.172 0.220
Chain 1: 6900 -8357.218 0.224 0.221
Chain 1: 7000 -7803.648 0.225 0.221
Chain 1: 7100 -11645.999 0.239 0.256
Chain 1: 7200 -8861.771 0.266 0.314
Chain 1: 7300 -8251.366 0.252 0.314
Chain 1: 7400 -9619.393 0.232 0.256
Chain 1: 7500 -8879.589 0.215 0.220
Chain 1: 7600 -10733.961 0.210 0.173
Chain 1: 7700 -7822.468 0.247 0.314
Chain 1: 7800 -8622.423 0.219 0.173
Chain 1: 7900 -8457.839 0.167 0.142
Chain 1: 8000 -7918.080 0.167 0.142
Chain 1: 8100 -9335.123 0.149 0.142
Chain 1: 8200 -10478.256 0.129 0.109
Chain 1: 8300 -8222.687 0.149 0.142
Chain 1: 8400 -7827.591 0.139 0.109
Chain 1: 8500 -7814.155 0.131 0.109
Chain 1: 8600 -8393.555 0.121 0.093
Chain 1: 8700 -8353.546 0.084 0.069
Chain 1: 8800 -9260.205 0.085 0.069
Chain 1: 8900 -12231.780 0.107 0.098
Chain 1: 9000 -8696.176 0.141 0.109
Chain 1: 9100 -8481.753 0.128 0.098
Chain 1: 9200 -8200.702 0.121 0.069
Chain 1: 9300 -9730.645 0.109 0.069
Chain 1: 9400 -11095.365 0.116 0.098
Chain 1: 9500 -9327.643 0.135 0.123
Chain 1: 9600 -8053.608 0.144 0.157
Chain 1: 9700 -9277.417 0.157 0.157
Chain 1: 9800 -8467.927 0.156 0.157
Chain 1: 9900 -9469.336 0.143 0.132
Chain 1: 10000 -7700.704 0.125 0.132
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001509 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.09 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57487.153 1.000 1.000
Chain 1: 200 -17079.993 1.683 2.366
Chain 1: 300 -8410.413 1.466 1.031
Chain 1: 400 -7841.490 1.117 1.031
Chain 1: 500 -8240.556 0.904 1.000
Chain 1: 600 -8495.046 0.758 1.000
Chain 1: 700 -7712.155 0.664 0.102
Chain 1: 800 -7889.989 0.584 0.102
Chain 1: 900 -7879.678 0.519 0.073
Chain 1: 1000 -7604.928 0.471 0.073
Chain 1: 1100 -7651.109 0.372 0.048
Chain 1: 1200 -7568.083 0.136 0.036
Chain 1: 1300 -7607.817 0.033 0.030
Chain 1: 1400 -7587.193 0.026 0.023
Chain 1: 1500 -7552.649 0.022 0.011
Chain 1: 1600 -7473.685 0.020 0.011
Chain 1: 1700 -7448.594 0.010 0.006 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003354 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.54 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86093.081 1.000 1.000
Chain 1: 200 -12917.329 3.332 5.665
Chain 1: 300 -9399.236 2.346 1.000
Chain 1: 400 -10268.431 1.781 1.000
Chain 1: 500 -8281.002 1.473 0.374
Chain 1: 600 -7988.753 1.233 0.374
Chain 1: 700 -8287.967 1.062 0.240
Chain 1: 800 -8462.380 0.932 0.240
Chain 1: 900 -8313.907 0.831 0.085
Chain 1: 1000 -7996.913 0.751 0.085
Chain 1: 1100 -8293.391 0.655 0.040 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -7956.675 0.093 0.040
Chain 1: 1300 -8171.460 0.058 0.037
Chain 1: 1400 -8159.213 0.050 0.036
Chain 1: 1500 -8063.984 0.027 0.036
Chain 1: 1600 -8151.348 0.024 0.026
Chain 1: 1700 -8256.819 0.022 0.021
Chain 1: 1800 -7868.742 0.025 0.026
Chain 1: 1900 -7968.464 0.024 0.026
Chain 1: 2000 -7938.390 0.021 0.013
Chain 1: 2100 -8080.311 0.019 0.013
Chain 1: 2200 -7860.043 0.017 0.013
Chain 1: 2300 -8002.379 0.017 0.013
Chain 1: 2400 -7888.049 0.018 0.014
Chain 1: 2500 -7945.240 0.017 0.014
Chain 1: 2600 -7958.946 0.017 0.014
Chain 1: 2700 -7880.601 0.016 0.014
Chain 1: 2800 -7863.591 0.012 0.013
Chain 1: 2900 -7872.635 0.010 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004321 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 43.21 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8412564.278 1.000 1.000
Chain 1: 200 -1589556.354 2.646 4.292
Chain 1: 300 -891022.362 2.025 1.000
Chain 1: 400 -457159.776 1.756 1.000
Chain 1: 500 -357020.394 1.461 0.949
Chain 1: 600 -231934.179 1.308 0.949
Chain 1: 700 -118348.092 1.258 0.949
Chain 1: 800 -85627.683 1.148 0.949
Chain 1: 900 -66011.384 1.054 0.784
Chain 1: 1000 -50841.084 0.978 0.784
Chain 1: 1100 -38357.494 0.911 0.539 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37532.606 0.484 0.382
Chain 1: 1300 -25541.439 0.452 0.382
Chain 1: 1400 -25262.483 0.359 0.325
Chain 1: 1500 -21863.853 0.346 0.325
Chain 1: 1600 -21083.634 0.296 0.298
Chain 1: 1700 -19964.324 0.205 0.297
Chain 1: 1800 -19909.701 0.167 0.155
Chain 1: 1900 -20235.146 0.139 0.056
Chain 1: 2000 -18751.208 0.117 0.056
Chain 1: 2100 -18989.376 0.086 0.037
Chain 1: 2200 -19214.739 0.085 0.037
Chain 1: 2300 -18833.049 0.040 0.020
Chain 1: 2400 -18605.443 0.040 0.020
Chain 1: 2500 -18407.302 0.026 0.016
Chain 1: 2600 -18038.476 0.024 0.016
Chain 1: 2700 -17995.711 0.019 0.013
Chain 1: 2800 -17712.817 0.020 0.016
Chain 1: 2900 -17993.663 0.020 0.016
Chain 1: 3000 -17979.929 0.012 0.013
Chain 1: 3100 -18064.818 0.011 0.012
Chain 1: 3200 -17756.050 0.012 0.016
Chain 1: 3300 -17960.339 0.011 0.012
Chain 1: 3400 -17436.193 0.013 0.016
Chain 1: 3500 -18046.631 0.015 0.016
Chain 1: 3600 -17355.126 0.017 0.016
Chain 1: 3700 -17740.547 0.019 0.017
Chain 1: 3800 -16703.090 0.024 0.022
Chain 1: 3900 -16699.264 0.022 0.022
Chain 1: 4000 -16816.589 0.023 0.022
Chain 1: 4100 -16730.491 0.023 0.022
Chain 1: 4200 -16547.338 0.022 0.022
Chain 1: 4300 -16685.326 0.022 0.022
Chain 1: 4400 -16642.654 0.019 0.011
Chain 1: 4500 -16545.242 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001405 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.05 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49063.640 1.000 1.000
Chain 1: 200 -19625.353 1.250 1.500
Chain 1: 300 -22384.739 0.874 1.000
Chain 1: 400 -29484.223 0.716 1.000
Chain 1: 500 -12925.228 0.829 1.000
Chain 1: 600 -17039.634 0.731 1.000
Chain 1: 700 -14440.728 0.652 0.241
Chain 1: 800 -16006.143 0.583 0.241
Chain 1: 900 -19925.158 0.540 0.241
Chain 1: 1000 -17921.018 0.497 0.241
Chain 1: 1100 -16442.608 0.406 0.197
Chain 1: 1200 -15427.828 0.263 0.180
Chain 1: 1300 -10942.661 0.292 0.197
Chain 1: 1400 -11038.513 0.268 0.180
Chain 1: 1500 -11926.565 0.148 0.112
Chain 1: 1600 -9984.759 0.143 0.112
Chain 1: 1700 -18716.691 0.172 0.112
Chain 1: 1800 -10059.281 0.248 0.194
Chain 1: 1900 -10243.728 0.230 0.112
Chain 1: 2000 -9829.647 0.223 0.090
Chain 1: 2100 -10359.562 0.219 0.074
Chain 1: 2200 -11045.834 0.219 0.074
Chain 1: 2300 -9420.241 0.195 0.074
Chain 1: 2400 -9853.101 0.199 0.074
Chain 1: 2500 -9588.454 0.194 0.062
Chain 1: 2600 -10217.009 0.181 0.062
Chain 1: 2700 -11220.009 0.143 0.062
Chain 1: 2800 -10269.073 0.066 0.062
Chain 1: 2900 -9581.350 0.071 0.062
Chain 1: 3000 -9190.114 0.072 0.062
Chain 1: 3100 -12593.325 0.093 0.072
Chain 1: 3200 -9671.499 0.117 0.089
Chain 1: 3300 -9565.902 0.101 0.072
Chain 1: 3400 -16802.818 0.140 0.089
Chain 1: 3500 -12754.129 0.169 0.093
Chain 1: 3600 -9151.457 0.202 0.270
Chain 1: 3700 -9276.878 0.195 0.270
Chain 1: 3800 -8716.757 0.192 0.270
Chain 1: 3900 -8837.231 0.186 0.270
Chain 1: 4000 -18674.020 0.234 0.302
Chain 1: 4100 -9147.830 0.311 0.317
Chain 1: 4200 -8928.679 0.284 0.317
Chain 1: 4300 -9800.759 0.291 0.317
Chain 1: 4400 -9005.730 0.257 0.089
Chain 1: 4500 -9790.278 0.234 0.088
Chain 1: 4600 -14623.691 0.227 0.088
Chain 1: 4700 -13379.835 0.235 0.089
Chain 1: 4800 -8805.188 0.281 0.093
Chain 1: 4900 -9112.733 0.283 0.093
Chain 1: 5000 -11031.877 0.247 0.093
Chain 1: 5100 -9003.326 0.166 0.093
Chain 1: 5200 -9242.803 0.166 0.093
Chain 1: 5300 -10922.188 0.172 0.154
Chain 1: 5400 -13653.322 0.184 0.174
Chain 1: 5500 -10896.442 0.201 0.200
Chain 1: 5600 -14636.774 0.193 0.200
Chain 1: 5700 -9114.963 0.245 0.225
Chain 1: 5800 -12468.542 0.220 0.225
Chain 1: 5900 -9940.622 0.242 0.253
Chain 1: 6000 -8905.755 0.236 0.253
Chain 1: 6100 -8483.476 0.218 0.253
Chain 1: 6200 -8601.991 0.217 0.253
Chain 1: 6300 -9508.889 0.211 0.253
Chain 1: 6400 -11914.897 0.211 0.253
Chain 1: 6500 -8936.930 0.219 0.254
Chain 1: 6600 -9128.167 0.196 0.202
Chain 1: 6700 -11389.729 0.155 0.199
Chain 1: 6800 -14315.855 0.149 0.199
Chain 1: 6900 -11553.058 0.147 0.199
Chain 1: 7000 -15367.647 0.161 0.202
Chain 1: 7100 -9936.195 0.210 0.204
Chain 1: 7200 -12627.871 0.230 0.213
Chain 1: 7300 -8754.607 0.265 0.239
Chain 1: 7400 -9004.067 0.247 0.239
Chain 1: 7500 -9364.429 0.218 0.213
Chain 1: 7600 -8764.488 0.223 0.213
Chain 1: 7700 -8938.401 0.205 0.213
Chain 1: 7800 -11995.907 0.210 0.239
Chain 1: 7900 -8399.368 0.229 0.248
Chain 1: 8000 -8259.017 0.206 0.213
Chain 1: 8100 -8530.617 0.154 0.068
Chain 1: 8200 -8903.579 0.137 0.042
Chain 1: 8300 -8671.389 0.095 0.038
Chain 1: 8400 -12140.268 0.121 0.042
Chain 1: 8500 -11695.600 0.121 0.042
Chain 1: 8600 -8674.704 0.149 0.042
Chain 1: 8700 -8656.408 0.147 0.042
Chain 1: 8800 -8431.732 0.125 0.038
Chain 1: 8900 -9213.227 0.090 0.038
Chain 1: 9000 -10813.289 0.103 0.042
Chain 1: 9100 -9294.398 0.117 0.085
Chain 1: 9200 -8762.592 0.118 0.085
Chain 1: 9300 -8709.178 0.116 0.085
Chain 1: 9400 -9140.051 0.093 0.061
Chain 1: 9500 -8072.460 0.102 0.085
Chain 1: 9600 -8535.145 0.073 0.061
Chain 1: 9700 -11204.065 0.096 0.085
Chain 1: 9800 -10825.890 0.097 0.085
Chain 1: 9900 -10500.870 0.092 0.061
Chain 1: 10000 -10134.311 0.080 0.054
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001399 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.99 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61886.573 1.000 1.000
Chain 1: 200 -18005.480 1.719 2.437
Chain 1: 300 -8959.751 1.482 1.010
Chain 1: 400 -9451.964 1.125 1.010
Chain 1: 500 -8106.861 0.933 1.000
Chain 1: 600 -8496.069 0.785 1.000
Chain 1: 700 -8241.630 0.677 0.166
Chain 1: 800 -7880.995 0.598 0.166
Chain 1: 900 -8061.964 0.534 0.052
Chain 1: 1000 -7862.928 0.483 0.052
Chain 1: 1100 -7759.554 0.385 0.046
Chain 1: 1200 -7871.746 0.143 0.046
Chain 1: 1300 -7684.716 0.044 0.031
Chain 1: 1400 -7868.972 0.041 0.025
Chain 1: 1500 -7666.098 0.027 0.025
Chain 1: 1600 -7852.302 0.025 0.024
Chain 1: 1700 -7487.926 0.027 0.024
Chain 1: 1800 -7699.756 0.025 0.024
Chain 1: 1900 -7654.208 0.023 0.024
Chain 1: 2000 -7698.128 0.021 0.024
Chain 1: 2100 -7628.520 0.021 0.024
Chain 1: 2200 -7760.205 0.021 0.024
Chain 1: 2300 -7664.868 0.020 0.023
Chain 1: 2400 -7666.390 0.018 0.017
Chain 1: 2500 -7837.927 0.017 0.017
Chain 1: 2600 -7574.619 0.018 0.017
Chain 1: 2700 -7610.474 0.014 0.012
Chain 1: 2800 -7625.601 0.011 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.0041 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 41 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86839.387 1.000 1.000
Chain 1: 200 -13688.299 3.172 5.344
Chain 1: 300 -10005.600 2.237 1.000
Chain 1: 400 -10926.973 1.699 1.000
Chain 1: 500 -9000.214 1.402 0.368
Chain 1: 600 -8433.559 1.180 0.368
Chain 1: 700 -8878.729 1.018 0.214
Chain 1: 800 -9031.476 0.893 0.214
Chain 1: 900 -8779.491 0.797 0.084
Chain 1: 1000 -8771.959 0.717 0.084
Chain 1: 1100 -8677.696 0.619 0.067 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8419.001 0.087 0.050
Chain 1: 1300 -8694.883 0.054 0.032
Chain 1: 1400 -8647.864 0.046 0.031
Chain 1: 1500 -8538.459 0.026 0.029
Chain 1: 1600 -8645.888 0.020 0.017
Chain 1: 1700 -8723.936 0.016 0.013
Chain 1: 1800 -8295.421 0.019 0.013
Chain 1: 1900 -8398.774 0.018 0.012
Chain 1: 2000 -8373.755 0.018 0.012
Chain 1: 2100 -8502.392 0.018 0.013
Chain 1: 2200 -8299.896 0.018 0.013
Chain 1: 2300 -8395.057 0.016 0.012
Chain 1: 2400 -8461.815 0.016 0.012
Chain 1: 2500 -8407.793 0.015 0.012
Chain 1: 2600 -8410.947 0.014 0.011
Chain 1: 2700 -8326.787 0.014 0.011
Chain 1: 2800 -8284.590 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003648 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 36.48 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8413272.620 1.000 1.000
Chain 1: 200 -1585062.464 2.654 4.308
Chain 1: 300 -890916.919 2.029 1.000
Chain 1: 400 -457672.006 1.758 1.000
Chain 1: 500 -358144.202 1.462 0.947
Chain 1: 600 -233123.666 1.308 0.947
Chain 1: 700 -119371.687 1.257 0.947
Chain 1: 800 -86607.768 1.147 0.947
Chain 1: 900 -66958.333 1.052 0.779
Chain 1: 1000 -51762.532 0.977 0.779
Chain 1: 1100 -39245.788 0.908 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38425.357 0.480 0.378
Chain 1: 1300 -26382.567 0.448 0.378
Chain 1: 1400 -26103.100 0.354 0.319
Chain 1: 1500 -22690.867 0.341 0.319
Chain 1: 1600 -21907.933 0.291 0.294
Chain 1: 1700 -20781.437 0.201 0.293
Chain 1: 1800 -20725.679 0.164 0.150
Chain 1: 1900 -21052.071 0.136 0.054
Chain 1: 2000 -19562.745 0.114 0.054
Chain 1: 2100 -19801.185 0.084 0.036
Chain 1: 2200 -20027.854 0.083 0.036
Chain 1: 2300 -19644.786 0.039 0.019
Chain 1: 2400 -19416.798 0.039 0.019
Chain 1: 2500 -19218.834 0.025 0.016
Chain 1: 2600 -18848.827 0.023 0.016
Chain 1: 2700 -18805.689 0.018 0.012
Chain 1: 2800 -18522.509 0.019 0.015
Chain 1: 2900 -18803.821 0.019 0.015
Chain 1: 3000 -18789.972 0.012 0.012
Chain 1: 3100 -18875.026 0.011 0.012
Chain 1: 3200 -18565.530 0.012 0.015
Chain 1: 3300 -18770.377 0.011 0.012
Chain 1: 3400 -18245.067 0.012 0.015
Chain 1: 3500 -18857.300 0.015 0.015
Chain 1: 3600 -18163.453 0.016 0.015
Chain 1: 3700 -18550.696 0.018 0.017
Chain 1: 3800 -17509.593 0.023 0.021
Chain 1: 3900 -17505.696 0.021 0.021
Chain 1: 4000 -17623.002 0.022 0.021
Chain 1: 4100 -17536.762 0.022 0.021
Chain 1: 4200 -17352.791 0.021 0.021
Chain 1: 4300 -17491.325 0.021 0.021
Chain 1: 4400 -17448.013 0.018 0.011
Chain 1: 4500 -17350.501 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001276 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.76 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12923.509 1.000 1.000
Chain 1: 200 -9422.015 0.686 1.000
Chain 1: 300 -8138.275 0.510 0.372
Chain 1: 400 -8117.421 0.383 0.372
Chain 1: 500 -8010.625 0.309 0.158
Chain 1: 600 -7885.819 0.260 0.158
Chain 1: 700 -7799.718 0.225 0.016
Chain 1: 800 -7827.179 0.197 0.016
Chain 1: 900 -7896.797 0.176 0.013
Chain 1: 1000 -7855.379 0.159 0.013
Chain 1: 1100 -7821.175 0.059 0.011
Chain 1: 1200 -7828.687 0.022 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001428 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.28 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58011.235 1.000 1.000
Chain 1: 200 -17693.450 1.639 2.279
Chain 1: 300 -8673.925 1.440 1.040
Chain 1: 400 -8046.253 1.099 1.040
Chain 1: 500 -8416.182 0.888 1.000
Chain 1: 600 -8246.127 0.744 1.000
Chain 1: 700 -8448.969 0.641 0.078
Chain 1: 800 -8258.044 0.564 0.078
Chain 1: 900 -7919.328 0.506 0.044
Chain 1: 1000 -7835.856 0.456 0.044
Chain 1: 1100 -7776.041 0.357 0.043
Chain 1: 1200 -7552.985 0.132 0.030
Chain 1: 1300 -7618.680 0.029 0.024
Chain 1: 1400 -7803.301 0.023 0.024
Chain 1: 1500 -7594.751 0.022 0.024
Chain 1: 1600 -7455.742 0.022 0.024
Chain 1: 1700 -7482.892 0.020 0.023
Chain 1: 1800 -7586.974 0.019 0.019
Chain 1: 1900 -7571.245 0.015 0.014
Chain 1: 2000 -7583.020 0.014 0.014
Chain 1: 2100 -7574.745 0.013 0.014
Chain 1: 2200 -7651.269 0.011 0.010
Chain 1: 2300 -7542.350 0.012 0.014
Chain 1: 2400 -7568.229 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003069 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.69 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86553.291 1.000 1.000
Chain 1: 200 -13513.082 3.203 5.405
Chain 1: 300 -9838.178 2.260 1.000
Chain 1: 400 -10998.814 1.721 1.000
Chain 1: 500 -8818.337 1.426 0.374
Chain 1: 600 -8423.092 1.196 0.374
Chain 1: 700 -8351.484 1.027 0.247
Chain 1: 800 -8780.207 0.904 0.247
Chain 1: 900 -8531.721 0.807 0.106
Chain 1: 1000 -8356.386 0.729 0.106
Chain 1: 1100 -8655.928 0.632 0.049 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8251.071 0.096 0.049
Chain 1: 1300 -8518.988 0.062 0.047
Chain 1: 1400 -8521.824 0.052 0.035
Chain 1: 1500 -8368.346 0.029 0.031
Chain 1: 1600 -8483.461 0.025 0.029
Chain 1: 1700 -8556.294 0.025 0.029
Chain 1: 1800 -8127.056 0.026 0.029
Chain 1: 1900 -8230.699 0.024 0.021
Chain 1: 2000 -8205.785 0.022 0.018
Chain 1: 2100 -8336.619 0.021 0.016
Chain 1: 2200 -8133.185 0.018 0.016
Chain 1: 2300 -8228.343 0.016 0.014
Chain 1: 2400 -8293.978 0.017 0.014
Chain 1: 2500 -8239.395 0.016 0.013
Chain 1: 2600 -8243.162 0.014 0.012
Chain 1: 2700 -8158.698 0.015 0.012
Chain 1: 2800 -8115.912 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003063 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.63 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8445341.535 1.000 1.000
Chain 1: 200 -1591479.709 2.653 4.307
Chain 1: 300 -890485.792 2.031 1.000
Chain 1: 400 -457211.791 1.760 1.000
Chain 1: 500 -356837.811 1.465 0.948
Chain 1: 600 -231930.755 1.310 0.948
Chain 1: 700 -118696.205 1.259 0.948
Chain 1: 800 -86023.045 1.149 0.948
Chain 1: 900 -66479.984 1.054 0.787
Chain 1: 1000 -51371.045 0.978 0.787
Chain 1: 1100 -38935.142 0.910 0.539 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38125.826 0.482 0.380
Chain 1: 1300 -26169.341 0.449 0.380
Chain 1: 1400 -25898.397 0.355 0.319
Chain 1: 1500 -22507.169 0.342 0.319
Chain 1: 1600 -21730.608 0.292 0.294
Chain 1: 1700 -20614.603 0.202 0.294
Chain 1: 1800 -20561.458 0.164 0.151
Chain 1: 1900 -20887.841 0.136 0.054
Chain 1: 2000 -19403.919 0.114 0.054
Chain 1: 2100 -19642.184 0.084 0.036
Chain 1: 2200 -19867.730 0.083 0.036
Chain 1: 2300 -19485.662 0.039 0.020
Chain 1: 2400 -19257.764 0.039 0.020
Chain 1: 2500 -19059.309 0.025 0.016
Chain 1: 2600 -18689.701 0.023 0.016
Chain 1: 2700 -18646.840 0.018 0.012
Chain 1: 2800 -18363.293 0.019 0.015
Chain 1: 2900 -18644.596 0.019 0.015
Chain 1: 3000 -18630.913 0.012 0.012
Chain 1: 3100 -18715.847 0.011 0.012
Chain 1: 3200 -18406.521 0.012 0.015
Chain 1: 3300 -18611.303 0.011 0.012
Chain 1: 3400 -18085.928 0.013 0.015
Chain 1: 3500 -18698.012 0.015 0.015
Chain 1: 3600 -18004.449 0.017 0.015
Chain 1: 3700 -18391.291 0.018 0.017
Chain 1: 3800 -17350.444 0.023 0.021
Chain 1: 3900 -17346.505 0.021 0.021
Chain 1: 4000 -17463.905 0.022 0.021
Chain 1: 4100 -17377.514 0.022 0.021
Chain 1: 4200 -17193.718 0.021 0.021
Chain 1: 4300 -17332.220 0.021 0.021
Chain 1: 4400 -17288.955 0.019 0.011
Chain 1: 4500 -17191.406 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001352 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.52 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48598.713 1.000 1.000
Chain 1: 200 -22137.911 1.098 1.195
Chain 1: 300 -19639.217 0.774 1.000
Chain 1: 400 -12603.005 0.720 1.000
Chain 1: 500 -23672.700 0.670 0.558
Chain 1: 600 -14010.176 0.673 0.690
Chain 1: 700 -15971.175 0.594 0.558
Chain 1: 800 -10369.372 0.588 0.558
Chain 1: 900 -12451.018 0.541 0.540
Chain 1: 1000 -17542.540 0.516 0.540
Chain 1: 1100 -16219.717 0.424 0.468
Chain 1: 1200 -12699.898 0.332 0.290
Chain 1: 1300 -11349.873 0.331 0.290
Chain 1: 1400 -10613.923 0.282 0.277
Chain 1: 1500 -10715.487 0.237 0.167
Chain 1: 1600 -11545.274 0.175 0.123
Chain 1: 1700 -9974.284 0.178 0.158
Chain 1: 1800 -13372.919 0.150 0.158
Chain 1: 1900 -10898.563 0.156 0.158
Chain 1: 2000 -9689.088 0.139 0.125
Chain 1: 2100 -9041.850 0.138 0.125
Chain 1: 2200 -10856.864 0.127 0.125
Chain 1: 2300 -9424.708 0.130 0.152
Chain 1: 2400 -10280.640 0.132 0.152
Chain 1: 2500 -10404.132 0.132 0.152
Chain 1: 2600 -8636.092 0.145 0.158
Chain 1: 2700 -10302.616 0.146 0.162
Chain 1: 2800 -20551.196 0.170 0.162
Chain 1: 2900 -9634.891 0.261 0.162
Chain 1: 3000 -10346.356 0.255 0.162
Chain 1: 3100 -15087.057 0.280 0.167
Chain 1: 3200 -12217.285 0.286 0.205
Chain 1: 3300 -9014.868 0.307 0.235
Chain 1: 3400 -8268.201 0.307 0.235
Chain 1: 3500 -9006.662 0.314 0.235
Chain 1: 3600 -8407.245 0.301 0.235
Chain 1: 3700 -8365.288 0.285 0.235
Chain 1: 3800 -12927.861 0.271 0.235
Chain 1: 3900 -9017.706 0.201 0.235
Chain 1: 4000 -8864.839 0.196 0.235
Chain 1: 4100 -8345.687 0.170 0.090
Chain 1: 4200 -10841.001 0.170 0.090
Chain 1: 4300 -9036.270 0.154 0.090
Chain 1: 4400 -9289.339 0.148 0.082
Chain 1: 4500 -8588.937 0.148 0.082
Chain 1: 4600 -13642.051 0.178 0.200
Chain 1: 4700 -11964.847 0.192 0.200
Chain 1: 4800 -8195.581 0.202 0.200
Chain 1: 4900 -8856.625 0.166 0.140
Chain 1: 5000 -14395.488 0.203 0.200
Chain 1: 5100 -9163.193 0.254 0.230
Chain 1: 5200 -8698.972 0.236 0.200
Chain 1: 5300 -8616.911 0.217 0.140
Chain 1: 5400 -13002.824 0.248 0.337
Chain 1: 5500 -8775.756 0.288 0.370
Chain 1: 5600 -8831.112 0.252 0.337
Chain 1: 5700 -8605.793 0.240 0.337
Chain 1: 5800 -10647.516 0.214 0.192
Chain 1: 5900 -11149.987 0.211 0.192
Chain 1: 6000 -9122.638 0.194 0.192
Chain 1: 6100 -7965.183 0.152 0.145
Chain 1: 6200 -7982.352 0.147 0.145
Chain 1: 6300 -8377.929 0.151 0.145
Chain 1: 6400 -9761.511 0.131 0.142
Chain 1: 6500 -12760.119 0.106 0.142
Chain 1: 6600 -9257.332 0.144 0.145
Chain 1: 6700 -9233.226 0.141 0.145
Chain 1: 6800 -10392.231 0.133 0.142
Chain 1: 6900 -11357.171 0.137 0.142
Chain 1: 7000 -8704.891 0.145 0.142
Chain 1: 7100 -7919.026 0.141 0.112
Chain 1: 7200 -10041.969 0.162 0.142
Chain 1: 7300 -10360.962 0.160 0.142
Chain 1: 7400 -7982.085 0.176 0.211
Chain 1: 7500 -8268.865 0.156 0.112
Chain 1: 7600 -10721.766 0.141 0.112
Chain 1: 7700 -10986.218 0.143 0.112
Chain 1: 7800 -10816.948 0.133 0.099
Chain 1: 7900 -10967.475 0.126 0.099
Chain 1: 8000 -9014.743 0.117 0.099
Chain 1: 8100 -7983.629 0.120 0.129
Chain 1: 8200 -8426.591 0.104 0.053
Chain 1: 8300 -8941.987 0.107 0.058
Chain 1: 8400 -8035.183 0.089 0.058
Chain 1: 8500 -8962.405 0.095 0.103
Chain 1: 8600 -7993.247 0.085 0.103
Chain 1: 8700 -7850.505 0.084 0.103
Chain 1: 8800 -8047.992 0.085 0.103
Chain 1: 8900 -10056.891 0.104 0.113
Chain 1: 9000 -7906.996 0.109 0.113
Chain 1: 9100 -7949.191 0.097 0.103
Chain 1: 9200 -8431.908 0.097 0.103
Chain 1: 9300 -9029.180 0.098 0.103
Chain 1: 9400 -9760.346 0.094 0.075
Chain 1: 9500 -9992.101 0.086 0.066
Chain 1: 9600 -8008.894 0.099 0.066
Chain 1: 9700 -10341.242 0.120 0.075
Chain 1: 9800 -10451.110 0.118 0.075
Chain 1: 9900 -8044.357 0.128 0.075
Chain 1: 10000 -7917.895 0.103 0.066
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00152 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.2 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -62548.905 1.000 1.000
Chain 1: 200 -17576.610 1.779 2.559
Chain 1: 300 -8511.061 1.541 1.065
Chain 1: 400 -8092.230 1.169 1.065
Chain 1: 500 -8195.637 0.938 1.000
Chain 1: 600 -8554.339 0.788 1.000
Chain 1: 700 -7870.444 0.688 0.087
Chain 1: 800 -7953.886 0.603 0.087
Chain 1: 900 -7747.193 0.539 0.052
Chain 1: 1000 -7673.945 0.486 0.052
Chain 1: 1100 -7633.467 0.387 0.042
Chain 1: 1200 -7530.334 0.132 0.027
Chain 1: 1300 -7692.782 0.028 0.021
Chain 1: 1400 -7734.035 0.023 0.014
Chain 1: 1500 -7583.096 0.024 0.020
Chain 1: 1600 -7491.706 0.021 0.014
Chain 1: 1700 -7463.765 0.013 0.012
Chain 1: 1800 -7493.440 0.012 0.012
Chain 1: 1900 -7560.319 0.010 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003199 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.99 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86211.509 1.000 1.000
Chain 1: 200 -12924.191 3.335 5.671
Chain 1: 300 -9417.940 2.348 1.000
Chain 1: 400 -10309.865 1.782 1.000
Chain 1: 500 -8255.802 1.476 0.372
Chain 1: 600 -8057.004 1.234 0.372
Chain 1: 700 -8307.018 1.062 0.249
Chain 1: 800 -8458.256 0.931 0.249
Chain 1: 900 -8317.268 0.830 0.087
Chain 1: 1000 -8046.649 0.750 0.087
Chain 1: 1100 -8249.729 0.653 0.034 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8042.644 0.088 0.030
Chain 1: 1300 -8205.486 0.053 0.026
Chain 1: 1400 -8131.850 0.045 0.025
Chain 1: 1500 -8080.978 0.021 0.025
Chain 1: 1600 -8078.491 0.018 0.020
Chain 1: 1700 -8021.486 0.016 0.018
Chain 1: 1800 -7900.751 0.016 0.017
Chain 1: 1900 -8012.889 0.016 0.015
Chain 1: 2000 -7975.844 0.013 0.014
Chain 1: 2100 -8120.480 0.012 0.014
Chain 1: 2200 -7901.882 0.012 0.014
Chain 1: 2300 -8032.195 0.012 0.014
Chain 1: 2400 -8048.437 0.011 0.014
Chain 1: 2500 -8014.206 0.011 0.014
Chain 1: 2600 -8007.551 0.011 0.014
Chain 1: 2700 -7919.222 0.011 0.014
Chain 1: 2800 -7905.227 0.010 0.011
Chain 1: 2900 -7904.004 0.009 0.005 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003483 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.83 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8428836.626 1.000 1.000
Chain 1: 200 -1588605.715 2.653 4.306
Chain 1: 300 -890999.529 2.030 1.000
Chain 1: 400 -457119.659 1.759 1.000
Chain 1: 500 -356954.308 1.464 0.949
Chain 1: 600 -231809.020 1.310 0.949
Chain 1: 700 -118286.407 1.260 0.949
Chain 1: 800 -85572.379 1.150 0.949
Chain 1: 900 -65966.940 1.055 0.783
Chain 1: 1000 -50802.575 0.980 0.783
Chain 1: 1100 -38328.272 0.912 0.540 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37501.555 0.484 0.382
Chain 1: 1300 -25522.012 0.452 0.382
Chain 1: 1400 -25242.793 0.359 0.325
Chain 1: 1500 -21848.080 0.346 0.325
Chain 1: 1600 -21068.759 0.296 0.298
Chain 1: 1700 -19951.077 0.205 0.297
Chain 1: 1800 -19896.686 0.167 0.155
Chain 1: 1900 -20221.963 0.139 0.056
Chain 1: 2000 -18739.364 0.117 0.056
Chain 1: 2100 -18977.226 0.086 0.037
Chain 1: 2200 -19202.484 0.085 0.037
Chain 1: 2300 -18820.988 0.040 0.020
Chain 1: 2400 -18593.505 0.040 0.020
Chain 1: 2500 -18395.332 0.026 0.016
Chain 1: 2600 -18026.631 0.024 0.016
Chain 1: 2700 -17983.922 0.019 0.013
Chain 1: 2800 -17701.136 0.020 0.016
Chain 1: 2900 -17981.858 0.020 0.016
Chain 1: 3000 -17968.128 0.012 0.013
Chain 1: 3100 -18052.992 0.011 0.012
Chain 1: 3200 -17744.315 0.012 0.016
Chain 1: 3300 -17948.529 0.011 0.012
Chain 1: 3400 -17424.556 0.013 0.016
Chain 1: 3500 -18034.699 0.015 0.016
Chain 1: 3600 -17343.594 0.017 0.016
Chain 1: 3700 -17728.728 0.019 0.017
Chain 1: 3800 -16691.851 0.024 0.022
Chain 1: 3900 -16688.052 0.022 0.022
Chain 1: 4000 -16805.374 0.023 0.022
Chain 1: 4100 -16719.326 0.023 0.022
Chain 1: 4200 -16536.297 0.022 0.022
Chain 1: 4300 -16674.188 0.022 0.022
Chain 1: 4400 -16631.618 0.019 0.011
Chain 1: 4500 -16534.249 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001365 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.65 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13003.081 1.000 1.000
Chain 1: 200 -9551.773 0.681 1.000
Chain 1: 300 -8123.345 0.512 0.361
Chain 1: 400 -8015.745 0.388 0.361
Chain 1: 500 -8102.902 0.312 0.176
Chain 1: 600 -7934.298 0.264 0.176
Chain 1: 700 -7851.892 0.228 0.021
Chain 1: 800 -7858.466 0.199 0.021
Chain 1: 900 -7781.625 0.178 0.013
Chain 1: 1000 -7962.859 0.163 0.021
Chain 1: 1100 -7989.045 0.063 0.013
Chain 1: 1200 -7886.552 0.028 0.013
Chain 1: 1300 -7820.778 0.011 0.011
Chain 1: 1400 -7844.303 0.010 0.010
Chain 1: 1500 -7931.402 0.010 0.010
Chain 1: 1600 -7894.550 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001535 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.35 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58046.399 1.000 1.000
Chain 1: 200 -17627.075 1.647 2.293
Chain 1: 300 -8675.121 1.442 1.032
Chain 1: 400 -8178.908 1.096 1.032
Chain 1: 500 -7804.937 0.887 1.000
Chain 1: 600 -8334.285 0.750 1.000
Chain 1: 700 -8075.543 0.647 0.064
Chain 1: 800 -8212.120 0.568 0.064
Chain 1: 900 -7874.151 0.510 0.061
Chain 1: 1000 -7556.753 0.463 0.061
Chain 1: 1100 -7791.390 0.366 0.048
Chain 1: 1200 -7829.782 0.137 0.043
Chain 1: 1300 -7687.014 0.036 0.042
Chain 1: 1400 -7794.025 0.031 0.032
Chain 1: 1500 -7617.741 0.029 0.030
Chain 1: 1600 -7747.758 0.024 0.023
Chain 1: 1700 -7530.238 0.024 0.023
Chain 1: 1800 -7614.598 0.023 0.023
Chain 1: 1900 -7610.228 0.019 0.019
Chain 1: 2000 -7591.783 0.015 0.017
Chain 1: 2100 -7586.739 0.012 0.014
Chain 1: 2200 -7693.508 0.013 0.014
Chain 1: 2300 -7596.940 0.012 0.014
Chain 1: 2400 -7641.293 0.012 0.013
Chain 1: 2500 -7472.228 0.012 0.013
Chain 1: 2600 -7517.537 0.010 0.011
Chain 1: 2700 -7557.781 0.008 0.006 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002698 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 26.98 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86577.131 1.000 1.000
Chain 1: 200 -13469.609 3.214 5.428
Chain 1: 300 -9832.317 2.266 1.000
Chain 1: 400 -10667.178 1.719 1.000
Chain 1: 500 -8795.294 1.418 0.370
Chain 1: 600 -8306.036 1.191 0.370
Chain 1: 700 -8409.653 1.023 0.213
Chain 1: 800 -8794.051 0.900 0.213
Chain 1: 900 -8650.510 0.802 0.078
Chain 1: 1000 -8360.606 0.725 0.078
Chain 1: 1100 -8585.955 0.628 0.059 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8448.677 0.087 0.044
Chain 1: 1300 -8546.512 0.051 0.035
Chain 1: 1400 -8555.643 0.043 0.026
Chain 1: 1500 -8388.792 0.024 0.020
Chain 1: 1600 -8509.196 0.020 0.017
Chain 1: 1700 -8592.184 0.019 0.017
Chain 1: 1800 -8178.737 0.020 0.017
Chain 1: 1900 -8274.623 0.020 0.016
Chain 1: 2000 -8248.050 0.016 0.014
Chain 1: 2100 -8370.827 0.015 0.014
Chain 1: 2200 -8190.872 0.016 0.014
Chain 1: 2300 -8269.627 0.016 0.014
Chain 1: 2400 -8339.330 0.016 0.014
Chain 1: 2500 -8284.775 0.015 0.012
Chain 1: 2600 -8284.270 0.014 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003622 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 36.22 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8441607.783 1.000 1.000
Chain 1: 200 -1587866.871 2.658 4.316
Chain 1: 300 -890554.004 2.033 1.000
Chain 1: 400 -457548.224 1.761 1.000
Chain 1: 500 -357304.051 1.465 0.946
Chain 1: 600 -232320.591 1.311 0.946
Chain 1: 700 -118849.203 1.260 0.946
Chain 1: 800 -86129.945 1.150 0.946
Chain 1: 900 -66538.253 1.055 0.783
Chain 1: 1000 -51391.814 0.979 0.783
Chain 1: 1100 -38925.030 0.911 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38107.280 0.481 0.380
Chain 1: 1300 -26121.732 0.449 0.380
Chain 1: 1400 -25846.610 0.355 0.320
Chain 1: 1500 -22448.589 0.342 0.320
Chain 1: 1600 -21669.636 0.292 0.295
Chain 1: 1700 -20550.291 0.202 0.294
Chain 1: 1800 -20496.081 0.164 0.151
Chain 1: 1900 -20822.203 0.137 0.054
Chain 1: 2000 -19337.051 0.115 0.054
Chain 1: 2100 -19575.250 0.084 0.036
Chain 1: 2200 -19801.058 0.083 0.036
Chain 1: 2300 -19418.838 0.039 0.020
Chain 1: 2400 -19190.999 0.039 0.020
Chain 1: 2500 -18992.744 0.025 0.016
Chain 1: 2600 -18623.224 0.023 0.016
Chain 1: 2700 -18580.338 0.018 0.012
Chain 1: 2800 -18297.073 0.020 0.015
Chain 1: 2900 -18578.221 0.020 0.015
Chain 1: 3000 -18564.516 0.012 0.012
Chain 1: 3100 -18649.461 0.011 0.012
Chain 1: 3200 -18340.250 0.012 0.015
Chain 1: 3300 -18544.894 0.011 0.012
Chain 1: 3400 -18019.878 0.013 0.015
Chain 1: 3500 -18631.549 0.015 0.015
Chain 1: 3600 -17938.478 0.017 0.015
Chain 1: 3700 -18325.027 0.019 0.017
Chain 1: 3800 -17285.044 0.023 0.021
Chain 1: 3900 -17281.146 0.022 0.021
Chain 1: 4000 -17398.507 0.022 0.021
Chain 1: 4100 -17312.243 0.022 0.021
Chain 1: 4200 -17128.569 0.022 0.021
Chain 1: 4300 -17266.947 0.021 0.021
Chain 1: 4400 -17223.836 0.019 0.011
Chain 1: 4500 -17126.326 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001569 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.69 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48609.787 1.000 1.000
Chain 1: 200 -16638.601 1.461 1.922
Chain 1: 300 -12834.803 1.073 1.000
Chain 1: 400 -22304.833 0.911 1.000
Chain 1: 500 -16110.708 0.805 0.425
Chain 1: 600 -11144.545 0.745 0.446
Chain 1: 700 -20557.343 0.704 0.446
Chain 1: 800 -22875.946 0.629 0.446
Chain 1: 900 -11487.879 0.669 0.446
Chain 1: 1000 -13049.753 0.614 0.446
Chain 1: 1100 -14349.805 0.523 0.425 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -14482.429 0.332 0.384
Chain 1: 1300 -10541.189 0.340 0.384
Chain 1: 1400 -15645.464 0.330 0.374
Chain 1: 1500 -9830.391 0.351 0.374
Chain 1: 1600 -10176.666 0.310 0.326
Chain 1: 1700 -10727.519 0.269 0.120
Chain 1: 1800 -12117.909 0.270 0.120
Chain 1: 1900 -9858.867 0.194 0.120
Chain 1: 2000 -9891.867 0.182 0.115
Chain 1: 2100 -9780.865 0.174 0.115
Chain 1: 2200 -11381.057 0.188 0.141
Chain 1: 2300 -9153.422 0.175 0.141
Chain 1: 2400 -9622.146 0.147 0.115
Chain 1: 2500 -21131.298 0.142 0.115
Chain 1: 2600 -9341.042 0.265 0.141
Chain 1: 2700 -9382.647 0.260 0.141
Chain 1: 2800 -9154.393 0.251 0.141
Chain 1: 2900 -10025.104 0.237 0.087
Chain 1: 3000 -9089.073 0.247 0.103
Chain 1: 3100 -9290.339 0.248 0.103
Chain 1: 3200 -9067.998 0.236 0.087
Chain 1: 3300 -12568.647 0.240 0.087
Chain 1: 3400 -12689.315 0.236 0.087
Chain 1: 3500 -14916.291 0.196 0.087
Chain 1: 3600 -9071.943 0.135 0.087
Chain 1: 3700 -8839.536 0.137 0.087
Chain 1: 3800 -8367.453 0.140 0.087
Chain 1: 3900 -9494.978 0.143 0.103
Chain 1: 4000 -8820.990 0.141 0.076
Chain 1: 4100 -9326.628 0.144 0.076
Chain 1: 4200 -9386.359 0.142 0.076
Chain 1: 4300 -9648.135 0.117 0.056
Chain 1: 4400 -14382.382 0.149 0.076
Chain 1: 4500 -8838.022 0.197 0.076
Chain 1: 4600 -8621.258 0.135 0.056
Chain 1: 4700 -9237.846 0.139 0.067
Chain 1: 4800 -10083.265 0.142 0.076
Chain 1: 4900 -9813.352 0.132 0.067
Chain 1: 5000 -14475.674 0.157 0.067
Chain 1: 5100 -15518.954 0.158 0.067
Chain 1: 5200 -14967.615 0.161 0.067
Chain 1: 5300 -8790.086 0.229 0.084
Chain 1: 5400 -9307.710 0.202 0.067
Chain 1: 5500 -12674.184 0.165 0.067
Chain 1: 5600 -12133.359 0.167 0.067
Chain 1: 5700 -12746.415 0.165 0.067
Chain 1: 5800 -8464.368 0.208 0.067
Chain 1: 5900 -10279.736 0.223 0.177
Chain 1: 6000 -10985.688 0.197 0.067
Chain 1: 6100 -11756.229 0.197 0.066
Chain 1: 6200 -8245.265 0.235 0.177
Chain 1: 6300 -8902.193 0.173 0.074
Chain 1: 6400 -9876.326 0.177 0.099
Chain 1: 6500 -11690.129 0.166 0.099
Chain 1: 6600 -13427.548 0.174 0.129
Chain 1: 6700 -8139.377 0.234 0.155
Chain 1: 6800 -8297.340 0.186 0.129
Chain 1: 6900 -11442.030 0.196 0.129
Chain 1: 7000 -8052.673 0.231 0.155
Chain 1: 7100 -9865.655 0.243 0.184
Chain 1: 7200 -8311.965 0.219 0.184
Chain 1: 7300 -8293.405 0.212 0.184
Chain 1: 7400 -8544.374 0.205 0.184
Chain 1: 7500 -7937.066 0.197 0.184
Chain 1: 7600 -10027.891 0.205 0.187
Chain 1: 7700 -8122.761 0.164 0.187
Chain 1: 7800 -9394.402 0.175 0.187
Chain 1: 7900 -8945.659 0.153 0.184
Chain 1: 8000 -8477.261 0.116 0.135
Chain 1: 8100 -8689.916 0.100 0.077
Chain 1: 8200 -8743.771 0.082 0.055
Chain 1: 8300 -12966.709 0.115 0.077
Chain 1: 8400 -8070.424 0.172 0.135
Chain 1: 8500 -9754.246 0.182 0.173
Chain 1: 8600 -9925.669 0.163 0.135
Chain 1: 8700 -8144.998 0.161 0.135
Chain 1: 8800 -8813.813 0.155 0.076
Chain 1: 8900 -11100.190 0.171 0.173
Chain 1: 9000 -7958.743 0.205 0.206
Chain 1: 9100 -8138.310 0.205 0.206
Chain 1: 9200 -9956.699 0.222 0.206
Chain 1: 9300 -8130.373 0.212 0.206
Chain 1: 9400 -8464.094 0.155 0.183
Chain 1: 9500 -12906.485 0.173 0.206
Chain 1: 9600 -9416.218 0.208 0.219
Chain 1: 9700 -9276.524 0.188 0.206
Chain 1: 9800 -7972.254 0.196 0.206
Chain 1: 9900 -9276.819 0.190 0.183
Chain 1: 10000 -7795.872 0.169 0.183
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001369 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.69 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56824.817 1.000 1.000
Chain 1: 200 -17290.024 1.643 2.287
Chain 1: 300 -8692.019 1.425 1.000
Chain 1: 400 -7944.848 1.092 1.000
Chain 1: 500 -8415.892 0.885 0.989
Chain 1: 600 -8853.932 0.746 0.989
Chain 1: 700 -7759.871 0.659 0.141
Chain 1: 800 -7708.245 0.578 0.141
Chain 1: 900 -7957.867 0.517 0.094
Chain 1: 1000 -7986.609 0.466 0.094
Chain 1: 1100 -7808.703 0.368 0.056
Chain 1: 1200 -7722.409 0.141 0.049
Chain 1: 1300 -7855.790 0.043 0.031
Chain 1: 1400 -7942.665 0.035 0.023
Chain 1: 1500 -7662.694 0.033 0.023
Chain 1: 1600 -7672.608 0.028 0.017
Chain 1: 1700 -7577.821 0.015 0.013
Chain 1: 1800 -7645.091 0.016 0.013
Chain 1: 1900 -7652.620 0.013 0.011
Chain 1: 2000 -7657.388 0.012 0.011
Chain 1: 2100 -7659.175 0.010 0.011
Chain 1: 2200 -7749.190 0.010 0.011
Chain 1: 2300 -7662.869 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003282 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.82 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86195.606 1.000 1.000
Chain 1: 200 -13309.060 3.238 5.476
Chain 1: 300 -9665.647 2.284 1.000
Chain 1: 400 -10567.381 1.735 1.000
Chain 1: 500 -8498.924 1.436 0.377
Chain 1: 600 -8113.606 1.205 0.377
Chain 1: 700 -8193.245 1.034 0.243
Chain 1: 800 -8546.353 0.910 0.243
Chain 1: 900 -8473.626 0.810 0.085
Chain 1: 1000 -8113.190 0.733 0.085
Chain 1: 1100 -8333.867 0.636 0.047 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8146.618 0.091 0.044
Chain 1: 1300 -8350.452 0.055 0.041
Chain 1: 1400 -8359.281 0.047 0.026
Chain 1: 1500 -8208.726 0.024 0.024
Chain 1: 1600 -8323.313 0.021 0.023
Chain 1: 1700 -8401.711 0.021 0.023
Chain 1: 1800 -7980.863 0.022 0.023
Chain 1: 1900 -8080.763 0.023 0.023
Chain 1: 2000 -8054.883 0.018 0.018
Chain 1: 2100 -8179.710 0.017 0.015
Chain 1: 2200 -7987.142 0.017 0.015
Chain 1: 2300 -8075.415 0.016 0.014
Chain 1: 2400 -8144.568 0.017 0.014
Chain 1: 2500 -8090.646 0.016 0.012
Chain 1: 2600 -8091.466 0.014 0.011
Chain 1: 2700 -8008.475 0.014 0.011
Chain 1: 2800 -7969.150 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003358 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.58 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8385613.955 1.000 1.000
Chain 1: 200 -1580743.458 2.652 4.305
Chain 1: 300 -889944.209 2.027 1.000
Chain 1: 400 -457364.053 1.757 1.000
Chain 1: 500 -358181.146 1.461 0.946
Chain 1: 600 -233147.187 1.307 0.946
Chain 1: 700 -119241.162 1.256 0.946
Chain 1: 800 -86421.480 1.147 0.946
Chain 1: 900 -66725.141 1.052 0.776
Chain 1: 1000 -51490.100 0.977 0.776
Chain 1: 1100 -38936.346 0.909 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38108.217 0.481 0.380
Chain 1: 1300 -26033.070 0.449 0.380
Chain 1: 1400 -25748.789 0.356 0.322
Chain 1: 1500 -22328.314 0.343 0.322
Chain 1: 1600 -21542.557 0.293 0.296
Chain 1: 1700 -20412.538 0.203 0.295
Chain 1: 1800 -20355.852 0.166 0.153
Chain 1: 1900 -20682.082 0.138 0.055
Chain 1: 2000 -19191.127 0.116 0.055
Chain 1: 2100 -19429.579 0.085 0.036
Chain 1: 2200 -19656.478 0.084 0.036
Chain 1: 2300 -19273.266 0.040 0.020
Chain 1: 2400 -19045.295 0.040 0.020
Chain 1: 2500 -18847.455 0.025 0.016
Chain 1: 2600 -18477.460 0.024 0.016
Chain 1: 2700 -18434.320 0.018 0.012
Chain 1: 2800 -18151.243 0.020 0.016
Chain 1: 2900 -18432.568 0.020 0.015
Chain 1: 3000 -18418.670 0.012 0.012
Chain 1: 3100 -18503.704 0.011 0.012
Chain 1: 3200 -18194.309 0.012 0.015
Chain 1: 3300 -18399.081 0.011 0.012
Chain 1: 3400 -17873.930 0.013 0.015
Chain 1: 3500 -18486.001 0.015 0.016
Chain 1: 3600 -17792.424 0.017 0.016
Chain 1: 3700 -18179.473 0.019 0.017
Chain 1: 3800 -17138.824 0.023 0.021
Chain 1: 3900 -17134.978 0.022 0.021
Chain 1: 4000 -17252.255 0.022 0.021
Chain 1: 4100 -17166.042 0.022 0.021
Chain 1: 4200 -16982.165 0.022 0.021
Chain 1: 4300 -17120.629 0.021 0.021
Chain 1: 4400 -17077.399 0.019 0.011
Chain 1: 4500 -16979.918 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001424 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.24 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -50156.626 1.000 1.000
Chain 1: 200 -20623.576 1.216 1.432
Chain 1: 300 -20446.267 0.814 1.000
Chain 1: 400 -14256.231 0.719 1.000
Chain 1: 500 -25775.117 0.664 0.447
Chain 1: 600 -39064.774 0.610 0.447
Chain 1: 700 -22376.992 0.630 0.447
Chain 1: 800 -17038.119 0.590 0.447
Chain 1: 900 -14566.272 0.543 0.434
Chain 1: 1000 -11686.895 0.514 0.434
Chain 1: 1100 -11630.789 0.414 0.340
Chain 1: 1200 -11568.919 0.272 0.313
Chain 1: 1300 -14212.204 0.289 0.313
Chain 1: 1400 -19822.647 0.274 0.283
Chain 1: 1500 -24168.419 0.247 0.246
Chain 1: 1600 -20667.840 0.230 0.186
Chain 1: 1700 -12436.412 0.222 0.186
Chain 1: 1800 -10449.424 0.210 0.186
Chain 1: 1900 -11581.778 0.202 0.186
Chain 1: 2000 -14825.858 0.200 0.186
Chain 1: 2100 -10815.138 0.236 0.190
Chain 1: 2200 -17897.787 0.275 0.219
Chain 1: 2300 -10914.722 0.321 0.283
Chain 1: 2400 -20601.942 0.339 0.371
Chain 1: 2500 -10742.974 0.413 0.396
Chain 1: 2600 -10428.232 0.399 0.396
Chain 1: 2700 -14196.387 0.360 0.371
Chain 1: 2800 -11897.787 0.360 0.371
Chain 1: 2900 -11379.694 0.355 0.371
Chain 1: 3000 -10024.339 0.346 0.371
Chain 1: 3100 -10475.985 0.314 0.265
Chain 1: 3200 -12809.363 0.292 0.193
Chain 1: 3300 -10252.448 0.253 0.193
Chain 1: 3400 -17108.871 0.246 0.193
Chain 1: 3500 -12755.069 0.189 0.193
Chain 1: 3600 -18366.739 0.216 0.249
Chain 1: 3700 -9982.947 0.274 0.249
Chain 1: 3800 -10680.248 0.261 0.249
Chain 1: 3900 -17135.565 0.294 0.306
Chain 1: 4000 -9666.059 0.358 0.341
Chain 1: 4100 -10503.159 0.361 0.341
Chain 1: 4200 -9694.156 0.351 0.341
Chain 1: 4300 -18421.503 0.374 0.377
Chain 1: 4400 -9651.663 0.425 0.377
Chain 1: 4500 -10002.082 0.394 0.377
Chain 1: 4600 -9733.404 0.366 0.377
Chain 1: 4700 -9590.988 0.284 0.083
Chain 1: 4800 -9564.705 0.278 0.083
Chain 1: 4900 -10565.922 0.249 0.083
Chain 1: 5000 -20042.544 0.219 0.083
Chain 1: 5100 -9661.725 0.319 0.095
Chain 1: 5200 -9752.593 0.311 0.095
Chain 1: 5300 -9505.910 0.267 0.035
Chain 1: 5400 -10508.336 0.185 0.035
Chain 1: 5500 -14563.111 0.210 0.095
Chain 1: 5600 -10623.319 0.244 0.095
Chain 1: 5700 -9958.609 0.249 0.095
Chain 1: 5800 -9624.477 0.252 0.095
Chain 1: 5900 -16365.680 0.284 0.278
Chain 1: 6000 -10370.301 0.295 0.278
Chain 1: 6100 -13545.937 0.211 0.234
Chain 1: 6200 -11622.125 0.226 0.234
Chain 1: 6300 -13469.127 0.237 0.234
Chain 1: 6400 -13336.009 0.229 0.234
Chain 1: 6500 -14493.667 0.209 0.166
Chain 1: 6600 -9479.132 0.225 0.166
Chain 1: 6700 -13775.359 0.249 0.234
Chain 1: 6800 -10737.281 0.274 0.283
Chain 1: 6900 -14259.800 0.258 0.247
Chain 1: 7000 -15080.251 0.205 0.234
Chain 1: 7100 -9086.626 0.248 0.247
Chain 1: 7200 -10512.089 0.245 0.247
Chain 1: 7300 -13050.142 0.250 0.247
Chain 1: 7400 -13348.466 0.252 0.247
Chain 1: 7500 -12529.636 0.250 0.247
Chain 1: 7600 -10843.653 0.213 0.194
Chain 1: 7700 -10876.143 0.182 0.155
Chain 1: 7800 -9306.219 0.171 0.155
Chain 1: 7900 -9442.879 0.147 0.136
Chain 1: 8000 -9397.703 0.142 0.136
Chain 1: 8100 -9806.673 0.081 0.065
Chain 1: 8200 -9408.705 0.071 0.042
Chain 1: 8300 -9270.536 0.053 0.042
Chain 1: 8400 -9543.114 0.054 0.042
Chain 1: 8500 -9583.621 0.048 0.029
Chain 1: 8600 -9675.511 0.033 0.015
Chain 1: 8700 -10580.201 0.041 0.029
Chain 1: 8800 -9112.591 0.041 0.029
Chain 1: 8900 -12807.811 0.068 0.042
Chain 1: 9000 -10506.124 0.090 0.042
Chain 1: 9100 -9345.980 0.098 0.086
Chain 1: 9200 -10977.800 0.108 0.124
Chain 1: 9300 -9744.882 0.120 0.127
Chain 1: 9400 -9294.171 0.122 0.127
Chain 1: 9500 -9684.785 0.125 0.127
Chain 1: 9600 -9928.070 0.127 0.127
Chain 1: 9700 -9151.264 0.127 0.127
Chain 1: 9800 -13537.652 0.143 0.127
Chain 1: 9900 -12865.125 0.119 0.124
Chain 1: 10000 -13630.251 0.103 0.085
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001661 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 16.61 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -47737.995 1.000 1.000
Chain 1: 200 -17025.498 1.402 1.804
Chain 1: 300 -9523.009 1.197 1.000
Chain 1: 400 -8502.376 0.928 1.000
Chain 1: 500 -8686.672 0.747 0.788
Chain 1: 600 -9958.746 0.643 0.788
Chain 1: 700 -9025.139 0.566 0.128
Chain 1: 800 -8652.260 0.501 0.128
Chain 1: 900 -8393.978 0.449 0.120
Chain 1: 1000 -8287.176 0.405 0.120
Chain 1: 1100 -8178.355 0.306 0.103
Chain 1: 1200 -8131.879 0.127 0.043
Chain 1: 1300 -7896.440 0.051 0.031
Chain 1: 1400 -7719.928 0.041 0.030
Chain 1: 1500 -7975.931 0.042 0.031
Chain 1: 1600 -8276.827 0.033 0.031
Chain 1: 1700 -8024.821 0.026 0.031
Chain 1: 1800 -7832.679 0.024 0.030
Chain 1: 1900 -7855.353 0.021 0.025
Chain 1: 2000 -7798.723 0.021 0.025
Chain 1: 2100 -7839.209 0.020 0.025
Chain 1: 2200 -8047.355 0.022 0.026
Chain 1: 2300 -7710.887 0.023 0.026
Chain 1: 2400 -7783.273 0.022 0.026
Chain 1: 2500 -7783.857 0.019 0.025
Chain 1: 2600 -7719.760 0.016 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003177 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.77 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86498.313 1.000 1.000
Chain 1: 200 -14930.478 2.897 4.793
Chain 1: 300 -11038.091 2.049 1.000
Chain 1: 400 -13585.644 1.583 1.000
Chain 1: 500 -9350.306 1.357 0.453
Chain 1: 600 -9472.668 1.133 0.453
Chain 1: 700 -9435.707 0.972 0.353
Chain 1: 800 -9116.899 0.855 0.353
Chain 1: 900 -9187.583 0.761 0.188
Chain 1: 1000 -9970.139 0.692 0.188
Chain 1: 1100 -9437.759 0.598 0.078 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -10041.083 0.125 0.060
Chain 1: 1300 -9209.622 0.099 0.060
Chain 1: 1400 -9369.657 0.081 0.056
Chain 1: 1500 -9321.557 0.037 0.035
Chain 1: 1600 -9279.102 0.036 0.035
Chain 1: 1700 -9146.112 0.037 0.035
Chain 1: 1800 -9184.306 0.034 0.017
Chain 1: 1900 -9197.942 0.033 0.017
Chain 1: 2000 -9362.908 0.027 0.017
Chain 1: 2100 -9205.747 0.023 0.017
Chain 1: 2200 -9126.785 0.018 0.015
Chain 1: 2300 -9330.446 0.011 0.015
Chain 1: 2400 -9063.013 0.012 0.015
Chain 1: 2500 -9147.767 0.013 0.015
Chain 1: 2600 -9050.341 0.013 0.015
Chain 1: 2700 -9068.155 0.012 0.011
Chain 1: 2800 -8926.609 0.013 0.016
Chain 1: 2900 -9115.484 0.015 0.017
Chain 1: 3000 -9023.766 0.015 0.016
Chain 1: 3100 -9121.979 0.014 0.011
Chain 1: 3200 -8988.354 0.015 0.015
Chain 1: 3300 -9257.634 0.015 0.015
Chain 1: 3400 -9327.626 0.013 0.011
Chain 1: 3500 -9137.137 0.014 0.015
Chain 1: 3600 -8940.749 0.015 0.016
Chain 1: 3700 -9110.098 0.017 0.019
Chain 1: 3800 -8944.682 0.017 0.019
Chain 1: 3900 -9170.364 0.018 0.019
Chain 1: 4000 -9171.632 0.017 0.019
Chain 1: 4100 -8955.642 0.018 0.021
Chain 1: 4200 -8940.643 0.017 0.021
Chain 1: 4300 -8942.060 0.014 0.019
Chain 1: 4400 -8896.113 0.014 0.019
Chain 1: 4500 -9036.799 0.013 0.018
Chain 1: 4600 -9063.878 0.011 0.016
Chain 1: 4700 -9181.772 0.011 0.013
Chain 1: 4800 -9004.560 0.011 0.013
Chain 1: 4900 -9030.055 0.009 0.005 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003136 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.36 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8357465.244 1.000 1.000
Chain 1: 200 -1579099.165 2.646 4.293
Chain 1: 300 -892530.956 2.021 1.000
Chain 1: 400 -459610.694 1.751 1.000
Chain 1: 500 -360639.332 1.456 0.942
Chain 1: 600 -235724.185 1.301 0.942
Chain 1: 700 -121390.273 1.250 0.942
Chain 1: 800 -88430.998 1.140 0.942
Chain 1: 900 -68671.463 1.046 0.769
Chain 1: 1000 -53387.315 0.970 0.769
Chain 1: 1100 -40766.719 0.901 0.530 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39944.112 0.473 0.373
Chain 1: 1300 -27780.305 0.440 0.373
Chain 1: 1400 -27494.374 0.347 0.310
Chain 1: 1500 -24048.378 0.334 0.310
Chain 1: 1600 -23256.501 0.284 0.288
Chain 1: 1700 -22114.702 0.195 0.286
Chain 1: 1800 -22055.988 0.158 0.143
Chain 1: 1900 -22383.396 0.131 0.052
Chain 1: 2000 -20883.549 0.110 0.052
Chain 1: 2100 -21122.813 0.080 0.034
Chain 1: 2200 -21351.381 0.079 0.034
Chain 1: 2300 -20966.351 0.037 0.018
Chain 1: 2400 -20737.794 0.037 0.018
Chain 1: 2500 -20540.154 0.024 0.015
Chain 1: 2600 -20168.740 0.022 0.015
Chain 1: 2700 -20125.125 0.017 0.011
Chain 1: 2800 -19841.545 0.018 0.014
Chain 1: 2900 -20123.558 0.018 0.014
Chain 1: 3000 -20109.660 0.011 0.011
Chain 1: 3100 -20194.844 0.010 0.011
Chain 1: 3200 -19884.556 0.011 0.014
Chain 1: 3300 -20089.997 0.010 0.011
Chain 1: 3400 -19563.312 0.012 0.014
Chain 1: 3500 -20177.796 0.014 0.014
Chain 1: 3600 -19481.131 0.015 0.014
Chain 1: 3700 -19870.507 0.017 0.016
Chain 1: 3800 -18825.098 0.021 0.020
Chain 1: 3900 -18821.143 0.020 0.020
Chain 1: 4000 -18938.415 0.020 0.020
Chain 1: 4100 -18851.959 0.021 0.020
Chain 1: 4200 -18667.040 0.020 0.020
Chain 1: 4300 -18806.234 0.020 0.020
Chain 1: 4400 -18762.144 0.017 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001342 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.42 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12109.271 1.000 1.000
Chain 1: 200 -8987.153 0.674 1.000
Chain 1: 300 -7849.894 0.497 0.347
Chain 1: 400 -7949.971 0.376 0.347
Chain 1: 500 -7833.614 0.304 0.145
Chain 1: 600 -7781.302 0.254 0.145
Chain 1: 700 -7699.248 0.220 0.015
Chain 1: 800 -7747.869 0.193 0.015
Chain 1: 900 -7775.304 0.172 0.013
Chain 1: 1000 -7726.695 0.155 0.013
Chain 1: 1100 -7807.330 0.056 0.011
Chain 1: 1200 -7720.352 0.023 0.011
Chain 1: 1300 -7669.182 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001499 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.99 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56897.100 1.000 1.000
Chain 1: 200 -17138.248 1.660 2.320
Chain 1: 300 -8613.149 1.437 1.000
Chain 1: 400 -7890.247 1.100 1.000
Chain 1: 500 -8516.622 0.895 0.990
Chain 1: 600 -8645.504 0.748 0.990
Chain 1: 700 -8312.196 0.647 0.092
Chain 1: 800 -8129.920 0.569 0.092
Chain 1: 900 -7835.696 0.510 0.074
Chain 1: 1000 -7746.724 0.460 0.074
Chain 1: 1100 -7721.445 0.360 0.040
Chain 1: 1200 -7623.913 0.130 0.038
Chain 1: 1300 -7720.145 0.032 0.022
Chain 1: 1400 -7666.718 0.024 0.015
Chain 1: 1500 -7599.219 0.017 0.013
Chain 1: 1600 -7547.073 0.016 0.012
Chain 1: 1700 -7535.487 0.012 0.011
Chain 1: 1800 -7569.318 0.011 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003778 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 37.78 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85735.434 1.000 1.000
Chain 1: 200 -13183.907 3.252 5.503
Chain 1: 300 -9613.720 2.291 1.000
Chain 1: 400 -10652.169 1.743 1.000
Chain 1: 500 -8562.810 1.443 0.371
Chain 1: 600 -8110.130 1.212 0.371
Chain 1: 700 -8323.231 1.042 0.244
Chain 1: 800 -8844.904 0.920 0.244
Chain 1: 900 -8426.742 0.823 0.097
Chain 1: 1000 -8156.265 0.744 0.097
Chain 1: 1100 -8491.110 0.648 0.059 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8046.431 0.103 0.056
Chain 1: 1300 -8328.110 0.069 0.055
Chain 1: 1400 -8322.528 0.060 0.050
Chain 1: 1500 -8230.482 0.036 0.039
Chain 1: 1600 -8332.982 0.032 0.034
Chain 1: 1700 -8416.560 0.030 0.034
Chain 1: 1800 -8022.252 0.029 0.034
Chain 1: 1900 -8123.583 0.026 0.033
Chain 1: 2000 -8094.291 0.023 0.012
Chain 1: 2100 -8216.619 0.020 0.012
Chain 1: 2200 -7997.561 0.018 0.012
Chain 1: 2300 -8152.426 0.016 0.012
Chain 1: 2400 -8166.189 0.016 0.012
Chain 1: 2500 -8135.765 0.015 0.012
Chain 1: 2600 -8138.420 0.014 0.012
Chain 1: 2700 -8044.649 0.014 0.012
Chain 1: 2800 -8015.710 0.010 0.012 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00362 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 36.2 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8419483.212 1.000 1.000
Chain 1: 200 -1587300.996 2.652 4.304
Chain 1: 300 -890764.816 2.029 1.000
Chain 1: 400 -457320.986 1.759 1.000
Chain 1: 500 -357422.703 1.463 0.948
Chain 1: 600 -232248.418 1.309 0.948
Chain 1: 700 -118636.365 1.259 0.948
Chain 1: 800 -85956.684 1.149 0.948
Chain 1: 900 -66333.908 1.054 0.782
Chain 1: 1000 -51166.134 0.978 0.782
Chain 1: 1100 -38681.326 0.911 0.539 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37857.799 0.482 0.380
Chain 1: 1300 -25849.281 0.451 0.380
Chain 1: 1400 -25570.556 0.357 0.323
Chain 1: 1500 -22168.575 0.344 0.323
Chain 1: 1600 -21388.382 0.294 0.296
Chain 1: 1700 -20266.061 0.204 0.296
Chain 1: 1800 -20211.093 0.166 0.153
Chain 1: 1900 -20536.954 0.138 0.055
Chain 1: 2000 -19051.116 0.116 0.055
Chain 1: 2100 -19289.092 0.085 0.036
Chain 1: 2200 -19515.267 0.084 0.036
Chain 1: 2300 -19132.819 0.040 0.020
Chain 1: 2400 -18905.064 0.040 0.020
Chain 1: 2500 -18707.193 0.025 0.016
Chain 1: 2600 -18337.656 0.024 0.016
Chain 1: 2700 -18294.669 0.019 0.012
Chain 1: 2800 -18011.785 0.020 0.016
Chain 1: 2900 -18292.769 0.020 0.015
Chain 1: 3000 -18278.928 0.012 0.012
Chain 1: 3100 -18363.945 0.011 0.012
Chain 1: 3200 -18054.803 0.012 0.015
Chain 1: 3300 -18259.369 0.011 0.012
Chain 1: 3400 -17734.739 0.013 0.015
Chain 1: 3500 -18345.985 0.015 0.016
Chain 1: 3600 -17653.389 0.017 0.016
Chain 1: 3700 -18039.671 0.019 0.017
Chain 1: 3800 -17000.617 0.023 0.021
Chain 1: 3900 -16996.802 0.022 0.021
Chain 1: 4000 -17114.079 0.022 0.021
Chain 1: 4100 -17027.977 0.023 0.021
Chain 1: 4200 -16844.452 0.022 0.021
Chain 1: 4300 -16982.657 0.022 0.021
Chain 1: 4400 -16939.673 0.019 0.011
Chain 1: 4500 -16842.266 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001401 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.01 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13033.937 1.000 1.000
Chain 1: 200 -9903.118 0.658 1.000
Chain 1: 300 -8523.378 0.493 0.316
Chain 1: 400 -8740.476 0.376 0.316
Chain 1: 500 -8611.761 0.304 0.162
Chain 1: 600 -8460.335 0.256 0.162
Chain 1: 700 -8569.367 0.221 0.025
Chain 1: 800 -8397.399 0.196 0.025
Chain 1: 900 -8432.743 0.175 0.020
Chain 1: 1000 -8399.440 0.158 0.020
Chain 1: 1100 -8502.836 0.059 0.018
Chain 1: 1200 -8383.218 0.029 0.015
Chain 1: 1300 -8326.962 0.013 0.014
Chain 1: 1400 -8342.787 0.011 0.013
Chain 1: 1500 -8440.060 0.011 0.012
Chain 1: 1600 -8354.453 0.010 0.012 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001543 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.43 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -50095.123 1.000 1.000
Chain 1: 200 -16701.629 1.500 1.999
Chain 1: 300 -9086.907 1.279 1.000
Chain 1: 400 -8395.251 0.980 1.000
Chain 1: 500 -8053.182 0.792 0.838
Chain 1: 600 -9441.456 0.685 0.838
Chain 1: 700 -8816.927 0.597 0.147
Chain 1: 800 -8514.499 0.527 0.147
Chain 1: 900 -7967.443 0.476 0.082
Chain 1: 1000 -8017.392 0.429 0.082
Chain 1: 1100 -8075.678 0.330 0.071
Chain 1: 1200 -7768.198 0.134 0.069
Chain 1: 1300 -7837.878 0.051 0.042
Chain 1: 1400 -7735.216 0.044 0.040
Chain 1: 1500 -7605.500 0.041 0.036
Chain 1: 1600 -7793.445 0.029 0.024
Chain 1: 1700 -7544.726 0.025 0.024
Chain 1: 1800 -7707.575 0.024 0.021
Chain 1: 1900 -7815.924 0.018 0.017
Chain 1: 2000 -7827.742 0.018 0.017
Chain 1: 2100 -7667.497 0.019 0.021
Chain 1: 2200 -7926.517 0.019 0.021
Chain 1: 2300 -7708.554 0.021 0.021
Chain 1: 2400 -7701.917 0.019 0.021
Chain 1: 2500 -7639.577 0.018 0.021
Chain 1: 2600 -7628.079 0.016 0.021
Chain 1: 2700 -7527.349 0.014 0.014
Chain 1: 2800 -7755.428 0.015 0.014
Chain 1: 2900 -7451.684 0.018 0.021
Chain 1: 3000 -7607.260 0.020 0.021
Chain 1: 3100 -7607.547 0.018 0.020
Chain 1: 3200 -7806.426 0.017 0.020
Chain 1: 3300 -7505.113 0.018 0.020
Chain 1: 3400 -7740.315 0.021 0.025
Chain 1: 3500 -7513.601 0.023 0.029
Chain 1: 3600 -7583.713 0.024 0.029
Chain 1: 3700 -7538.332 0.023 0.029
Chain 1: 3800 -7509.938 0.021 0.025
Chain 1: 3900 -7483.902 0.017 0.020
Chain 1: 4000 -7480.191 0.015 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003235 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.35 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86862.588 1.000 1.000
Chain 1: 200 -14241.374 3.050 5.099
Chain 1: 300 -10486.226 2.152 1.000
Chain 1: 400 -12113.465 1.648 1.000
Chain 1: 500 -9200.270 1.382 0.358
Chain 1: 600 -9865.559 1.163 0.358
Chain 1: 700 -9013.535 1.010 0.317
Chain 1: 800 -9743.317 0.893 0.317
Chain 1: 900 -9188.648 0.801 0.134
Chain 1: 1000 -9183.724 0.721 0.134
Chain 1: 1100 -9237.761 0.621 0.095 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8791.062 0.116 0.075
Chain 1: 1300 -9110.845 0.084 0.067
Chain 1: 1400 -9138.963 0.071 0.060
Chain 1: 1500 -8972.261 0.041 0.051
Chain 1: 1600 -9086.342 0.036 0.035
Chain 1: 1700 -9139.815 0.027 0.019
Chain 1: 1800 -8691.127 0.024 0.019
Chain 1: 1900 -8800.323 0.020 0.013
Chain 1: 2000 -8783.542 0.020 0.013
Chain 1: 2100 -8923.038 0.021 0.016
Chain 1: 2200 -8695.357 0.018 0.016
Chain 1: 2300 -8795.167 0.016 0.013
Chain 1: 2400 -8867.768 0.016 0.013
Chain 1: 2500 -8807.498 0.015 0.012
Chain 1: 2600 -8824.512 0.014 0.011
Chain 1: 2700 -8730.671 0.015 0.011
Chain 1: 2800 -8676.456 0.010 0.011
Chain 1: 2900 -8782.446 0.010 0.011
Chain 1: 3000 -8620.552 0.012 0.011
Chain 1: 3100 -8760.769 0.012 0.011
Chain 1: 3200 -8630.145 0.011 0.011
Chain 1: 3300 -8858.414 0.012 0.012
Chain 1: 3400 -8875.570 0.012 0.012
Chain 1: 3500 -8732.533 0.012 0.015
Chain 1: 3600 -8587.422 0.014 0.016
Chain 1: 3700 -8734.445 0.015 0.016
Chain 1: 3800 -8590.135 0.016 0.017
Chain 1: 3900 -8522.017 0.015 0.017
Chain 1: 4000 -8631.379 0.015 0.016
Chain 1: 4100 -8597.184 0.013 0.016
Chain 1: 4200 -8582.990 0.012 0.016
Chain 1: 4300 -8616.442 0.010 0.013 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003439 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.39 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8387364.357 1.000 1.000
Chain 1: 200 -1585177.612 2.646 4.291
Chain 1: 300 -892690.087 2.022 1.000
Chain 1: 400 -458823.593 1.753 1.000
Chain 1: 500 -359373.394 1.458 0.946
Chain 1: 600 -234109.737 1.304 0.946
Chain 1: 700 -120169.777 1.253 0.946
Chain 1: 800 -87310.549 1.144 0.946
Chain 1: 900 -67635.438 1.049 0.776
Chain 1: 1000 -52423.894 0.973 0.776
Chain 1: 1100 -39877.242 0.904 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39059.312 0.477 0.376
Chain 1: 1300 -26983.957 0.445 0.376
Chain 1: 1400 -26702.967 0.351 0.315
Chain 1: 1500 -23281.105 0.338 0.315
Chain 1: 1600 -22495.393 0.288 0.291
Chain 1: 1700 -21365.133 0.199 0.290
Chain 1: 1800 -21308.702 0.161 0.147
Chain 1: 1900 -21635.447 0.134 0.053
Chain 1: 2000 -20143.148 0.112 0.053
Chain 1: 2100 -20381.840 0.082 0.035
Chain 1: 2200 -20608.990 0.081 0.035
Chain 1: 2300 -20225.444 0.038 0.019
Chain 1: 2400 -19997.292 0.038 0.019
Chain 1: 2500 -19799.294 0.024 0.015
Chain 1: 2600 -19428.885 0.023 0.015
Chain 1: 2700 -19385.676 0.018 0.012
Chain 1: 2800 -19102.226 0.019 0.015
Chain 1: 2900 -19383.845 0.019 0.015
Chain 1: 3000 -19369.998 0.011 0.012
Chain 1: 3100 -19455.047 0.011 0.011
Chain 1: 3200 -19145.325 0.011 0.015
Chain 1: 3300 -19350.378 0.010 0.011
Chain 1: 3400 -18824.533 0.012 0.015
Chain 1: 3500 -19437.546 0.014 0.015
Chain 1: 3600 -18742.839 0.016 0.015
Chain 1: 3700 -19130.679 0.018 0.016
Chain 1: 3800 -18088.138 0.022 0.020
Chain 1: 3900 -18084.235 0.021 0.020
Chain 1: 4000 -18201.549 0.021 0.020
Chain 1: 4100 -18115.170 0.021 0.020
Chain 1: 4200 -17930.946 0.021 0.020
Chain 1: 4300 -18069.673 0.020 0.020
Chain 1: 4400 -18026.106 0.018 0.010
Chain 1: 4500 -17928.579 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00141 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.1 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13078.804 1.000 1.000
Chain 1: 200 -9811.081 0.667 1.000
Chain 1: 300 -8492.678 0.496 0.333
Chain 1: 400 -8663.046 0.377 0.333
Chain 1: 500 -8488.095 0.306 0.155
Chain 1: 600 -8385.590 0.257 0.155
Chain 1: 700 -8261.659 0.222 0.021
Chain 1: 800 -8259.532 0.195 0.021
Chain 1: 900 -8284.976 0.173 0.020
Chain 1: 1000 -8358.589 0.157 0.020
Chain 1: 1100 -8398.276 0.057 0.015
Chain 1: 1200 -8309.632 0.025 0.012
Chain 1: 1300 -8231.743 0.010 0.011
Chain 1: 1400 -8253.572 0.009 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001617 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 16.17 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57631.344 1.000 1.000
Chain 1: 200 -18152.048 1.587 2.175
Chain 1: 300 -9107.327 1.389 1.000
Chain 1: 400 -8368.823 1.064 1.000
Chain 1: 500 -8471.635 0.854 0.993
Chain 1: 600 -8806.573 0.718 0.993
Chain 1: 700 -8272.193 0.624 0.088
Chain 1: 800 -8274.255 0.546 0.088
Chain 1: 900 -8230.278 0.486 0.065
Chain 1: 1000 -8222.409 0.438 0.065
Chain 1: 1100 -7774.183 0.344 0.058
Chain 1: 1200 -8180.410 0.131 0.050
Chain 1: 1300 -7983.928 0.034 0.038
Chain 1: 1400 -7934.391 0.026 0.025
Chain 1: 1500 -7642.924 0.029 0.038
Chain 1: 1600 -7918.152 0.028 0.035
Chain 1: 1700 -7514.875 0.027 0.035
Chain 1: 1800 -7751.376 0.030 0.035
Chain 1: 1900 -7658.506 0.031 0.035
Chain 1: 2000 -7802.583 0.033 0.035
Chain 1: 2100 -7688.436 0.028 0.031
Chain 1: 2200 -7885.463 0.026 0.025
Chain 1: 2300 -7730.923 0.025 0.025
Chain 1: 2400 -7819.169 0.026 0.025
Chain 1: 2500 -7709.723 0.023 0.020
Chain 1: 2600 -7622.019 0.021 0.018
Chain 1: 2700 -7613.249 0.016 0.015
Chain 1: 2800 -7744.641 0.015 0.015
Chain 1: 2900 -7469.858 0.017 0.017
Chain 1: 3000 -7628.036 0.017 0.017
Chain 1: 3100 -7630.498 0.016 0.017
Chain 1: 3200 -7824.124 0.016 0.017
Chain 1: 3300 -7533.037 0.018 0.017
Chain 1: 3400 -7629.796 0.018 0.017
Chain 1: 3500 -7590.845 0.017 0.017
Chain 1: 3600 -7557.994 0.016 0.017
Chain 1: 3700 -7546.260 0.016 0.017
Chain 1: 3800 -7520.436 0.015 0.013
Chain 1: 3900 -7497.852 0.011 0.005 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00415 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 41.5 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86729.898 1.000 1.000
Chain 1: 200 -14256.967 3.042 5.083
Chain 1: 300 -10461.465 2.149 1.000
Chain 1: 400 -12379.396 1.650 1.000
Chain 1: 500 -8869.479 1.399 0.396
Chain 1: 600 -8759.173 1.168 0.396
Chain 1: 700 -8819.424 1.002 0.363
Chain 1: 800 -9141.696 0.881 0.363
Chain 1: 900 -9308.102 0.785 0.155
Chain 1: 1000 -8776.146 0.713 0.155
Chain 1: 1100 -9172.365 0.617 0.061 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8795.741 0.113 0.043
Chain 1: 1300 -9078.590 0.080 0.043
Chain 1: 1400 -8904.495 0.067 0.035
Chain 1: 1500 -8923.246 0.027 0.031
Chain 1: 1600 -9000.919 0.027 0.031
Chain 1: 1700 -9064.417 0.027 0.031
Chain 1: 1800 -8612.434 0.029 0.031
Chain 1: 1900 -8722.178 0.028 0.031
Chain 1: 2000 -8738.357 0.022 0.020
Chain 1: 2100 -8826.135 0.019 0.013
Chain 1: 2200 -8610.372 0.017 0.013
Chain 1: 2300 -8773.450 0.016 0.013
Chain 1: 2400 -8618.560 0.016 0.013
Chain 1: 2500 -8692.441 0.016 0.013
Chain 1: 2600 -8603.197 0.016 0.013
Chain 1: 2700 -8637.218 0.016 0.013
Chain 1: 2800 -8588.368 0.011 0.010
Chain 1: 2900 -8703.073 0.012 0.010
Chain 1: 3000 -8616.577 0.012 0.010
Chain 1: 3100 -8580.490 0.012 0.010
Chain 1: 3200 -8552.507 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003194 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.94 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8403543.759 1.000 1.000
Chain 1: 200 -1583718.024 2.653 4.306
Chain 1: 300 -891291.036 2.028 1.000
Chain 1: 400 -458263.415 1.757 1.000
Chain 1: 500 -358710.022 1.461 0.945
Chain 1: 600 -233654.384 1.307 0.945
Chain 1: 700 -119956.596 1.256 0.945
Chain 1: 800 -87185.906 1.146 0.945
Chain 1: 900 -67550.002 1.051 0.777
Chain 1: 1000 -52373.468 0.974 0.777
Chain 1: 1100 -39861.399 0.906 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39049.736 0.477 0.376
Chain 1: 1300 -26999.896 0.444 0.376
Chain 1: 1400 -26722.758 0.351 0.314
Chain 1: 1500 -23307.609 0.338 0.314
Chain 1: 1600 -22524.613 0.288 0.291
Chain 1: 1700 -21396.726 0.198 0.290
Chain 1: 1800 -21341.213 0.161 0.147
Chain 1: 1900 -21668.198 0.133 0.053
Chain 1: 2000 -20177.007 0.112 0.053
Chain 1: 2100 -20415.535 0.081 0.035
Chain 1: 2200 -20642.674 0.080 0.035
Chain 1: 2300 -20259.102 0.038 0.019
Chain 1: 2400 -20030.911 0.038 0.019
Chain 1: 2500 -19832.875 0.024 0.015
Chain 1: 2600 -19462.169 0.023 0.015
Chain 1: 2700 -19418.980 0.018 0.012
Chain 1: 2800 -19135.401 0.019 0.015
Chain 1: 2900 -19417.087 0.019 0.015
Chain 1: 3000 -19403.228 0.011 0.012
Chain 1: 3100 -19488.298 0.011 0.011
Chain 1: 3200 -19178.429 0.011 0.015
Chain 1: 3300 -19383.636 0.010 0.011
Chain 1: 3400 -18857.500 0.012 0.015
Chain 1: 3500 -19470.895 0.014 0.015
Chain 1: 3600 -18775.697 0.016 0.015
Chain 1: 3700 -19163.853 0.018 0.016
Chain 1: 3800 -18120.528 0.022 0.020
Chain 1: 3900 -18116.617 0.021 0.020
Chain 1: 4000 -18233.931 0.021 0.020
Chain 1: 4100 -18147.476 0.021 0.020
Chain 1: 4200 -17963.142 0.021 0.020
Chain 1: 4300 -18101.958 0.020 0.020
Chain 1: 4400 -18058.236 0.018 0.010
Chain 1: 4500 -17960.698 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001361 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.61 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12329.146 1.000 1.000
Chain 1: 200 -9343.657 0.660 1.000
Chain 1: 300 -7915.145 0.500 0.320
Chain 1: 400 -8002.988 0.378 0.320
Chain 1: 500 -7897.225 0.305 0.180
Chain 1: 600 -7823.815 0.256 0.180
Chain 1: 700 -7709.832 0.221 0.015
Chain 1: 800 -7713.704 0.194 0.015
Chain 1: 900 -7728.356 0.172 0.013
Chain 1: 1000 -7931.141 0.158 0.015
Chain 1: 1100 -7839.529 0.059 0.013
Chain 1: 1200 -7737.666 0.028 0.013
Chain 1: 1300 -7682.796 0.011 0.012
Chain 1: 1400 -7707.133 0.010 0.012
Chain 1: 1500 -7794.601 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001515 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.15 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61545.489 1.000 1.000
Chain 1: 200 -17870.073 1.722 2.444
Chain 1: 300 -8861.776 1.487 1.017
Chain 1: 400 -9533.755 1.133 1.017
Chain 1: 500 -8658.767 0.926 1.000
Chain 1: 600 -8796.065 0.775 1.000
Chain 1: 700 -8306.514 0.672 0.101
Chain 1: 800 -8313.833 0.588 0.101
Chain 1: 900 -7598.459 0.534 0.094
Chain 1: 1000 -7863.144 0.484 0.094
Chain 1: 1100 -7699.071 0.386 0.070
Chain 1: 1200 -7563.041 0.143 0.059
Chain 1: 1300 -7717.531 0.043 0.034
Chain 1: 1400 -7649.289 0.037 0.021
Chain 1: 1500 -7554.589 0.028 0.020
Chain 1: 1600 -7787.463 0.030 0.021
Chain 1: 1700 -7454.008 0.028 0.021
Chain 1: 1800 -7632.523 0.031 0.023
Chain 1: 1900 -7649.906 0.021 0.021
Chain 1: 2000 -7657.962 0.018 0.020
Chain 1: 2100 -7608.622 0.017 0.018
Chain 1: 2200 -7713.091 0.016 0.014
Chain 1: 2300 -7605.649 0.016 0.014
Chain 1: 2400 -7656.928 0.015 0.014
Chain 1: 2500 -7750.854 0.015 0.014
Chain 1: 2600 -7523.687 0.015 0.014
Chain 1: 2700 -7565.195 0.012 0.012
Chain 1: 2800 -7629.009 0.010 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003649 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 36.49 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86617.010 1.000 1.000
Chain 1: 200 -13505.475 3.207 5.413
Chain 1: 300 -9788.364 2.264 1.000
Chain 1: 400 -11172.529 1.729 1.000
Chain 1: 500 -8716.544 1.440 0.380
Chain 1: 600 -8193.998 1.210 0.380
Chain 1: 700 -8569.366 1.044 0.282
Chain 1: 800 -9312.639 0.923 0.282
Chain 1: 900 -8522.696 0.831 0.124
Chain 1: 1000 -8593.312 0.749 0.124
Chain 1: 1100 -8538.146 0.649 0.093 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8054.986 0.114 0.080
Chain 1: 1300 -8434.150 0.081 0.064
Chain 1: 1400 -8434.072 0.068 0.060
Chain 1: 1500 -8310.447 0.041 0.045
Chain 1: 1600 -8421.472 0.036 0.044
Chain 1: 1700 -8482.198 0.033 0.015
Chain 1: 1800 -8047.717 0.030 0.015
Chain 1: 1900 -8151.258 0.022 0.013
Chain 1: 2000 -8126.569 0.022 0.013
Chain 1: 2100 -8272.536 0.023 0.015
Chain 1: 2200 -8058.501 0.019 0.015
Chain 1: 2300 -8214.689 0.017 0.015
Chain 1: 2400 -8053.850 0.019 0.018
Chain 1: 2500 -8124.746 0.018 0.018
Chain 1: 2600 -8036.996 0.018 0.018
Chain 1: 2700 -8071.012 0.018 0.018
Chain 1: 2800 -8031.167 0.013 0.013
Chain 1: 2900 -8124.305 0.013 0.011
Chain 1: 3000 -7956.183 0.014 0.018
Chain 1: 3100 -8113.854 0.015 0.019
Chain 1: 3200 -7985.850 0.014 0.016
Chain 1: 3300 -7993.557 0.012 0.011
Chain 1: 3400 -8152.019 0.012 0.011
Chain 1: 3500 -8157.592 0.011 0.011
Chain 1: 3600 -7942.460 0.013 0.016
Chain 1: 3700 -8087.980 0.014 0.018
Chain 1: 3800 -7949.047 0.015 0.018
Chain 1: 3900 -7883.703 0.015 0.018
Chain 1: 4000 -7958.796 0.014 0.017
Chain 1: 4100 -7949.507 0.012 0.016
Chain 1: 4200 -7939.368 0.010 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003542 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.42 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8408180.848 1.000 1.000
Chain 1: 200 -1582197.592 2.657 4.314
Chain 1: 300 -890865.625 2.030 1.000
Chain 1: 400 -457838.936 1.759 1.000
Chain 1: 500 -358529.904 1.463 0.946
Chain 1: 600 -233385.920 1.308 0.946
Chain 1: 700 -119459.133 1.258 0.946
Chain 1: 800 -86631.385 1.148 0.946
Chain 1: 900 -66927.264 1.053 0.776
Chain 1: 1000 -51697.638 0.977 0.776
Chain 1: 1100 -39146.437 0.909 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38320.904 0.480 0.379
Chain 1: 1300 -26240.143 0.448 0.379
Chain 1: 1400 -25956.445 0.355 0.321
Chain 1: 1500 -22534.110 0.342 0.321
Chain 1: 1600 -21748.370 0.292 0.295
Chain 1: 1700 -20617.320 0.202 0.294
Chain 1: 1800 -20560.449 0.165 0.152
Chain 1: 1900 -20887.048 0.137 0.055
Chain 1: 2000 -19394.894 0.115 0.055
Chain 1: 2100 -19633.460 0.084 0.036
Chain 1: 2200 -19860.656 0.083 0.036
Chain 1: 2300 -19477.081 0.039 0.020
Chain 1: 2400 -19249.007 0.039 0.020
Chain 1: 2500 -19051.167 0.025 0.016
Chain 1: 2600 -18680.860 0.024 0.016
Chain 1: 2700 -18637.619 0.018 0.012
Chain 1: 2800 -18354.441 0.020 0.015
Chain 1: 2900 -18635.860 0.020 0.015
Chain 1: 3000 -18621.952 0.012 0.012
Chain 1: 3100 -18707.063 0.011 0.012
Chain 1: 3200 -18397.416 0.012 0.015
Chain 1: 3300 -18602.373 0.011 0.012
Chain 1: 3400 -18076.847 0.013 0.015
Chain 1: 3500 -18689.468 0.015 0.015
Chain 1: 3600 -17995.137 0.017 0.015
Chain 1: 3700 -18382.790 0.019 0.017
Chain 1: 3800 -17340.967 0.023 0.021
Chain 1: 3900 -17337.085 0.021 0.021
Chain 1: 4000 -17454.355 0.022 0.021
Chain 1: 4100 -17368.106 0.022 0.021
Chain 1: 4200 -17183.978 0.022 0.021
Chain 1: 4300 -17322.618 0.021 0.021
Chain 1: 4400 -17279.193 0.019 0.011
Chain 1: 4500 -17181.666 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001678 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 16.78 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12689.661 1.000 1.000
Chain 1: 200 -9375.297 0.677 1.000
Chain 1: 300 -7940.222 0.511 0.354
Chain 1: 400 -8052.121 0.387 0.354
Chain 1: 500 -7965.523 0.312 0.181
Chain 1: 600 -7841.529 0.262 0.181
Chain 1: 700 -7790.570 0.226 0.016
Chain 1: 800 -7772.214 0.198 0.016
Chain 1: 900 -7849.979 0.177 0.014
Chain 1: 1000 -7842.829 0.159 0.014
Chain 1: 1100 -7891.685 0.060 0.011
Chain 1: 1200 -7799.020 0.026 0.011
Chain 1: 1300 -7765.905 0.008 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00165 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 16.5 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57859.557 1.000 1.000
Chain 1: 200 -17382.692 1.664 2.329
Chain 1: 300 -8541.482 1.455 1.035
Chain 1: 400 -8163.042 1.103 1.035
Chain 1: 500 -7940.485 0.888 1.000
Chain 1: 600 -8639.445 0.753 1.000
Chain 1: 700 -7732.182 0.662 0.117
Chain 1: 800 -7979.806 0.583 0.117
Chain 1: 900 -7898.522 0.520 0.081
Chain 1: 1000 -7854.502 0.468 0.081
Chain 1: 1100 -7696.497 0.370 0.046
Chain 1: 1200 -7572.278 0.139 0.031
Chain 1: 1300 -7683.424 0.037 0.028
Chain 1: 1400 -7701.335 0.033 0.021
Chain 1: 1500 -7576.865 0.032 0.016
Chain 1: 1600 -7522.171 0.024 0.016
Chain 1: 1700 -7506.925 0.013 0.014
Chain 1: 1800 -7530.153 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003293 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.93 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86173.612 1.000 1.000
Chain 1: 200 -13198.463 3.265 5.529
Chain 1: 300 -9653.456 2.299 1.000
Chain 1: 400 -10604.697 1.746 1.000
Chain 1: 500 -8585.164 1.444 0.367
Chain 1: 600 -8218.464 1.211 0.367
Chain 1: 700 -8400.922 1.041 0.235
Chain 1: 800 -8706.784 0.915 0.235
Chain 1: 900 -8543.004 0.816 0.090
Chain 1: 1000 -8227.104 0.738 0.090
Chain 1: 1100 -8568.122 0.642 0.045 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8239.432 0.093 0.040
Chain 1: 1300 -8280.926 0.057 0.040
Chain 1: 1400 -8427.610 0.050 0.038
Chain 1: 1500 -8288.714 0.028 0.035
Chain 1: 1600 -8398.006 0.025 0.022
Chain 1: 1700 -8480.393 0.023 0.019
Chain 1: 1800 -8095.814 0.025 0.019
Chain 1: 1900 -8198.351 0.024 0.017
Chain 1: 2000 -8168.003 0.021 0.017
Chain 1: 2100 -8302.795 0.018 0.016
Chain 1: 2200 -8087.146 0.017 0.016
Chain 1: 2300 -8228.287 0.018 0.017
Chain 1: 2400 -8239.212 0.016 0.016
Chain 1: 2500 -8207.570 0.015 0.013
Chain 1: 2600 -8205.565 0.014 0.013
Chain 1: 2700 -8114.864 0.014 0.013
Chain 1: 2800 -8093.007 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002945 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.45 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8422295.585 1.000 1.000
Chain 1: 200 -1589767.769 2.649 4.298
Chain 1: 300 -890660.681 2.028 1.000
Chain 1: 400 -457266.708 1.758 1.000
Chain 1: 500 -357143.673 1.462 0.948
Chain 1: 600 -232071.469 1.308 0.948
Chain 1: 700 -118593.705 1.258 0.948
Chain 1: 800 -85856.046 1.148 0.948
Chain 1: 900 -66259.521 1.054 0.785
Chain 1: 1000 -51100.350 0.978 0.785
Chain 1: 1100 -38625.317 0.910 0.539 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37801.700 0.483 0.381
Chain 1: 1300 -25823.594 0.451 0.381
Chain 1: 1400 -25545.228 0.357 0.323
Chain 1: 1500 -22149.759 0.344 0.323
Chain 1: 1600 -21370.241 0.294 0.297
Chain 1: 1700 -20252.856 0.204 0.296
Chain 1: 1800 -20198.589 0.166 0.153
Chain 1: 1900 -20524.191 0.138 0.055
Chain 1: 2000 -19040.804 0.116 0.055
Chain 1: 2100 -19278.946 0.085 0.036
Chain 1: 2200 -19504.236 0.084 0.036
Chain 1: 2300 -19122.585 0.040 0.020
Chain 1: 2400 -18894.977 0.040 0.020
Chain 1: 2500 -18696.634 0.025 0.016
Chain 1: 2600 -18327.842 0.024 0.016
Chain 1: 2700 -18285.045 0.019 0.012
Chain 1: 2800 -18002.029 0.020 0.016
Chain 1: 2900 -18282.911 0.020 0.015
Chain 1: 3000 -18269.220 0.012 0.012
Chain 1: 3100 -18354.123 0.011 0.012
Chain 1: 3200 -18045.286 0.012 0.015
Chain 1: 3300 -18249.602 0.011 0.012
Chain 1: 3400 -17725.262 0.013 0.015
Chain 1: 3500 -18335.927 0.015 0.016
Chain 1: 3600 -17644.138 0.017 0.016
Chain 1: 3700 -18029.787 0.019 0.017
Chain 1: 3800 -16991.787 0.023 0.021
Chain 1: 3900 -16987.927 0.022 0.021
Chain 1: 4000 -17105.291 0.022 0.021
Chain 1: 4100 -17019.165 0.023 0.021
Chain 1: 4200 -16835.879 0.022 0.021
Chain 1: 4300 -16973.987 0.022 0.021
Chain 1: 4400 -16931.244 0.019 0.011
Chain 1: 4500 -16833.791 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001288 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.88 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12673.279 1.000 1.000
Chain 1: 200 -9521.116 0.666 1.000
Chain 1: 300 -8169.164 0.499 0.331
Chain 1: 400 -8273.883 0.377 0.331
Chain 1: 500 -8200.269 0.304 0.165
Chain 1: 600 -8115.792 0.255 0.165
Chain 1: 700 -8020.038 0.220 0.013
Chain 1: 800 -8058.940 0.193 0.013
Chain 1: 900 -8198.460 0.174 0.013
Chain 1: 1000 -8075.661 0.158 0.015
Chain 1: 1100 -8113.512 0.058 0.013
Chain 1: 1200 -8061.293 0.026 0.012
Chain 1: 1300 -7988.778 0.010 0.010
Chain 1: 1400 -8013.056 0.009 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002225 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 22.25 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -46532.665 1.000 1.000
Chain 1: 200 -15696.297 1.482 1.965
Chain 1: 300 -8766.993 1.252 1.000
Chain 1: 400 -8800.452 0.940 1.000
Chain 1: 500 -7686.071 0.781 0.790
Chain 1: 600 -8821.937 0.672 0.790
Chain 1: 700 -8149.773 0.588 0.145
Chain 1: 800 -8176.838 0.515 0.145
Chain 1: 900 -8062.684 0.459 0.129
Chain 1: 1000 -7938.316 0.415 0.129
Chain 1: 1100 -7912.710 0.315 0.082
Chain 1: 1200 -7784.365 0.120 0.016
Chain 1: 1300 -7811.900 0.042 0.016
Chain 1: 1400 -7816.279 0.041 0.016
Chain 1: 1500 -7682.554 0.029 0.016
Chain 1: 1600 -7685.236 0.016 0.014
Chain 1: 1700 -7606.261 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004363 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 43.63 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86174.740 1.000 1.000
Chain 1: 200 -13598.272 3.169 5.337
Chain 1: 300 -9976.986 2.233 1.000
Chain 1: 400 -10693.155 1.692 1.000
Chain 1: 500 -8963.672 1.392 0.363
Chain 1: 600 -8471.949 1.170 0.363
Chain 1: 700 -8848.962 1.009 0.193
Chain 1: 800 -9321.608 0.889 0.193
Chain 1: 900 -8740.825 0.798 0.067
Chain 1: 1000 -8609.782 0.719 0.067
Chain 1: 1100 -8845.778 0.622 0.066 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8380.055 0.094 0.058
Chain 1: 1300 -8695.992 0.061 0.056
Chain 1: 1400 -8695.218 0.054 0.051
Chain 1: 1500 -8568.788 0.037 0.043
Chain 1: 1600 -8675.998 0.032 0.036
Chain 1: 1700 -8762.311 0.029 0.027
Chain 1: 1800 -8355.805 0.029 0.027
Chain 1: 1900 -8452.635 0.023 0.015
Chain 1: 2000 -8424.757 0.022 0.015
Chain 1: 2100 -8545.249 0.021 0.014
Chain 1: 2200 -8355.879 0.017 0.014
Chain 1: 2300 -8492.359 0.015 0.014
Chain 1: 2400 -8499.676 0.015 0.014
Chain 1: 2500 -8465.920 0.014 0.012
Chain 1: 2600 -8463.929 0.013 0.011
Chain 1: 2700 -8377.955 0.013 0.011
Chain 1: 2800 -8343.174 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003551 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.51 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8368633.612 1.000 1.000
Chain 1: 200 -1579479.029 2.649 4.298
Chain 1: 300 -890515.966 2.024 1.000
Chain 1: 400 -457408.074 1.755 1.000
Chain 1: 500 -358205.013 1.459 0.947
Chain 1: 600 -233391.905 1.305 0.947
Chain 1: 700 -119529.429 1.255 0.947
Chain 1: 800 -86677.291 1.145 0.947
Chain 1: 900 -66996.854 1.051 0.774
Chain 1: 1000 -51763.010 0.975 0.774
Chain 1: 1100 -39204.676 0.907 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38377.263 0.479 0.379
Chain 1: 1300 -26301.632 0.448 0.379
Chain 1: 1400 -26017.365 0.354 0.320
Chain 1: 1500 -22595.381 0.342 0.320
Chain 1: 1600 -21808.603 0.292 0.294
Chain 1: 1700 -20678.854 0.202 0.294
Chain 1: 1800 -20622.031 0.164 0.151
Chain 1: 1900 -20948.047 0.137 0.055
Chain 1: 2000 -19457.197 0.115 0.055
Chain 1: 2100 -19695.916 0.084 0.036
Chain 1: 2200 -19922.473 0.083 0.036
Chain 1: 2300 -19539.571 0.039 0.020
Chain 1: 2400 -19311.630 0.039 0.020
Chain 1: 2500 -19113.689 0.025 0.016
Chain 1: 2600 -18744.118 0.023 0.016
Chain 1: 2700 -18701.076 0.018 0.012
Chain 1: 2800 -18418.017 0.019 0.015
Chain 1: 2900 -18699.231 0.019 0.015
Chain 1: 3000 -18685.483 0.012 0.012
Chain 1: 3100 -18770.440 0.011 0.012
Chain 1: 3200 -18461.241 0.012 0.015
Chain 1: 3300 -18665.834 0.011 0.012
Chain 1: 3400 -18140.985 0.012 0.015
Chain 1: 3500 -18752.613 0.015 0.015
Chain 1: 3600 -18059.608 0.017 0.015
Chain 1: 3700 -18446.214 0.018 0.017
Chain 1: 3800 -17406.451 0.023 0.021
Chain 1: 3900 -17402.587 0.021 0.021
Chain 1: 4000 -17519.898 0.022 0.021
Chain 1: 4100 -17433.705 0.022 0.021
Chain 1: 4200 -17250.019 0.021 0.021
Chain 1: 4300 -17388.378 0.021 0.021
Chain 1: 4400 -17345.310 0.018 0.011
Chain 1: 4500 -17247.824 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001497 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.97 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13444.875 1.000 1.000
Chain 1: 200 -10323.179 0.651 1.000
Chain 1: 300 -8768.874 0.493 0.302
Chain 1: 400 -9024.054 0.377 0.302
Chain 1: 500 -8805.285 0.307 0.177
Chain 1: 600 -8667.509 0.258 0.177
Chain 1: 700 -8541.210 0.223 0.028
Chain 1: 800 -8638.098 0.197 0.028
Chain 1: 900 -8544.680 0.176 0.025
Chain 1: 1000 -8635.354 0.160 0.025
Chain 1: 1100 -8612.839 0.060 0.016
Chain 1: 1200 -8608.501 0.030 0.015
Chain 1: 1300 -8536.695 0.013 0.011
Chain 1: 1400 -8547.582 0.010 0.011
Chain 1: 1500 -8621.588 0.008 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001638 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 16.38 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -65113.257 1.000 1.000
Chain 1: 200 -19551.555 1.665 2.330
Chain 1: 300 -9487.607 1.464 1.061
Chain 1: 400 -8560.099 1.125 1.061
Chain 1: 500 -8366.048 0.905 1.000
Chain 1: 600 -8529.812 0.757 1.000
Chain 1: 700 -9277.426 0.660 0.108
Chain 1: 800 -7737.927 0.603 0.199
Chain 1: 900 -9221.325 0.554 0.161
Chain 1: 1000 -8026.105 0.513 0.161
Chain 1: 1100 -8078.655 0.414 0.149
Chain 1: 1200 -7920.251 0.183 0.108
Chain 1: 1300 -7974.384 0.077 0.081
Chain 1: 1400 -7934.356 0.067 0.023
Chain 1: 1500 -7664.967 0.068 0.035
Chain 1: 1600 -7868.110 0.069 0.035
Chain 1: 1700 -7752.078 0.062 0.026
Chain 1: 1800 -7688.899 0.043 0.020
Chain 1: 1900 -7695.905 0.027 0.015
Chain 1: 2000 -7700.882 0.012 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003153 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.53 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87404.355 1.000 1.000
Chain 1: 200 -14715.356 2.970 4.940
Chain 1: 300 -10843.404 2.099 1.000
Chain 1: 400 -13216.666 1.619 1.000
Chain 1: 500 -9170.605 1.384 0.441
Chain 1: 600 -9464.180 1.158 0.441
Chain 1: 700 -9290.904 0.995 0.357
Chain 1: 800 -9426.218 0.873 0.357
Chain 1: 900 -9569.729 0.777 0.180
Chain 1: 1000 -9019.273 0.706 0.180
Chain 1: 1100 -9243.992 0.608 0.061 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8878.197 0.118 0.041
Chain 1: 1300 -9353.956 0.088 0.041
Chain 1: 1400 -9223.267 0.071 0.031
Chain 1: 1500 -9268.018 0.028 0.024
Chain 1: 1600 -9266.993 0.024 0.019
Chain 1: 1700 -9384.645 0.024 0.015
Chain 1: 1800 -8900.394 0.028 0.024
Chain 1: 1900 -9024.964 0.028 0.024
Chain 1: 2000 -9033.959 0.022 0.014
Chain 1: 2100 -9155.080 0.021 0.014
Chain 1: 2200 -8895.496 0.019 0.014
Chain 1: 2300 -8987.275 0.015 0.013
Chain 1: 2400 -9075.792 0.015 0.013
Chain 1: 2500 -8995.923 0.015 0.013
Chain 1: 2600 -9019.857 0.016 0.013
Chain 1: 2700 -8933.949 0.015 0.010
Chain 1: 2800 -8894.440 0.010 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003747 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 37.47 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8401028.145 1.000 1.000
Chain 1: 200 -1582968.148 2.654 4.307
Chain 1: 300 -891776.379 2.027 1.000
Chain 1: 400 -459090.355 1.756 1.000
Chain 1: 500 -359737.073 1.460 0.942
Chain 1: 600 -234583.744 1.306 0.942
Chain 1: 700 -120654.044 1.254 0.942
Chain 1: 800 -87864.817 1.144 0.942
Chain 1: 900 -68175.089 1.049 0.775
Chain 1: 1000 -52962.007 0.973 0.775
Chain 1: 1100 -40417.793 0.904 0.534 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39601.830 0.475 0.373
Chain 1: 1300 -27509.861 0.442 0.373
Chain 1: 1400 -27230.044 0.348 0.310
Chain 1: 1500 -23804.868 0.335 0.310
Chain 1: 1600 -23019.971 0.285 0.289
Chain 1: 1700 -21886.058 0.196 0.287
Chain 1: 1800 -21829.490 0.159 0.144
Chain 1: 1900 -22156.837 0.132 0.052
Chain 1: 2000 -20662.121 0.110 0.052
Chain 1: 2100 -20900.686 0.080 0.034
Chain 1: 2200 -21128.783 0.079 0.034
Chain 1: 2300 -20744.279 0.037 0.019
Chain 1: 2400 -20515.881 0.037 0.019
Chain 1: 2500 -20318.227 0.024 0.015
Chain 1: 2600 -19946.832 0.022 0.015
Chain 1: 2700 -19903.312 0.017 0.011
Chain 1: 2800 -19619.778 0.018 0.014
Chain 1: 2900 -19901.671 0.018 0.014
Chain 1: 3000 -19887.644 0.011 0.011
Chain 1: 3100 -19972.885 0.010 0.011
Chain 1: 3200 -19662.646 0.011 0.014
Chain 1: 3300 -19868.107 0.010 0.011
Chain 1: 3400 -19341.538 0.012 0.014
Chain 1: 3500 -19955.743 0.014 0.014
Chain 1: 3600 -19259.386 0.016 0.014
Chain 1: 3700 -19648.494 0.017 0.016
Chain 1: 3800 -18603.594 0.022 0.020
Chain 1: 3900 -18599.673 0.020 0.020
Chain 1: 4000 -18716.929 0.021 0.020
Chain 1: 4100 -18630.510 0.021 0.020
Chain 1: 4200 -18445.728 0.020 0.020
Chain 1: 4300 -18584.806 0.020 0.020
Chain 1: 4400 -18540.787 0.017 0.010
Chain 1: 4500 -18443.224 0.015 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001315 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.15 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49430.328 1.000 1.000
Chain 1: 200 -20853.478 1.185 1.370
Chain 1: 300 -21748.481 0.804 1.000
Chain 1: 400 -13032.291 0.770 1.000
Chain 1: 500 -16256.824 0.656 0.669
Chain 1: 600 -13222.191 0.585 0.669
Chain 1: 700 -12860.521 0.505 0.230
Chain 1: 800 -11826.830 0.453 0.230
Chain 1: 900 -11713.919 0.404 0.198
Chain 1: 1000 -12477.159 0.369 0.198
Chain 1: 1100 -28835.360 0.326 0.198
Chain 1: 1200 -10766.220 0.357 0.198
Chain 1: 1300 -10229.897 0.358 0.198
Chain 1: 1400 -11020.783 0.298 0.087
Chain 1: 1500 -9777.952 0.291 0.087
Chain 1: 1600 -15210.548 0.304 0.087
Chain 1: 1700 -21088.802 0.329 0.127
Chain 1: 1800 -12784.597 0.385 0.279
Chain 1: 1900 -10316.164 0.408 0.279
Chain 1: 2000 -19234.455 0.449 0.357
Chain 1: 2100 -11556.156 0.458 0.357
Chain 1: 2200 -10641.282 0.299 0.279
Chain 1: 2300 -17787.044 0.334 0.357
Chain 1: 2400 -9651.597 0.411 0.402
Chain 1: 2500 -9911.322 0.401 0.402
Chain 1: 2600 -13881.871 0.394 0.402
Chain 1: 2700 -9831.040 0.407 0.412
Chain 1: 2800 -17180.817 0.385 0.412
Chain 1: 2900 -9941.857 0.434 0.428
Chain 1: 3000 -11029.678 0.397 0.412
Chain 1: 3100 -9927.386 0.342 0.402
Chain 1: 3200 -10609.627 0.340 0.402
Chain 1: 3300 -17052.754 0.337 0.378
Chain 1: 3400 -9210.861 0.338 0.378
Chain 1: 3500 -9269.399 0.336 0.378
Chain 1: 3600 -9199.470 0.309 0.378
Chain 1: 3700 -9379.232 0.269 0.111
Chain 1: 3800 -9340.581 0.227 0.099
Chain 1: 3900 -10657.891 0.166 0.099
Chain 1: 4000 -9004.502 0.175 0.111
Chain 1: 4100 -8941.209 0.165 0.064
Chain 1: 4200 -10047.746 0.169 0.110
Chain 1: 4300 -9980.291 0.132 0.019
Chain 1: 4400 -8975.443 0.058 0.019
Chain 1: 4500 -9162.218 0.059 0.020
Chain 1: 4600 -13790.867 0.092 0.110
Chain 1: 4700 -8811.909 0.147 0.112
Chain 1: 4800 -8781.332 0.147 0.112
Chain 1: 4900 -14527.768 0.174 0.112
Chain 1: 5000 -9682.848 0.206 0.112
Chain 1: 5100 -13163.535 0.231 0.264
Chain 1: 5200 -13725.465 0.224 0.264
Chain 1: 5300 -8951.856 0.277 0.336
Chain 1: 5400 -16734.311 0.312 0.396
Chain 1: 5500 -13434.008 0.335 0.396
Chain 1: 5600 -8717.973 0.355 0.465
Chain 1: 5700 -10219.325 0.314 0.396
Chain 1: 5800 -11265.069 0.323 0.396
Chain 1: 5900 -15047.252 0.308 0.264
Chain 1: 6000 -9733.269 0.313 0.264
Chain 1: 6100 -9882.423 0.288 0.251
Chain 1: 6200 -9084.033 0.292 0.251
Chain 1: 6300 -8636.490 0.244 0.246
Chain 1: 6400 -8430.586 0.200 0.147
Chain 1: 6500 -9398.322 0.186 0.103
Chain 1: 6600 -8778.953 0.139 0.093
Chain 1: 6700 -8672.568 0.126 0.088
Chain 1: 6800 -8651.609 0.116 0.071
Chain 1: 6900 -8602.641 0.092 0.052
Chain 1: 7000 -12204.857 0.067 0.052
Chain 1: 7100 -8792.493 0.104 0.071
Chain 1: 7200 -8940.121 0.097 0.052
Chain 1: 7300 -8715.585 0.094 0.026
Chain 1: 7400 -9008.455 0.095 0.033
Chain 1: 7500 -8554.468 0.090 0.033
Chain 1: 7600 -8776.855 0.086 0.026
Chain 1: 7700 -8371.717 0.089 0.033
Chain 1: 7800 -8850.632 0.094 0.048
Chain 1: 7900 -8429.340 0.099 0.050
Chain 1: 8000 -8757.330 0.073 0.048
Chain 1: 8100 -8711.487 0.035 0.037
Chain 1: 8200 -9822.605 0.044 0.048
Chain 1: 8300 -8549.024 0.057 0.050
Chain 1: 8400 -9015.202 0.059 0.052
Chain 1: 8500 -8432.423 0.060 0.052
Chain 1: 8600 -9422.246 0.068 0.054
Chain 1: 8700 -10522.686 0.074 0.069
Chain 1: 8800 -8588.733 0.091 0.105
Chain 1: 8900 -8971.879 0.090 0.105
Chain 1: 9000 -8542.457 0.092 0.105
Chain 1: 9100 -8718.645 0.093 0.105
Chain 1: 9200 -8411.279 0.085 0.069
Chain 1: 9300 -8567.176 0.072 0.052
Chain 1: 9400 -10194.422 0.083 0.069
Chain 1: 9500 -9112.888 0.088 0.105
Chain 1: 9600 -10546.554 0.091 0.105
Chain 1: 9700 -8643.806 0.103 0.119
Chain 1: 9800 -13127.194 0.114 0.119
Chain 1: 9900 -8378.418 0.167 0.136
Chain 1: 10000 -11081.526 0.186 0.160
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00149 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.9 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58937.213 1.000 1.000
Chain 1: 200 -18191.287 1.620 2.240
Chain 1: 300 -8908.107 1.427 1.042
Chain 1: 400 -8170.422 1.093 1.042
Chain 1: 500 -8761.635 0.888 1.000
Chain 1: 600 -8125.829 0.753 1.000
Chain 1: 700 -8337.442 0.649 0.090
Chain 1: 800 -8469.111 0.570 0.090
Chain 1: 900 -7737.235 0.517 0.090
Chain 1: 1000 -7983.583 0.468 0.090
Chain 1: 1100 -7740.785 0.372 0.078
Chain 1: 1200 -7719.399 0.148 0.067
Chain 1: 1300 -7795.589 0.045 0.031
Chain 1: 1400 -7681.052 0.037 0.031
Chain 1: 1500 -7594.697 0.031 0.025
Chain 1: 1600 -7737.093 0.025 0.018
Chain 1: 1700 -7615.438 0.025 0.016
Chain 1: 1800 -7757.499 0.025 0.018
Chain 1: 1900 -7637.454 0.017 0.016
Chain 1: 2000 -7715.465 0.015 0.016
Chain 1: 2100 -7607.653 0.013 0.015
Chain 1: 2200 -7830.243 0.016 0.016
Chain 1: 2300 -7573.691 0.018 0.016
Chain 1: 2400 -7592.343 0.017 0.016
Chain 1: 2500 -7654.139 0.017 0.016
Chain 1: 2600 -7574.293 0.016 0.016
Chain 1: 2700 -7497.908 0.015 0.014
Chain 1: 2800 -7530.946 0.014 0.011
Chain 1: 2900 -7454.204 0.013 0.010
Chain 1: 3000 -7592.514 0.014 0.011
Chain 1: 3100 -7573.971 0.013 0.010
Chain 1: 3200 -7778.887 0.013 0.010
Chain 1: 3300 -7494.650 0.013 0.010
Chain 1: 3400 -7729.499 0.016 0.011
Chain 1: 3500 -7481.403 0.018 0.018
Chain 1: 3600 -7546.085 0.018 0.018
Chain 1: 3700 -7497.197 0.018 0.018
Chain 1: 3800 -7497.072 0.017 0.018
Chain 1: 3900 -7456.788 0.017 0.018
Chain 1: 4000 -7448.700 0.015 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003267 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.67 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86276.716 1.000 1.000
Chain 1: 200 -13916.310 3.100 5.200
Chain 1: 300 -10160.063 2.190 1.000
Chain 1: 400 -11666.719 1.675 1.000
Chain 1: 500 -8918.737 1.401 0.370
Chain 1: 600 -8580.429 1.174 0.370
Chain 1: 700 -8489.898 1.008 0.308
Chain 1: 800 -8908.477 0.888 0.308
Chain 1: 900 -8889.511 0.790 0.129
Chain 1: 1000 -9108.223 0.713 0.129
Chain 1: 1100 -8805.186 0.616 0.047 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8484.403 0.100 0.039
Chain 1: 1300 -8779.864 0.067 0.038
Chain 1: 1400 -8606.330 0.056 0.034
Chain 1: 1500 -8634.530 0.025 0.034
Chain 1: 1600 -8742.856 0.023 0.024
Chain 1: 1700 -8794.076 0.022 0.024
Chain 1: 1800 -8341.829 0.023 0.024
Chain 1: 1900 -8450.616 0.024 0.024
Chain 1: 2000 -8450.306 0.021 0.020
Chain 1: 2100 -8617.338 0.020 0.019
Chain 1: 2200 -8346.637 0.019 0.019
Chain 1: 2300 -8526.656 0.018 0.019
Chain 1: 2400 -8346.250 0.018 0.019
Chain 1: 2500 -8422.882 0.019 0.019
Chain 1: 2600 -8333.524 0.019 0.019
Chain 1: 2700 -8367.046 0.019 0.019
Chain 1: 2800 -8318.746 0.014 0.013
Chain 1: 2900 -8430.284 0.014 0.013
Chain 1: 3000 -8367.322 0.014 0.013
Chain 1: 3100 -8311.033 0.013 0.011
Chain 1: 3200 -8284.065 0.010 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003597 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.97 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8403003.436 1.000 1.000
Chain 1: 200 -1584668.866 2.651 4.303
Chain 1: 300 -890843.215 2.027 1.000
Chain 1: 400 -458166.513 1.756 1.000
Chain 1: 500 -358639.302 1.461 0.944
Chain 1: 600 -233478.222 1.307 0.944
Chain 1: 700 -119692.722 1.256 0.944
Chain 1: 800 -86915.330 1.146 0.944
Chain 1: 900 -67256.486 1.051 0.779
Chain 1: 1000 -52061.868 0.975 0.779
Chain 1: 1100 -39540.729 0.907 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38722.281 0.479 0.377
Chain 1: 1300 -26668.192 0.446 0.377
Chain 1: 1400 -26388.527 0.353 0.317
Chain 1: 1500 -22973.325 0.340 0.317
Chain 1: 1600 -22190.240 0.290 0.292
Chain 1: 1700 -21061.974 0.200 0.292
Chain 1: 1800 -21006.200 0.162 0.149
Chain 1: 1900 -21333.051 0.135 0.054
Chain 1: 2000 -19841.953 0.113 0.054
Chain 1: 2100 -20080.393 0.083 0.035
Chain 1: 2200 -20307.608 0.082 0.035
Chain 1: 2300 -19923.969 0.038 0.019
Chain 1: 2400 -19695.822 0.038 0.019
Chain 1: 2500 -19497.898 0.025 0.015
Chain 1: 2600 -19127.318 0.023 0.015
Chain 1: 2700 -19084.014 0.018 0.012
Chain 1: 2800 -18800.622 0.019 0.015
Chain 1: 2900 -19082.205 0.019 0.015
Chain 1: 3000 -19068.266 0.012 0.012
Chain 1: 3100 -19153.417 0.011 0.012
Chain 1: 3200 -18843.603 0.011 0.015
Chain 1: 3300 -19048.712 0.011 0.012
Chain 1: 3400 -18522.788 0.012 0.015
Chain 1: 3500 -19135.976 0.014 0.015
Chain 1: 3600 -18440.918 0.016 0.015
Chain 1: 3700 -18829.042 0.018 0.016
Chain 1: 3800 -17786.109 0.022 0.021
Chain 1: 3900 -17782.204 0.021 0.021
Chain 1: 4000 -17899.488 0.022 0.021
Chain 1: 4100 -17813.161 0.022 0.021
Chain 1: 4200 -17628.803 0.021 0.021
Chain 1: 4300 -17767.615 0.021 0.021
Chain 1: 4400 -17723.963 0.018 0.010
Chain 1: 4500 -17626.425 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001198 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 11.98 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -11950.309 1.000 1.000
Chain 1: 200 -8908.292 0.671 1.000
Chain 1: 300 -7955.638 0.487 0.341
Chain 1: 400 -8045.048 0.368 0.341
Chain 1: 500 -7855.076 0.299 0.120
Chain 1: 600 -7790.837 0.251 0.120
Chain 1: 700 -7731.234 0.216 0.024
Chain 1: 800 -7746.022 0.189 0.024
Chain 1: 900 -7814.442 0.169 0.011
Chain 1: 1000 -7771.170 0.153 0.011
Chain 1: 1100 -7834.849 0.054 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00139 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.9 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56850.893 1.000 1.000
Chain 1: 200 -16994.501 1.673 2.345
Chain 1: 300 -8514.635 1.447 1.000
Chain 1: 400 -8696.547 1.091 1.000
Chain 1: 500 -8396.931 0.880 0.996
Chain 1: 600 -8441.804 0.734 0.996
Chain 1: 700 -8217.046 0.633 0.036
Chain 1: 800 -8052.964 0.556 0.036
Chain 1: 900 -7840.678 0.498 0.027
Chain 1: 1000 -7730.262 0.449 0.027
Chain 1: 1100 -7751.061 0.349 0.027
Chain 1: 1200 -7598.379 0.117 0.021
Chain 1: 1300 -7639.026 0.018 0.020
Chain 1: 1400 -7862.545 0.019 0.020
Chain 1: 1500 -7596.357 0.019 0.020
Chain 1: 1600 -7495.963 0.019 0.020
Chain 1: 1700 -7481.471 0.017 0.020
Chain 1: 1800 -7506.798 0.015 0.014
Chain 1: 1900 -7562.858 0.013 0.013
Chain 1: 2000 -7564.115 0.012 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002632 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 26.32 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85551.442 1.000 1.000
Chain 1: 200 -13015.102 3.287 5.573
Chain 1: 300 -9526.701 2.313 1.000
Chain 1: 400 -10175.454 1.751 1.000
Chain 1: 500 -8427.445 1.442 0.366
Chain 1: 600 -8432.742 1.202 0.366
Chain 1: 700 -8459.944 1.031 0.207
Chain 1: 800 -8640.262 0.904 0.207
Chain 1: 900 -8433.028 0.807 0.064
Chain 1: 1000 -8165.339 0.729 0.064
Chain 1: 1100 -8329.631 0.631 0.033 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8109.623 0.077 0.027
Chain 1: 1300 -8185.705 0.041 0.025
Chain 1: 1400 -8313.887 0.036 0.021
Chain 1: 1500 -8219.463 0.017 0.020
Chain 1: 1600 -8302.273 0.017 0.020
Chain 1: 1700 -8401.281 0.018 0.020
Chain 1: 1800 -8025.316 0.021 0.020
Chain 1: 1900 -8122.016 0.020 0.015
Chain 1: 2000 -8092.662 0.017 0.012
Chain 1: 2100 -8240.009 0.017 0.012
Chain 1: 2200 -8016.324 0.017 0.012
Chain 1: 2300 -8105.405 0.017 0.012
Chain 1: 2400 -8169.831 0.016 0.012
Chain 1: 2500 -8129.592 0.015 0.012
Chain 1: 2600 -8122.484 0.014 0.012
Chain 1: 2700 -8035.174 0.014 0.011
Chain 1: 2800 -8019.886 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004813 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 48.13 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8406559.686 1.000 1.000
Chain 1: 200 -1586011.718 2.650 4.300
Chain 1: 300 -890914.533 2.027 1.000
Chain 1: 400 -457419.558 1.757 1.000
Chain 1: 500 -357531.942 1.462 0.948
Chain 1: 600 -232424.168 1.308 0.948
Chain 1: 700 -118662.506 1.258 0.948
Chain 1: 800 -85882.862 1.148 0.948
Chain 1: 900 -66226.315 1.054 0.780
Chain 1: 1000 -51017.751 0.978 0.780
Chain 1: 1100 -38501.239 0.911 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37668.305 0.483 0.382
Chain 1: 1300 -25648.719 0.452 0.382
Chain 1: 1400 -25364.632 0.358 0.325
Chain 1: 1500 -21959.711 0.346 0.325
Chain 1: 1600 -21177.072 0.295 0.298
Chain 1: 1700 -20054.843 0.205 0.297
Chain 1: 1800 -19999.287 0.167 0.155
Chain 1: 1900 -20324.503 0.139 0.056
Chain 1: 2000 -18839.524 0.117 0.056
Chain 1: 2100 -19077.500 0.086 0.037
Chain 1: 2200 -19303.135 0.085 0.037
Chain 1: 2300 -18921.332 0.040 0.020
Chain 1: 2400 -18693.817 0.040 0.020
Chain 1: 2500 -18495.799 0.026 0.016
Chain 1: 2600 -18126.976 0.024 0.016
Chain 1: 2700 -18084.197 0.019 0.012
Chain 1: 2800 -17801.501 0.020 0.016
Chain 1: 2900 -18082.274 0.020 0.016
Chain 1: 3000 -18068.493 0.012 0.012
Chain 1: 3100 -18153.368 0.011 0.012
Chain 1: 3200 -17844.672 0.012 0.016
Chain 1: 3300 -18048.897 0.011 0.012
Chain 1: 3400 -17524.948 0.013 0.016
Chain 1: 3500 -18135.120 0.015 0.016
Chain 1: 3600 -17444.018 0.017 0.016
Chain 1: 3700 -17829.180 0.019 0.017
Chain 1: 3800 -16792.332 0.024 0.022
Chain 1: 3900 -16788.575 0.022 0.022
Chain 1: 4000 -16905.864 0.023 0.022
Chain 1: 4100 -16819.820 0.023 0.022
Chain 1: 4200 -16636.803 0.022 0.022
Chain 1: 4300 -16774.673 0.022 0.022
Chain 1: 4400 -16732.109 0.019 0.011
Chain 1: 4500 -16634.765 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001277 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.77 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12663.776 1.000 1.000
Chain 1: 200 -9598.379 0.660 1.000
Chain 1: 300 -8236.391 0.495 0.319
Chain 1: 400 -8470.562 0.378 0.319
Chain 1: 500 -8322.799 0.306 0.165
Chain 1: 600 -8181.659 0.258 0.165
Chain 1: 700 -8081.743 0.223 0.028
Chain 1: 800 -8085.250 0.195 0.028
Chain 1: 900 -8012.615 0.174 0.018
Chain 1: 1000 -8205.944 0.159 0.024
Chain 1: 1100 -8231.210 0.060 0.018
Chain 1: 1200 -8098.466 0.029 0.017
Chain 1: 1300 -8067.037 0.013 0.016
Chain 1: 1400 -8075.422 0.010 0.012
Chain 1: 1500 -8164.993 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001487 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.87 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57398.923 1.000 1.000
Chain 1: 200 -17752.220 1.617 2.233
Chain 1: 300 -8857.660 1.413 1.004
Chain 1: 400 -8149.910 1.081 1.004
Chain 1: 500 -8443.883 0.872 1.000
Chain 1: 600 -8723.537 0.732 1.000
Chain 1: 700 -7952.707 0.641 0.097
Chain 1: 800 -8447.304 0.568 0.097
Chain 1: 900 -7971.551 0.512 0.087
Chain 1: 1000 -7654.899 0.465 0.087
Chain 1: 1100 -7893.912 0.368 0.060
Chain 1: 1200 -8040.879 0.146 0.059
Chain 1: 1300 -7698.269 0.050 0.045
Chain 1: 1400 -7832.182 0.043 0.041
Chain 1: 1500 -7598.059 0.043 0.041
Chain 1: 1600 -7764.994 0.042 0.041
Chain 1: 1700 -7550.066 0.035 0.031
Chain 1: 1800 -7582.766 0.030 0.030
Chain 1: 1900 -7601.567 0.024 0.028
Chain 1: 2000 -7756.694 0.022 0.021
Chain 1: 2100 -7596.375 0.021 0.021
Chain 1: 2200 -7848.842 0.022 0.021
Chain 1: 2300 -7545.695 0.022 0.021
Chain 1: 2400 -7608.125 0.021 0.021
Chain 1: 2500 -7628.562 0.018 0.021
Chain 1: 2600 -7526.553 0.017 0.020
Chain 1: 2700 -7514.244 0.015 0.014
Chain 1: 2800 -7502.663 0.014 0.014
Chain 1: 2900 -7408.652 0.015 0.014
Chain 1: 3000 -7542.528 0.015 0.014
Chain 1: 3100 -7532.264 0.013 0.013
Chain 1: 3200 -7733.606 0.013 0.013
Chain 1: 3300 -7454.268 0.012 0.013
Chain 1: 3400 -7681.837 0.014 0.014
Chain 1: 3500 -7438.637 0.017 0.018
Chain 1: 3600 -7504.833 0.017 0.018
Chain 1: 3700 -7454.185 0.017 0.018
Chain 1: 3800 -7452.393 0.017 0.018
Chain 1: 3900 -7417.998 0.017 0.018
Chain 1: 4000 -7414.291 0.015 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003058 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.58 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86840.627 1.000 1.000
Chain 1: 200 -13826.548 3.140 5.281
Chain 1: 300 -10129.794 2.215 1.000
Chain 1: 400 -11343.884 1.688 1.000
Chain 1: 500 -9129.483 1.399 0.365
Chain 1: 600 -8696.847 1.174 0.365
Chain 1: 700 -8531.769 1.009 0.243
Chain 1: 800 -9570.529 0.897 0.243
Chain 1: 900 -8821.658 0.806 0.109
Chain 1: 1000 -8799.869 0.726 0.109
Chain 1: 1100 -8853.019 0.627 0.107 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8523.099 0.102 0.085
Chain 1: 1300 -8803.920 0.069 0.050
Chain 1: 1400 -8807.308 0.058 0.039
Chain 1: 1500 -8653.631 0.036 0.032
Chain 1: 1600 -8768.779 0.032 0.019
Chain 1: 1700 -8837.145 0.031 0.018
Chain 1: 1800 -8404.678 0.025 0.018
Chain 1: 1900 -8508.739 0.018 0.013
Chain 1: 2000 -8484.179 0.018 0.013
Chain 1: 2100 -8460.843 0.018 0.013
Chain 1: 2200 -8426.586 0.014 0.012
Chain 1: 2300 -8556.369 0.013 0.012
Chain 1: 2400 -8411.446 0.014 0.013
Chain 1: 2500 -8480.375 0.013 0.012
Chain 1: 2600 -8399.541 0.013 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003034 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.34 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8406859.334 1.000 1.000
Chain 1: 200 -1584848.073 2.652 4.305
Chain 1: 300 -890859.035 2.028 1.000
Chain 1: 400 -457945.266 1.757 1.000
Chain 1: 500 -358472.129 1.461 0.945
Chain 1: 600 -233317.620 1.307 0.945
Chain 1: 700 -119533.997 1.256 0.945
Chain 1: 800 -86800.292 1.146 0.945
Chain 1: 900 -67138.551 1.052 0.779
Chain 1: 1000 -51939.912 0.976 0.779
Chain 1: 1100 -39421.301 0.907 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38597.826 0.479 0.377
Chain 1: 1300 -26544.725 0.447 0.377
Chain 1: 1400 -26264.163 0.353 0.318
Chain 1: 1500 -22849.983 0.340 0.318
Chain 1: 1600 -22067.112 0.290 0.293
Chain 1: 1700 -20938.910 0.201 0.293
Chain 1: 1800 -20882.823 0.163 0.149
Chain 1: 1900 -21209.348 0.135 0.054
Chain 1: 2000 -19719.174 0.114 0.054
Chain 1: 2100 -19957.443 0.083 0.035
Chain 1: 2200 -20184.532 0.082 0.035
Chain 1: 2300 -19801.066 0.039 0.019
Chain 1: 2400 -19573.039 0.039 0.019
Chain 1: 2500 -19375.304 0.025 0.015
Chain 1: 2600 -19005.004 0.023 0.015
Chain 1: 2700 -18961.758 0.018 0.012
Chain 1: 2800 -18678.682 0.019 0.015
Chain 1: 2900 -18959.996 0.019 0.015
Chain 1: 3000 -18946.071 0.012 0.012
Chain 1: 3100 -19031.198 0.011 0.012
Chain 1: 3200 -18721.612 0.011 0.015
Chain 1: 3300 -18926.518 0.011 0.012
Chain 1: 3400 -18401.135 0.012 0.015
Chain 1: 3500 -19013.606 0.015 0.015
Chain 1: 3600 -18319.403 0.016 0.015
Chain 1: 3700 -18706.922 0.018 0.017
Chain 1: 3800 -17665.435 0.023 0.021
Chain 1: 3900 -17661.567 0.021 0.021
Chain 1: 4000 -17778.816 0.022 0.021
Chain 1: 4100 -17692.638 0.022 0.021
Chain 1: 4200 -17508.536 0.021 0.021
Chain 1: 4300 -17647.130 0.021 0.021
Chain 1: 4400 -17603.731 0.018 0.011
Chain 1: 4500 -17506.237 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00138 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.8 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49647.186 1.000 1.000
Chain 1: 200 -19213.032 1.292 1.584
Chain 1: 300 -13758.477 0.993 1.000
Chain 1: 400 -48829.236 0.925 1.000
Chain 1: 500 -20102.928 1.026 1.000
Chain 1: 600 -27008.781 0.897 1.000
Chain 1: 700 -12210.073 0.942 1.000
Chain 1: 800 -13617.352 0.837 1.000
Chain 1: 900 -14195.548 0.749 0.718
Chain 1: 1000 -12848.179 0.684 0.718
Chain 1: 1100 -11301.333 0.598 0.396 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -12577.915 0.450 0.256
Chain 1: 1300 -14504.655 0.424 0.137
Chain 1: 1400 -14145.431 0.354 0.133
Chain 1: 1500 -12536.720 0.224 0.128
Chain 1: 1600 -20354.493 0.237 0.128
Chain 1: 1700 -10661.538 0.207 0.128
Chain 1: 1800 -10351.579 0.199 0.128
Chain 1: 1900 -19570.619 0.242 0.133
Chain 1: 2000 -13909.121 0.273 0.137
Chain 1: 2100 -19058.128 0.286 0.270
Chain 1: 2200 -12166.176 0.332 0.384
Chain 1: 2300 -10600.295 0.334 0.384
Chain 1: 2400 -10220.154 0.335 0.384
Chain 1: 2500 -10145.432 0.323 0.384
Chain 1: 2600 -10171.721 0.285 0.270
Chain 1: 2700 -10497.114 0.197 0.148
Chain 1: 2800 -11506.027 0.203 0.148
Chain 1: 2900 -9770.363 0.173 0.148
Chain 1: 3000 -10771.102 0.142 0.093
Chain 1: 3100 -10102.567 0.122 0.088
Chain 1: 3200 -10356.493 0.067 0.066
Chain 1: 3300 -9720.552 0.059 0.065
Chain 1: 3400 -17118.354 0.099 0.066
Chain 1: 3500 -10117.856 0.167 0.088
Chain 1: 3600 -11965.505 0.182 0.093
Chain 1: 3700 -9831.549 0.201 0.154
Chain 1: 3800 -18183.587 0.238 0.178
Chain 1: 3900 -9733.322 0.307 0.217
Chain 1: 4000 -9683.277 0.298 0.217
Chain 1: 4100 -10434.944 0.299 0.217
Chain 1: 4200 -17103.026 0.336 0.390
Chain 1: 4300 -10961.398 0.385 0.432
Chain 1: 4400 -9541.052 0.357 0.390
Chain 1: 4500 -9926.962 0.291 0.217
Chain 1: 4600 -9245.240 0.283 0.217
Chain 1: 4700 -9558.937 0.265 0.149
Chain 1: 4800 -9068.164 0.224 0.074
Chain 1: 4900 -9943.830 0.146 0.074
Chain 1: 5000 -9725.419 0.148 0.074
Chain 1: 5100 -12394.512 0.162 0.088
Chain 1: 5200 -13630.888 0.133 0.088
Chain 1: 5300 -9557.655 0.119 0.088
Chain 1: 5400 -9115.898 0.109 0.074
Chain 1: 5500 -9973.455 0.114 0.086
Chain 1: 5600 -11779.975 0.122 0.088
Chain 1: 5700 -16766.047 0.148 0.091
Chain 1: 5800 -9818.706 0.214 0.153
Chain 1: 5900 -9992.817 0.206 0.153
Chain 1: 6000 -12749.446 0.226 0.215
Chain 1: 6100 -9938.753 0.233 0.216
Chain 1: 6200 -9991.250 0.224 0.216
Chain 1: 6300 -13574.975 0.208 0.216
Chain 1: 6400 -13559.868 0.203 0.216
Chain 1: 6500 -12759.163 0.201 0.216
Chain 1: 6600 -9262.863 0.223 0.264
Chain 1: 6700 -13209.948 0.223 0.264
Chain 1: 6800 -14205.855 0.160 0.216
Chain 1: 6900 -9081.817 0.214 0.264
Chain 1: 7000 -14933.315 0.232 0.283
Chain 1: 7100 -8778.797 0.274 0.299
Chain 1: 7200 -11254.295 0.295 0.299
Chain 1: 7300 -10111.363 0.280 0.299
Chain 1: 7400 -8801.585 0.295 0.299
Chain 1: 7500 -9004.029 0.291 0.299
Chain 1: 7600 -8942.459 0.254 0.220
Chain 1: 7700 -10121.221 0.235 0.149
Chain 1: 7800 -13692.998 0.255 0.220
Chain 1: 7900 -8882.160 0.252 0.220
Chain 1: 8000 -8809.694 0.214 0.149
Chain 1: 8100 -11277.767 0.166 0.149
Chain 1: 8200 -8861.180 0.171 0.149
Chain 1: 8300 -8754.782 0.161 0.149
Chain 1: 8400 -13117.185 0.179 0.219
Chain 1: 8500 -9153.231 0.220 0.261
Chain 1: 8600 -9795.084 0.226 0.261
Chain 1: 8700 -9964.348 0.216 0.261
Chain 1: 8800 -12063.540 0.208 0.219
Chain 1: 8900 -9817.424 0.176 0.219
Chain 1: 9000 -10711.738 0.184 0.219
Chain 1: 9100 -9380.888 0.176 0.174
Chain 1: 9200 -8665.792 0.157 0.142
Chain 1: 9300 -9116.016 0.161 0.142
Chain 1: 9400 -9518.173 0.132 0.083
Chain 1: 9500 -12986.465 0.115 0.083
Chain 1: 9600 -11449.109 0.122 0.134
Chain 1: 9700 -8789.194 0.151 0.142
Chain 1: 9800 -9442.742 0.140 0.134
Chain 1: 9900 -11116.284 0.132 0.134
Chain 1: 10000 -9416.861 0.142 0.142
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002141 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 21.41 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -62821.644 1.000 1.000
Chain 1: 200 -18848.536 1.666 2.333
Chain 1: 300 -9398.980 1.446 1.005
Chain 1: 400 -8665.041 1.106 1.005
Chain 1: 500 -9455.366 0.901 1.000
Chain 1: 600 -8273.626 0.775 1.000
Chain 1: 700 -9166.804 0.678 0.143
Chain 1: 800 -7931.448 0.613 0.156
Chain 1: 900 -8038.636 0.546 0.143
Chain 1: 1000 -8379.252 0.496 0.143
Chain 1: 1100 -8097.664 0.399 0.097
Chain 1: 1200 -7612.392 0.172 0.085
Chain 1: 1300 -7727.827 0.073 0.084
Chain 1: 1400 -8109.011 0.069 0.064
Chain 1: 1500 -7670.325 0.067 0.057
Chain 1: 1600 -7880.684 0.055 0.047
Chain 1: 1700 -7550.675 0.050 0.044
Chain 1: 1800 -7685.163 0.036 0.041
Chain 1: 1900 -7659.592 0.035 0.041
Chain 1: 2000 -7830.175 0.033 0.035
Chain 1: 2100 -7722.361 0.031 0.027
Chain 1: 2200 -7902.376 0.027 0.023
Chain 1: 2300 -7721.554 0.028 0.023
Chain 1: 2400 -7782.273 0.024 0.023
Chain 1: 2500 -7627.759 0.020 0.022
Chain 1: 2600 -7636.267 0.018 0.020
Chain 1: 2700 -7643.996 0.013 0.017
Chain 1: 2800 -7743.346 0.013 0.014
Chain 1: 2900 -7470.079 0.016 0.020
Chain 1: 3000 -7646.872 0.016 0.020
Chain 1: 3100 -7624.169 0.015 0.020
Chain 1: 3200 -7842.184 0.016 0.020
Chain 1: 3300 -7497.942 0.018 0.020
Chain 1: 3400 -7623.924 0.019 0.020
Chain 1: 3500 -7551.297 0.018 0.017
Chain 1: 3600 -7565.055 0.018 0.017
Chain 1: 3700 -7547.962 0.018 0.017
Chain 1: 3800 -7494.803 0.017 0.017
Chain 1: 3900 -7491.161 0.014 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003277 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.77 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86021.087 1.000 1.000
Chain 1: 200 -14359.336 2.995 4.991
Chain 1: 300 -10560.321 2.117 1.000
Chain 1: 400 -12576.510 1.628 1.000
Chain 1: 500 -9230.062 1.375 0.363
Chain 1: 600 -9700.423 1.154 0.363
Chain 1: 700 -8872.296 1.002 0.360
Chain 1: 800 -9704.257 0.888 0.360
Chain 1: 900 -9244.994 0.794 0.160
Chain 1: 1000 -9554.691 0.718 0.160
Chain 1: 1100 -9360.197 0.620 0.093 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8909.915 0.126 0.086
Chain 1: 1300 -9158.787 0.093 0.051
Chain 1: 1400 -8997.729 0.079 0.050
Chain 1: 1500 -9025.763 0.043 0.048
Chain 1: 1600 -9116.760 0.039 0.032
Chain 1: 1700 -9150.592 0.030 0.027
Chain 1: 1800 -8694.878 0.027 0.027
Chain 1: 1900 -8806.681 0.023 0.021
Chain 1: 2000 -8826.705 0.020 0.018
Chain 1: 2100 -8911.119 0.019 0.013
Chain 1: 2200 -8689.580 0.016 0.013
Chain 1: 2300 -8898.027 0.016 0.013
Chain 1: 2400 -8697.622 0.017 0.013
Chain 1: 2500 -8775.315 0.017 0.013
Chain 1: 2600 -8682.934 0.017 0.013
Chain 1: 2700 -8720.301 0.017 0.013
Chain 1: 2800 -8672.078 0.013 0.011
Chain 1: 2900 -8786.213 0.013 0.011
Chain 1: 3000 -8695.534 0.013 0.011
Chain 1: 3100 -8662.809 0.013 0.011
Chain 1: 3200 -8633.849 0.011 0.010
Chain 1: 3300 -8897.314 0.011 0.010
Chain 1: 3400 -8943.804 0.009 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002861 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 28.61 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8378518.665 1.000 1.000
Chain 1: 200 -1579452.845 2.652 4.305
Chain 1: 300 -892437.846 2.025 1.000
Chain 1: 400 -459734.019 1.754 1.000
Chain 1: 500 -360512.108 1.458 0.941
Chain 1: 600 -235208.432 1.304 0.941
Chain 1: 700 -120787.055 1.253 0.941
Chain 1: 800 -87864.060 1.143 0.941
Chain 1: 900 -68077.469 1.048 0.770
Chain 1: 1000 -52784.723 0.973 0.770
Chain 1: 1100 -40170.595 0.904 0.533 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39341.061 0.476 0.375
Chain 1: 1300 -27185.474 0.443 0.375
Chain 1: 1400 -26898.017 0.350 0.314
Chain 1: 1500 -23456.514 0.337 0.314
Chain 1: 1600 -22665.936 0.288 0.291
Chain 1: 1700 -21525.001 0.198 0.290
Chain 1: 1800 -21466.331 0.161 0.147
Chain 1: 1900 -21793.346 0.133 0.053
Chain 1: 2000 -20295.330 0.112 0.053
Chain 1: 2100 -20534.053 0.082 0.035
Chain 1: 2200 -20762.594 0.081 0.035
Chain 1: 2300 -20377.773 0.038 0.019
Chain 1: 2400 -20149.390 0.038 0.019
Chain 1: 2500 -19952.003 0.024 0.015
Chain 1: 2600 -19580.699 0.023 0.015
Chain 1: 2700 -19537.136 0.018 0.012
Chain 1: 2800 -19253.903 0.019 0.015
Chain 1: 2900 -19535.682 0.019 0.014
Chain 1: 3000 -19521.665 0.011 0.012
Chain 1: 3100 -19606.868 0.011 0.011
Chain 1: 3200 -19296.776 0.011 0.014
Chain 1: 3300 -19502.067 0.010 0.011
Chain 1: 3400 -18975.888 0.012 0.014
Chain 1: 3500 -19589.669 0.014 0.015
Chain 1: 3600 -18893.869 0.016 0.015
Chain 1: 3700 -19282.631 0.018 0.016
Chain 1: 3800 -18238.678 0.022 0.020
Chain 1: 3900 -18234.809 0.021 0.020
Chain 1: 4000 -18352.026 0.021 0.020
Chain 1: 4100 -18265.714 0.021 0.020
Chain 1: 4200 -18081.081 0.021 0.020
Chain 1: 4300 -18220.011 0.020 0.020
Chain 1: 4400 -18176.160 0.018 0.010
Chain 1: 4500 -18078.656 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00162 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 16.2 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49367.616 1.000 1.000
Chain 1: 200 -19307.544 1.278 1.557
Chain 1: 300 -44460.798 1.041 1.000
Chain 1: 400 -19484.853 1.101 1.282
Chain 1: 500 -13944.468 0.960 1.000
Chain 1: 600 -12840.473 0.815 1.000
Chain 1: 700 -18743.711 0.743 0.566
Chain 1: 800 -12937.575 0.706 0.566
Chain 1: 900 -14305.723 0.639 0.449
Chain 1: 1000 -11100.828 0.604 0.449
Chain 1: 1100 -16831.009 0.538 0.397 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -16744.097 0.382 0.340
Chain 1: 1300 -11506.578 0.371 0.340
Chain 1: 1400 -10602.066 0.252 0.315
Chain 1: 1500 -13140.203 0.231 0.289
Chain 1: 1600 -15886.200 0.240 0.289
Chain 1: 1700 -14902.563 0.215 0.193
Chain 1: 1800 -11651.698 0.198 0.193
Chain 1: 1900 -19556.609 0.229 0.279
Chain 1: 2000 -25432.263 0.223 0.231
Chain 1: 2100 -11208.389 0.316 0.231
Chain 1: 2200 -11330.252 0.317 0.231
Chain 1: 2300 -13006.998 0.284 0.193
Chain 1: 2400 -10153.015 0.304 0.231
Chain 1: 2500 -16848.029 0.324 0.279
Chain 1: 2600 -10458.050 0.368 0.281
Chain 1: 2700 -9467.352 0.372 0.281
Chain 1: 2800 -10972.544 0.358 0.281
Chain 1: 2900 -10091.458 0.326 0.231
Chain 1: 3000 -9560.271 0.308 0.137
Chain 1: 3100 -14028.288 0.213 0.137
Chain 1: 3200 -13849.478 0.213 0.137
Chain 1: 3300 -12210.983 0.214 0.137
Chain 1: 3400 -9749.027 0.211 0.137
Chain 1: 3500 -11294.731 0.185 0.137
Chain 1: 3600 -11304.015 0.124 0.134
Chain 1: 3700 -8910.127 0.140 0.137
Chain 1: 3800 -9039.185 0.128 0.134
Chain 1: 3900 -9509.121 0.124 0.134
Chain 1: 4000 -9245.304 0.122 0.134
Chain 1: 4100 -10067.929 0.098 0.082
Chain 1: 4200 -9846.063 0.099 0.082
Chain 1: 4300 -10406.334 0.091 0.054
Chain 1: 4400 -9091.291 0.080 0.054
Chain 1: 4500 -9254.691 0.068 0.049
Chain 1: 4600 -10893.463 0.083 0.054
Chain 1: 4700 -9534.400 0.071 0.054
Chain 1: 4800 -9297.178 0.072 0.054
Chain 1: 4900 -9553.865 0.069 0.054
Chain 1: 5000 -12314.824 0.089 0.082
Chain 1: 5100 -11810.238 0.085 0.054
Chain 1: 5200 -10171.951 0.099 0.143
Chain 1: 5300 -11455.921 0.105 0.143
Chain 1: 5400 -9548.213 0.110 0.143
Chain 1: 5500 -13010.092 0.135 0.150
Chain 1: 5600 -9356.816 0.159 0.161
Chain 1: 5700 -9340.197 0.145 0.161
Chain 1: 5800 -9375.968 0.143 0.161
Chain 1: 5900 -13492.843 0.171 0.200
Chain 1: 6000 -9313.728 0.193 0.200
Chain 1: 6100 -10344.659 0.199 0.200
Chain 1: 6200 -9235.436 0.195 0.200
Chain 1: 6300 -11364.261 0.202 0.200
Chain 1: 6400 -10490.438 0.191 0.187
Chain 1: 6500 -9405.486 0.176 0.120
Chain 1: 6600 -8730.129 0.144 0.115
Chain 1: 6700 -9470.239 0.152 0.115
Chain 1: 6800 -9173.120 0.155 0.115
Chain 1: 6900 -8740.022 0.129 0.100
Chain 1: 7000 -8780.542 0.085 0.083
Chain 1: 7100 -9632.638 0.084 0.083
Chain 1: 7200 -8910.834 0.080 0.081
Chain 1: 7300 -12061.213 0.087 0.081
Chain 1: 7400 -13111.049 0.087 0.080
Chain 1: 7500 -8716.708 0.126 0.080
Chain 1: 7600 -11764.912 0.144 0.081
Chain 1: 7700 -12706.498 0.143 0.081
Chain 1: 7800 -8714.745 0.186 0.088
Chain 1: 7900 -8757.695 0.182 0.088
Chain 1: 8000 -8730.920 0.181 0.088
Chain 1: 8100 -10755.573 0.191 0.188
Chain 1: 8200 -8890.714 0.204 0.210
Chain 1: 8300 -8635.263 0.181 0.188
Chain 1: 8400 -8451.930 0.175 0.188
Chain 1: 8500 -8789.420 0.129 0.074
Chain 1: 8600 -9310.537 0.108 0.056
Chain 1: 8700 -9067.072 0.104 0.038
Chain 1: 8800 -8574.289 0.064 0.038
Chain 1: 8900 -8790.367 0.066 0.038
Chain 1: 9000 -11557.967 0.089 0.056
Chain 1: 9100 -9753.530 0.089 0.056
Chain 1: 9200 -8974.893 0.077 0.056
Chain 1: 9300 -8481.425 0.079 0.057
Chain 1: 9400 -8608.644 0.079 0.057
Chain 1: 9500 -11697.920 0.101 0.058
Chain 1: 9600 -8664.531 0.131 0.087
Chain 1: 9700 -8431.748 0.131 0.087
Chain 1: 9800 -11140.581 0.149 0.185
Chain 1: 9900 -9641.983 0.162 0.185
Chain 1: 10000 -8596.468 0.151 0.155
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001644 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 16.44 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57953.227 1.000 1.000
Chain 1: 200 -18165.153 1.595 2.190
Chain 1: 300 -9060.987 1.398 1.005
Chain 1: 400 -8224.519 1.074 1.005
Chain 1: 500 -8602.314 0.868 1.000
Chain 1: 600 -9182.322 0.734 1.000
Chain 1: 700 -9211.175 0.630 0.102
Chain 1: 800 -8383.710 0.563 0.102
Chain 1: 900 -8210.064 0.503 0.099
Chain 1: 1000 -8240.273 0.453 0.099
Chain 1: 1100 -7874.540 0.358 0.063
Chain 1: 1200 -7719.071 0.141 0.046
Chain 1: 1300 -7737.073 0.040 0.044
Chain 1: 1400 -7786.893 0.031 0.021
Chain 1: 1500 -7689.229 0.028 0.020
Chain 1: 1600 -7780.552 0.023 0.013
Chain 1: 1700 -7818.994 0.023 0.013
Chain 1: 1800 -7676.921 0.015 0.013
Chain 1: 1900 -7651.650 0.013 0.012
Chain 1: 2000 -7876.895 0.016 0.013
Chain 1: 2100 -7687.875 0.013 0.013
Chain 1: 2200 -7810.934 0.013 0.013
Chain 1: 2300 -7581.513 0.016 0.016
Chain 1: 2400 -7660.712 0.016 0.016
Chain 1: 2500 -7615.382 0.015 0.016
Chain 1: 2600 -7591.092 0.015 0.016
Chain 1: 2700 -7498.876 0.015 0.016
Chain 1: 2800 -7717.891 0.016 0.016
Chain 1: 2900 -7465.670 0.019 0.025
Chain 1: 3000 -7594.925 0.018 0.017
Chain 1: 3100 -7594.705 0.016 0.016
Chain 1: 3200 -7800.348 0.017 0.017
Chain 1: 3300 -7499.721 0.018 0.017
Chain 1: 3400 -7719.456 0.020 0.026
Chain 1: 3500 -7496.331 0.022 0.028
Chain 1: 3600 -7561.512 0.022 0.028
Chain 1: 3700 -7513.084 0.022 0.028
Chain 1: 3800 -7487.601 0.019 0.026
Chain 1: 3900 -7463.091 0.016 0.017
Chain 1: 4000 -7459.512 0.015 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003362 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.62 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86260.577 1.000 1.000
Chain 1: 200 -14129.249 3.053 5.105
Chain 1: 300 -10352.887 2.157 1.000
Chain 1: 400 -11961.999 1.651 1.000
Chain 1: 500 -9178.615 1.382 0.365
Chain 1: 600 -8776.252 1.159 0.365
Chain 1: 700 -8612.447 0.996 0.303
Chain 1: 800 -9362.638 0.882 0.303
Chain 1: 900 -8995.367 0.788 0.135
Chain 1: 1000 -9345.028 0.713 0.135
Chain 1: 1100 -9052.417 0.616 0.080 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8650.978 0.110 0.046
Chain 1: 1300 -8958.756 0.077 0.046
Chain 1: 1400 -8893.322 0.065 0.041
Chain 1: 1500 -8841.421 0.035 0.037
Chain 1: 1600 -8919.690 0.031 0.034
Chain 1: 1700 -8970.694 0.030 0.034
Chain 1: 1800 -8513.707 0.027 0.034
Chain 1: 1900 -8623.004 0.024 0.032
Chain 1: 2000 -8635.472 0.021 0.013
Chain 1: 2100 -8727.192 0.019 0.011
Chain 1: 2200 -8512.742 0.017 0.011
Chain 1: 2300 -8713.199 0.015 0.011
Chain 1: 2400 -8518.508 0.017 0.013
Chain 1: 2500 -8594.462 0.017 0.013
Chain 1: 2600 -8504.448 0.017 0.013
Chain 1: 2700 -8537.506 0.017 0.013
Chain 1: 2800 -8488.644 0.012 0.011
Chain 1: 2900 -8602.884 0.013 0.011
Chain 1: 3000 -8520.593 0.013 0.011
Chain 1: 3100 -8480.868 0.013 0.011
Chain 1: 3200 -8453.272 0.011 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.0032 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8371918.593 1.000 1.000
Chain 1: 200 -1578242.662 2.652 4.305
Chain 1: 300 -890614.018 2.026 1.000
Chain 1: 400 -458126.056 1.755 1.000
Chain 1: 500 -359243.926 1.459 0.944
Chain 1: 600 -234215.903 1.305 0.944
Chain 1: 700 -120207.896 1.254 0.944
Chain 1: 800 -87387.348 1.144 0.944
Chain 1: 900 -67673.800 1.049 0.772
Chain 1: 1000 -52432.990 0.974 0.772
Chain 1: 1100 -39862.927 0.905 0.534 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39040.350 0.477 0.376
Chain 1: 1300 -26924.754 0.445 0.376
Chain 1: 1400 -26641.123 0.351 0.315
Chain 1: 1500 -23209.742 0.338 0.315
Chain 1: 1600 -22422.491 0.289 0.291
Chain 1: 1700 -21286.101 0.199 0.291
Chain 1: 1800 -21228.598 0.162 0.148
Chain 1: 1900 -21555.625 0.134 0.053
Chain 1: 2000 -20060.184 0.113 0.053
Chain 1: 2100 -20298.790 0.082 0.035
Chain 1: 2200 -20526.864 0.081 0.035
Chain 1: 2300 -20142.450 0.038 0.019
Chain 1: 2400 -19914.150 0.038 0.019
Chain 1: 2500 -19716.634 0.024 0.015
Chain 1: 2600 -19345.561 0.023 0.015
Chain 1: 2700 -19302.136 0.018 0.012
Chain 1: 2800 -19018.862 0.019 0.015
Chain 1: 2900 -19300.573 0.019 0.015
Chain 1: 3000 -19286.547 0.012 0.012
Chain 1: 3100 -19371.730 0.011 0.011
Chain 1: 3200 -19061.768 0.011 0.015
Chain 1: 3300 -19267.006 0.010 0.011
Chain 1: 3400 -18740.964 0.012 0.015
Chain 1: 3500 -19354.490 0.014 0.015
Chain 1: 3600 -18659.022 0.016 0.015
Chain 1: 3700 -19047.503 0.018 0.016
Chain 1: 3800 -18004.040 0.022 0.020
Chain 1: 3900 -18000.181 0.021 0.020
Chain 1: 4000 -18117.399 0.021 0.020
Chain 1: 4100 -18031.084 0.021 0.020
Chain 1: 4200 -17846.606 0.021 0.020
Chain 1: 4300 -17985.456 0.020 0.020
Chain 1: 4400 -17941.700 0.018 0.010
Chain 1: 4500 -17844.185 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001463 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.63 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13485.905 1.000 1.000
Chain 1: 200 -10274.168 0.656 1.000
Chain 1: 300 -8663.582 0.500 0.313
Chain 1: 400 -8889.214 0.381 0.313
Chain 1: 500 -8861.708 0.305 0.186
Chain 1: 600 -8543.504 0.261 0.186
Chain 1: 700 -8469.841 0.225 0.037
Chain 1: 800 -8485.648 0.197 0.037
Chain 1: 900 -8478.849 0.175 0.025
Chain 1: 1000 -8570.879 0.159 0.025
Chain 1: 1100 -8535.841 0.059 0.011
Chain 1: 1200 -8510.311 0.028 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001571 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.71 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -47480.734 1.000 1.000
Chain 1: 200 -16812.639 1.412 1.824
Chain 1: 300 -9387.854 1.205 1.000
Chain 1: 400 -8320.834 0.936 1.000
Chain 1: 500 -8952.255 0.763 0.791
Chain 1: 600 -9150.525 0.639 0.791
Chain 1: 700 -8214.323 0.564 0.128
Chain 1: 800 -8840.478 0.503 0.128
Chain 1: 900 -7860.448 0.461 0.125
Chain 1: 1000 -7995.745 0.416 0.125
Chain 1: 1100 -7916.101 0.317 0.114
Chain 1: 1200 -7880.490 0.135 0.071
Chain 1: 1300 -8062.863 0.058 0.071
Chain 1: 1400 -7959.356 0.047 0.023
Chain 1: 1500 -7685.207 0.043 0.023
Chain 1: 1600 -7851.484 0.043 0.023
Chain 1: 1700 -7783.840 0.033 0.021
Chain 1: 1800 -7715.878 0.027 0.017
Chain 1: 1900 -7700.825 0.014 0.013
Chain 1: 2000 -7800.137 0.014 0.013
Chain 1: 2100 -7672.955 0.015 0.013
Chain 1: 2200 -8058.281 0.019 0.017
Chain 1: 2300 -7745.929 0.021 0.017
Chain 1: 2400 -7723.389 0.020 0.017
Chain 1: 2500 -7672.510 0.017 0.013
Chain 1: 2600 -7762.398 0.016 0.012
Chain 1: 2700 -7553.345 0.018 0.013
Chain 1: 2800 -7640.650 0.018 0.013
Chain 1: 2900 -7473.741 0.020 0.017
Chain 1: 3000 -7712.157 0.022 0.022
Chain 1: 3100 -7627.162 0.021 0.022
Chain 1: 3200 -7741.725 0.018 0.015
Chain 1: 3300 -7489.084 0.017 0.015
Chain 1: 3400 -7886.495 0.022 0.022
Chain 1: 3500 -7548.085 0.026 0.028
Chain 1: 3600 -7693.736 0.027 0.028
Chain 1: 3700 -7494.050 0.027 0.027
Chain 1: 3800 -7601.762 0.027 0.027
Chain 1: 3900 -7506.880 0.026 0.027
Chain 1: 4000 -7496.768 0.023 0.019
Chain 1: 4100 -7498.999 0.022 0.019
Chain 1: 4200 -7704.651 0.023 0.027
Chain 1: 4300 -7484.971 0.023 0.027
Chain 1: 4400 -7539.070 0.018 0.019
Chain 1: 4500 -7671.089 0.015 0.017
Chain 1: 4600 -7559.253 0.015 0.015
Chain 1: 4700 -7551.055 0.012 0.014
Chain 1: 4800 -7514.665 0.012 0.013
Chain 1: 4900 -7785.496 0.014 0.015
Chain 1: 5000 -7704.600 0.015 0.015
Chain 1: 5100 -7578.995 0.016 0.017
Chain 1: 5200 -7598.266 0.014 0.015
Chain 1: 5300 -7585.537 0.011 0.010
Chain 1: 5400 -7547.774 0.011 0.010
Chain 1: 5500 -7484.569 0.010 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.0036 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 36 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87810.250 1.000 1.000
Chain 1: 200 -14716.658 2.983 4.967
Chain 1: 300 -10818.810 2.109 1.000
Chain 1: 400 -13283.097 1.628 1.000
Chain 1: 500 -9145.000 1.393 0.452
Chain 1: 600 -9562.432 1.168 0.452
Chain 1: 700 -9515.944 1.002 0.360
Chain 1: 800 -8925.540 0.885 0.360
Chain 1: 900 -8948.139 0.787 0.186
Chain 1: 1000 -9800.758 0.717 0.186
Chain 1: 1100 -9292.357 0.622 0.087 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -9597.169 0.129 0.066
Chain 1: 1300 -8949.732 0.100 0.066
Chain 1: 1400 -9021.377 0.082 0.055
Chain 1: 1500 -9126.631 0.038 0.044
Chain 1: 1600 -9053.015 0.035 0.032
Chain 1: 1700 -8917.940 0.036 0.032
Chain 1: 1800 -8958.796 0.030 0.015
Chain 1: 1900 -8954.754 0.029 0.015
Chain 1: 2000 -9198.870 0.023 0.015
Chain 1: 2100 -8899.318 0.021 0.015
Chain 1: 2200 -8862.709 0.018 0.012
Chain 1: 2300 -9120.045 0.014 0.012
Chain 1: 2400 -8840.079 0.016 0.015
Chain 1: 2500 -8920.004 0.016 0.015
Chain 1: 2600 -8824.790 0.016 0.015
Chain 1: 2700 -8843.031 0.015 0.011
Chain 1: 2800 -8702.731 0.016 0.016
Chain 1: 2900 -8889.832 0.018 0.021
Chain 1: 3000 -8797.579 0.017 0.016
Chain 1: 3100 -8896.043 0.014 0.011
Chain 1: 3200 -8762.314 0.016 0.015
Chain 1: 3300 -9030.874 0.016 0.015
Chain 1: 3400 -9100.411 0.013 0.011
Chain 1: 3500 -8908.685 0.015 0.015
Chain 1: 3600 -8714.786 0.016 0.016
Chain 1: 3700 -8884.364 0.017 0.019
Chain 1: 3800 -8718.634 0.018 0.019
Chain 1: 3900 -8944.082 0.018 0.019
Chain 1: 4000 -8945.841 0.017 0.019
Chain 1: 4100 -8729.651 0.018 0.022
Chain 1: 4200 -8714.644 0.017 0.022
Chain 1: 4300 -8716.031 0.014 0.019
Chain 1: 4400 -8670.068 0.014 0.019
Chain 1: 4500 -8810.804 0.013 0.019
Chain 1: 4600 -8837.869 0.011 0.016
Chain 1: 4700 -8955.838 0.011 0.013
Chain 1: 4800 -8778.519 0.011 0.013
Chain 1: 4900 -8804.270 0.009 0.005 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004106 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 41.06 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8415537.539 1.000 1.000
Chain 1: 200 -1587904.242 2.650 4.300
Chain 1: 300 -893094.756 2.026 1.000
Chain 1: 400 -459302.957 1.756 1.000
Chain 1: 500 -359413.808 1.460 0.944
Chain 1: 600 -234348.606 1.306 0.944
Chain 1: 700 -120531.765 1.254 0.944
Chain 1: 800 -87718.312 1.144 0.944
Chain 1: 900 -68077.725 1.049 0.778
Chain 1: 1000 -52893.508 0.973 0.778
Chain 1: 1100 -40368.934 0.904 0.534 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39559.225 0.476 0.374
Chain 1: 1300 -27492.327 0.442 0.374
Chain 1: 1400 -27215.249 0.349 0.310
Chain 1: 1500 -23795.345 0.335 0.310
Chain 1: 1600 -23011.340 0.285 0.289
Chain 1: 1700 -21881.106 0.196 0.287
Chain 1: 1800 -21825.105 0.159 0.144
Chain 1: 1900 -22152.588 0.131 0.052
Chain 1: 2000 -20659.192 0.110 0.052
Chain 1: 2100 -20897.920 0.080 0.034
Chain 1: 2200 -21125.545 0.079 0.034
Chain 1: 2300 -20741.404 0.037 0.019
Chain 1: 2400 -20513.027 0.037 0.019
Chain 1: 2500 -20314.974 0.024 0.015
Chain 1: 2600 -19943.806 0.022 0.015
Chain 1: 2700 -19900.414 0.017 0.011
Chain 1: 2800 -19616.645 0.018 0.014
Chain 1: 2900 -19898.564 0.018 0.014
Chain 1: 3000 -19884.706 0.011 0.011
Chain 1: 3100 -19969.843 0.010 0.011
Chain 1: 3200 -19659.616 0.011 0.014
Chain 1: 3300 -19865.086 0.010 0.011
Chain 1: 3400 -19338.338 0.012 0.014
Chain 1: 3500 -19952.627 0.014 0.014
Chain 1: 3600 -19256.271 0.016 0.014
Chain 1: 3700 -19645.312 0.017 0.016
Chain 1: 3800 -18600.152 0.022 0.020
Chain 1: 3900 -18596.178 0.020 0.020
Chain 1: 4000 -18713.512 0.021 0.020
Chain 1: 4100 -18626.976 0.021 0.020
Chain 1: 4200 -18442.205 0.020 0.020
Chain 1: 4300 -18581.331 0.020 0.020
Chain 1: 4400 -18537.292 0.017 0.010
Chain 1: 4500 -18439.682 0.015 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001401 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.01 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49033.578 1.000 1.000
Chain 1: 200 -15317.652 1.601 2.201
Chain 1: 300 -40312.725 1.274 1.000
Chain 1: 400 -20541.841 1.196 1.000
Chain 1: 500 -18250.214 0.982 0.962
Chain 1: 600 -27355.850 0.874 0.962
Chain 1: 700 -15537.873 0.858 0.761
Chain 1: 800 -12422.909 0.782 0.761
Chain 1: 900 -15594.259 0.717 0.620
Chain 1: 1000 -10216.901 0.698 0.620
Chain 1: 1100 -10333.703 0.599 0.526 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -12299.175 0.395 0.333
Chain 1: 1300 -10092.829 0.355 0.251
Chain 1: 1400 -10765.747 0.265 0.219
Chain 1: 1500 -9953.331 0.261 0.219
Chain 1: 1600 -9915.489 0.228 0.203
Chain 1: 1700 -11450.518 0.165 0.160
Chain 1: 1800 -14438.830 0.161 0.160
Chain 1: 1900 -14991.828 0.144 0.134
Chain 1: 2000 -10209.498 0.138 0.134
Chain 1: 2100 -9743.422 0.142 0.134
Chain 1: 2200 -17986.022 0.172 0.134
Chain 1: 2300 -11312.215 0.209 0.134
Chain 1: 2400 -9352.430 0.224 0.207
Chain 1: 2500 -13622.213 0.247 0.210
Chain 1: 2600 -15657.942 0.260 0.210
Chain 1: 2700 -9401.296 0.313 0.313
Chain 1: 2800 -19970.564 0.345 0.458
Chain 1: 2900 -17278.377 0.357 0.458
Chain 1: 3000 -9168.905 0.398 0.458
Chain 1: 3100 -8669.545 0.399 0.458
Chain 1: 3200 -9311.327 0.360 0.313
Chain 1: 3300 -9983.320 0.308 0.210
Chain 1: 3400 -13224.062 0.312 0.245
Chain 1: 3500 -9769.572 0.316 0.245
Chain 1: 3600 -9275.973 0.308 0.245
Chain 1: 3700 -9210.837 0.242 0.156
Chain 1: 3800 -9137.678 0.190 0.069
Chain 1: 3900 -9005.091 0.176 0.067
Chain 1: 4000 -8965.534 0.088 0.058
Chain 1: 4100 -8804.605 0.084 0.053
Chain 1: 4200 -10764.062 0.095 0.053
Chain 1: 4300 -15579.211 0.120 0.053
Chain 1: 4400 -13750.780 0.108 0.053
Chain 1: 4500 -8931.705 0.127 0.053
Chain 1: 4600 -8785.715 0.123 0.018
Chain 1: 4700 -13062.665 0.155 0.133
Chain 1: 4800 -9158.824 0.197 0.182
Chain 1: 4900 -9819.009 0.202 0.182
Chain 1: 5000 -8538.042 0.217 0.182
Chain 1: 5100 -12611.533 0.247 0.309
Chain 1: 5200 -13600.423 0.236 0.309
Chain 1: 5300 -10756.359 0.232 0.264
Chain 1: 5400 -8602.836 0.244 0.264
Chain 1: 5500 -8701.959 0.191 0.250
Chain 1: 5600 -8661.865 0.190 0.250
Chain 1: 5700 -10866.471 0.177 0.203
Chain 1: 5800 -8930.200 0.156 0.203
Chain 1: 5900 -8750.492 0.152 0.203
Chain 1: 6000 -9918.583 0.148 0.203
Chain 1: 6100 -10339.674 0.120 0.118
Chain 1: 6200 -9671.375 0.120 0.118
Chain 1: 6300 -8662.858 0.105 0.116
Chain 1: 6400 -11478.307 0.105 0.116
Chain 1: 6500 -8650.413 0.136 0.118
Chain 1: 6600 -10430.856 0.153 0.171
Chain 1: 6700 -8444.483 0.156 0.171
Chain 1: 6800 -8565.786 0.136 0.118
Chain 1: 6900 -8591.227 0.134 0.118
Chain 1: 7000 -13337.581 0.158 0.171
Chain 1: 7100 -8219.727 0.216 0.235
Chain 1: 7200 -8273.923 0.210 0.235
Chain 1: 7300 -8224.368 0.199 0.235
Chain 1: 7400 -8931.125 0.182 0.171
Chain 1: 7500 -9302.089 0.153 0.079
Chain 1: 7600 -8771.710 0.142 0.060
Chain 1: 7700 -8434.524 0.123 0.040
Chain 1: 7800 -13018.288 0.157 0.060
Chain 1: 7900 -8392.939 0.211 0.079
Chain 1: 8000 -8223.276 0.178 0.060
Chain 1: 8100 -8249.379 0.116 0.040
Chain 1: 8200 -8423.052 0.117 0.040
Chain 1: 8300 -9872.282 0.131 0.060
Chain 1: 8400 -8314.319 0.142 0.060
Chain 1: 8500 -8844.196 0.144 0.060
Chain 1: 8600 -8996.570 0.140 0.060
Chain 1: 8700 -8649.523 0.140 0.060
Chain 1: 8800 -8484.561 0.107 0.040
Chain 1: 8900 -8315.326 0.054 0.021
Chain 1: 9000 -9219.963 0.061 0.040
Chain 1: 9100 -8294.086 0.072 0.060
Chain 1: 9200 -11674.583 0.099 0.098
Chain 1: 9300 -8671.808 0.119 0.098
Chain 1: 9400 -8400.640 0.103 0.060
Chain 1: 9500 -8555.513 0.099 0.040
Chain 1: 9600 -9173.940 0.104 0.067
Chain 1: 9700 -8213.339 0.112 0.098
Chain 1: 9800 -11581.859 0.139 0.112
Chain 1: 9900 -9661.816 0.157 0.117
Chain 1: 10000 -8407.274 0.162 0.149
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001421 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.21 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57962.502 1.000 1.000
Chain 1: 200 -17721.744 1.635 2.271
Chain 1: 300 -8727.270 1.434 1.031
Chain 1: 400 -8208.415 1.091 1.031
Chain 1: 500 -8468.724 0.879 1.000
Chain 1: 600 -8355.838 0.735 1.000
Chain 1: 700 -7878.070 0.638 0.063
Chain 1: 800 -8209.057 0.564 0.063
Chain 1: 900 -7871.289 0.506 0.061
Chain 1: 1000 -8074.266 0.458 0.061
Chain 1: 1100 -7752.530 0.362 0.043
Chain 1: 1200 -7703.819 0.135 0.042
Chain 1: 1300 -7718.750 0.033 0.040
Chain 1: 1400 -7980.745 0.030 0.033
Chain 1: 1500 -7607.736 0.031 0.040
Chain 1: 1600 -7821.230 0.033 0.040
Chain 1: 1700 -7560.228 0.030 0.035
Chain 1: 1800 -7658.808 0.027 0.033
Chain 1: 1900 -7589.737 0.024 0.027
Chain 1: 2000 -7648.321 0.022 0.027
Chain 1: 2100 -7619.633 0.019 0.013
Chain 1: 2200 -7742.151 0.019 0.016
Chain 1: 2300 -7646.157 0.021 0.016
Chain 1: 2400 -7688.879 0.018 0.013
Chain 1: 2500 -7602.127 0.014 0.013
Chain 1: 2600 -7567.983 0.012 0.011
Chain 1: 2700 -7536.972 0.009 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003244 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.44 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86433.861 1.000 1.000
Chain 1: 200 -13553.921 3.189 5.377
Chain 1: 300 -9907.369 2.248 1.000
Chain 1: 400 -10695.368 1.705 1.000
Chain 1: 500 -8901.461 1.404 0.368
Chain 1: 600 -8367.210 1.181 0.368
Chain 1: 700 -8547.365 1.015 0.202
Chain 1: 800 -8939.730 0.894 0.202
Chain 1: 900 -8690.118 0.798 0.074
Chain 1: 1000 -8558.926 0.719 0.074
Chain 1: 1100 -8721.751 0.621 0.064 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8238.439 0.089 0.059
Chain 1: 1300 -8605.030 0.057 0.044
Chain 1: 1400 -8593.708 0.050 0.043
Chain 1: 1500 -8463.518 0.031 0.029
Chain 1: 1600 -8570.639 0.026 0.021
Chain 1: 1700 -8649.802 0.025 0.019
Chain 1: 1800 -8229.072 0.025 0.019
Chain 1: 1900 -8328.287 0.024 0.015
Chain 1: 2000 -8302.494 0.022 0.015
Chain 1: 2100 -8427.199 0.022 0.015
Chain 1: 2200 -8235.189 0.019 0.015
Chain 1: 2300 -8323.007 0.015 0.012
Chain 1: 2400 -8392.231 0.016 0.012
Chain 1: 2500 -8338.328 0.015 0.012
Chain 1: 2600 -8339.095 0.014 0.011
Chain 1: 2700 -8256.088 0.014 0.011
Chain 1: 2800 -8216.870 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003436 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.36 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8424692.962 1.000 1.000
Chain 1: 200 -1590748.966 2.648 4.296
Chain 1: 300 -890922.571 2.027 1.000
Chain 1: 400 -457296.324 1.757 1.000
Chain 1: 500 -357035.046 1.462 0.948
Chain 1: 600 -232049.215 1.308 0.948
Chain 1: 700 -118781.307 1.258 0.948
Chain 1: 800 -86107.361 1.148 0.948
Chain 1: 900 -66554.249 1.053 0.786
Chain 1: 1000 -51437.985 0.977 0.786
Chain 1: 1100 -38991.271 0.909 0.539 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38179.398 0.481 0.379
Chain 1: 1300 -26210.403 0.449 0.379
Chain 1: 1400 -25937.263 0.355 0.319
Chain 1: 1500 -22543.235 0.342 0.319
Chain 1: 1600 -21765.658 0.291 0.294
Chain 1: 1700 -20648.091 0.202 0.294
Chain 1: 1800 -20594.423 0.164 0.151
Chain 1: 1900 -20920.693 0.136 0.054
Chain 1: 2000 -19436.205 0.114 0.054
Chain 1: 2100 -19674.428 0.084 0.036
Chain 1: 2200 -19900.144 0.083 0.036
Chain 1: 2300 -19517.966 0.039 0.020
Chain 1: 2400 -19290.093 0.039 0.020
Chain 1: 2500 -19091.788 0.025 0.016
Chain 1: 2600 -18722.266 0.023 0.016
Chain 1: 2700 -18679.387 0.018 0.012
Chain 1: 2800 -18396.014 0.019 0.015
Chain 1: 2900 -18677.214 0.019 0.015
Chain 1: 3000 -18663.516 0.012 0.012
Chain 1: 3100 -18748.468 0.011 0.012
Chain 1: 3200 -18439.237 0.012 0.015
Chain 1: 3300 -18643.917 0.011 0.012
Chain 1: 3400 -18118.789 0.012 0.015
Chain 1: 3500 -18730.616 0.015 0.015
Chain 1: 3600 -18037.361 0.017 0.015
Chain 1: 3700 -18424.008 0.018 0.017
Chain 1: 3800 -17383.748 0.023 0.021
Chain 1: 3900 -17379.845 0.021 0.021
Chain 1: 4000 -17497.212 0.022 0.021
Chain 1: 4100 -17410.908 0.022 0.021
Chain 1: 4200 -17227.205 0.021 0.021
Chain 1: 4300 -17365.620 0.021 0.021
Chain 1: 4400 -17322.449 0.018 0.011
Chain 1: 4500 -17224.934 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001448 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.48 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49491.608 1.000 1.000
Chain 1: 200 -14749.378 1.678 2.356
Chain 1: 300 -20788.481 1.215 1.000
Chain 1: 400 -17617.051 0.957 1.000
Chain 1: 500 -19403.462 0.784 0.291
Chain 1: 600 -12217.098 0.751 0.588
Chain 1: 700 -15261.798 0.672 0.291
Chain 1: 800 -15120.728 0.589 0.291
Chain 1: 900 -19917.827 0.551 0.241
Chain 1: 1000 -11742.565 0.565 0.291
Chain 1: 1100 -14819.790 0.486 0.241
Chain 1: 1200 -21887.615 0.283 0.241
Chain 1: 1300 -11198.247 0.349 0.241
Chain 1: 1400 -12221.266 0.339 0.241
Chain 1: 1500 -13193.407 0.338 0.241
Chain 1: 1600 -10161.145 0.309 0.241
Chain 1: 1700 -19942.747 0.338 0.298
Chain 1: 1800 -10365.432 0.429 0.323
Chain 1: 1900 -10763.200 0.409 0.323
Chain 1: 2000 -11208.541 0.343 0.298
Chain 1: 2100 -10652.350 0.328 0.298
Chain 1: 2200 -9997.728 0.302 0.084
Chain 1: 2300 -11848.730 0.222 0.084
Chain 1: 2400 -12318.444 0.218 0.074
Chain 1: 2500 -14659.911 0.226 0.156
Chain 1: 2600 -9576.468 0.249 0.156
Chain 1: 2700 -10071.108 0.205 0.065
Chain 1: 2800 -9245.288 0.122 0.065
Chain 1: 2900 -12565.100 0.144 0.089
Chain 1: 3000 -13006.130 0.144 0.089
Chain 1: 3100 -11310.301 0.154 0.150
Chain 1: 3200 -11533.943 0.149 0.150
Chain 1: 3300 -9243.371 0.158 0.150
Chain 1: 3400 -9633.935 0.158 0.150
Chain 1: 3500 -13135.554 0.169 0.150
Chain 1: 3600 -9020.033 0.162 0.150
Chain 1: 3700 -9608.385 0.163 0.150
Chain 1: 3800 -9012.036 0.161 0.150
Chain 1: 3900 -9111.399 0.135 0.066
Chain 1: 4000 -9744.623 0.138 0.066
Chain 1: 4100 -9014.141 0.131 0.066
Chain 1: 4200 -13960.819 0.165 0.081
Chain 1: 4300 -9833.825 0.182 0.081
Chain 1: 4400 -11730.668 0.194 0.162
Chain 1: 4500 -12391.337 0.173 0.081
Chain 1: 4600 -9178.539 0.162 0.081
Chain 1: 4700 -9381.464 0.158 0.081
Chain 1: 4800 -10251.294 0.160 0.085
Chain 1: 4900 -9365.577 0.169 0.095
Chain 1: 5000 -12933.333 0.190 0.162
Chain 1: 5100 -8642.981 0.231 0.276
Chain 1: 5200 -10612.574 0.214 0.186
Chain 1: 5300 -8664.810 0.195 0.186
Chain 1: 5400 -11851.105 0.206 0.225
Chain 1: 5500 -9243.311 0.228 0.269
Chain 1: 5600 -8791.675 0.199 0.225
Chain 1: 5700 -14980.755 0.238 0.269
Chain 1: 5800 -9232.892 0.292 0.276
Chain 1: 5900 -8379.725 0.292 0.276
Chain 1: 6000 -9242.872 0.274 0.269
Chain 1: 6100 -8856.059 0.229 0.225
Chain 1: 6200 -8481.369 0.215 0.225
Chain 1: 6300 -13075.237 0.227 0.269
Chain 1: 6400 -15549.069 0.216 0.159
Chain 1: 6500 -9162.657 0.258 0.159
Chain 1: 6600 -9776.712 0.259 0.159
Chain 1: 6700 -12467.809 0.239 0.159
Chain 1: 6800 -9611.627 0.207 0.159
Chain 1: 6900 -9911.547 0.199 0.159
Chain 1: 7000 -8602.887 0.205 0.159
Chain 1: 7100 -9611.808 0.211 0.159
Chain 1: 7200 -10594.886 0.216 0.159
Chain 1: 7300 -8670.458 0.203 0.159
Chain 1: 7400 -9457.551 0.196 0.152
Chain 1: 7500 -8354.252 0.139 0.132
Chain 1: 7600 -8671.664 0.137 0.132
Chain 1: 7700 -9214.127 0.121 0.105
Chain 1: 7800 -10097.439 0.100 0.093
Chain 1: 7900 -8249.109 0.119 0.105
Chain 1: 8000 -10422.212 0.125 0.105
Chain 1: 8100 -8462.585 0.138 0.132
Chain 1: 8200 -8986.665 0.134 0.132
Chain 1: 8300 -8328.653 0.120 0.087
Chain 1: 8400 -8408.076 0.113 0.087
Chain 1: 8500 -11488.716 0.126 0.087
Chain 1: 8600 -9625.544 0.142 0.194
Chain 1: 8700 -8559.393 0.148 0.194
Chain 1: 8800 -8196.222 0.144 0.194
Chain 1: 8900 -10647.147 0.145 0.194
Chain 1: 9000 -9379.160 0.137 0.135
Chain 1: 9100 -9339.605 0.115 0.125
Chain 1: 9200 -8448.489 0.119 0.125
Chain 1: 9300 -8741.840 0.115 0.125
Chain 1: 9400 -10831.924 0.133 0.135
Chain 1: 9500 -8408.072 0.135 0.135
Chain 1: 9600 -10210.825 0.134 0.135
Chain 1: 9700 -8593.327 0.140 0.177
Chain 1: 9800 -8920.171 0.139 0.177
Chain 1: 9900 -8315.647 0.123 0.135
Chain 1: 10000 -8139.242 0.112 0.105
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001857 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 18.57 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57249.588 1.000 1.000
Chain 1: 200 -17775.554 1.610 2.221
Chain 1: 300 -8868.457 1.408 1.004
Chain 1: 400 -8201.968 1.077 1.004
Chain 1: 500 -9225.938 0.883 1.000
Chain 1: 600 -8863.802 0.743 1.000
Chain 1: 700 -8586.042 0.642 0.111
Chain 1: 800 -8130.749 0.568 0.111
Chain 1: 900 -7871.603 0.509 0.081
Chain 1: 1000 -7663.582 0.461 0.081
Chain 1: 1100 -7697.449 0.361 0.056
Chain 1: 1200 -7646.655 0.140 0.041
Chain 1: 1300 -7729.139 0.040 0.033
Chain 1: 1400 -7861.446 0.034 0.032
Chain 1: 1500 -7521.424 0.027 0.032
Chain 1: 1600 -7705.795 0.026 0.027
Chain 1: 1700 -7432.279 0.026 0.027
Chain 1: 1800 -7657.358 0.023 0.027
Chain 1: 1900 -7718.026 0.021 0.024
Chain 1: 2000 -7690.137 0.019 0.017
Chain 1: 2100 -7570.517 0.020 0.017
Chain 1: 2200 -7792.709 0.022 0.024
Chain 1: 2300 -7506.834 0.025 0.029
Chain 1: 2400 -7552.470 0.024 0.029
Chain 1: 2500 -7545.445 0.019 0.024
Chain 1: 2600 -7498.189 0.017 0.016
Chain 1: 2700 -7453.293 0.014 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003645 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 36.45 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87109.163 1.000 1.000
Chain 1: 200 -13810.024 3.154 5.308
Chain 1: 300 -10018.127 2.229 1.000
Chain 1: 400 -11746.118 1.708 1.000
Chain 1: 500 -8514.408 1.443 0.380
Chain 1: 600 -8277.745 1.207 0.380
Chain 1: 700 -8777.783 1.043 0.379
Chain 1: 800 -9098.504 0.917 0.379
Chain 1: 900 -8709.753 0.820 0.147
Chain 1: 1000 -8577.129 0.739 0.147
Chain 1: 1100 -8799.626 0.642 0.057 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8285.654 0.117 0.057
Chain 1: 1300 -8612.201 0.083 0.045
Chain 1: 1400 -8488.472 0.070 0.038
Chain 1: 1500 -8492.593 0.032 0.035
Chain 1: 1600 -8577.880 0.030 0.035
Chain 1: 1700 -8627.459 0.025 0.025
Chain 1: 1800 -8169.709 0.027 0.025
Chain 1: 1900 -8279.887 0.024 0.015
Chain 1: 2000 -8296.374 0.023 0.015
Chain 1: 2100 -8423.775 0.022 0.015
Chain 1: 2200 -8174.823 0.019 0.015
Chain 1: 2300 -8358.207 0.017 0.015
Chain 1: 2400 -8174.480 0.018 0.015
Chain 1: 2500 -8250.398 0.019 0.015
Chain 1: 2600 -8159.503 0.019 0.015
Chain 1: 2700 -8193.983 0.019 0.015
Chain 1: 2800 -8144.856 0.014 0.013
Chain 1: 2900 -8259.488 0.014 0.014
Chain 1: 3000 -8173.258 0.015 0.014
Chain 1: 3100 -8137.037 0.013 0.011
Chain 1: 3200 -8108.881 0.011 0.011
Chain 1: 3300 -8368.683 0.012 0.011
Chain 1: 3400 -8410.004 0.010 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003062 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.62 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8418607.244 1.000 1.000
Chain 1: 200 -1588693.889 2.650 4.299
Chain 1: 300 -891540.864 2.027 1.000
Chain 1: 400 -458194.208 1.757 1.000
Chain 1: 500 -358203.594 1.461 0.946
Chain 1: 600 -233010.548 1.307 0.946
Chain 1: 700 -119360.781 1.256 0.946
Chain 1: 800 -86654.005 1.147 0.946
Chain 1: 900 -67030.103 1.052 0.782
Chain 1: 1000 -51870.762 0.976 0.782
Chain 1: 1100 -39383.372 0.907 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38570.893 0.480 0.377
Chain 1: 1300 -26542.150 0.447 0.377
Chain 1: 1400 -26267.285 0.353 0.317
Chain 1: 1500 -22858.633 0.340 0.317
Chain 1: 1600 -22077.985 0.290 0.293
Chain 1: 1700 -20952.124 0.200 0.292
Chain 1: 1800 -20897.100 0.163 0.149
Chain 1: 1900 -21224.145 0.135 0.054
Chain 1: 2000 -19734.202 0.113 0.054
Chain 1: 2100 -19972.579 0.083 0.035
Chain 1: 2200 -20199.704 0.082 0.035
Chain 1: 2300 -19816.118 0.038 0.019
Chain 1: 2400 -19587.888 0.039 0.019
Chain 1: 2500 -19389.966 0.025 0.015
Chain 1: 2600 -19019.229 0.023 0.015
Chain 1: 2700 -18975.958 0.018 0.012
Chain 1: 2800 -18692.476 0.019 0.015
Chain 1: 2900 -18974.096 0.019 0.015
Chain 1: 3000 -18960.178 0.012 0.012
Chain 1: 3100 -19045.310 0.011 0.012
Chain 1: 3200 -18735.427 0.011 0.015
Chain 1: 3300 -18940.615 0.011 0.012
Chain 1: 3400 -18414.527 0.012 0.015
Chain 1: 3500 -19027.918 0.015 0.015
Chain 1: 3600 -18332.602 0.016 0.015
Chain 1: 3700 -18720.863 0.018 0.017
Chain 1: 3800 -17677.496 0.023 0.021
Chain 1: 3900 -17673.558 0.021 0.021
Chain 1: 4000 -17790.874 0.022 0.021
Chain 1: 4100 -17704.490 0.022 0.021
Chain 1: 4200 -17520.047 0.021 0.021
Chain 1: 4300 -17658.910 0.021 0.021
Chain 1: 4400 -17615.163 0.018 0.011
Chain 1: 4500 -17517.604 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001542 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.42 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12478.725 1.000 1.000
Chain 1: 200 -9082.667 0.687 1.000
Chain 1: 300 -7934.756 0.506 0.374
Chain 1: 400 -8105.732 0.385 0.374
Chain 1: 500 -8014.493 0.310 0.145
Chain 1: 600 -7883.697 0.261 0.145
Chain 1: 700 -7809.005 0.225 0.021
Chain 1: 800 -7783.734 0.198 0.021
Chain 1: 900 -7700.403 0.177 0.017
Chain 1: 1000 -7867.058 0.161 0.021
Chain 1: 1100 -7915.099 0.062 0.017
Chain 1: 1200 -7849.856 0.025 0.011
Chain 1: 1300 -7776.506 0.012 0.011
Chain 1: 1400 -7779.891 0.010 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001657 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 16.57 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -51356.814 1.000 1.000
Chain 1: 200 -16461.228 1.560 2.120
Chain 1: 300 -8563.184 1.347 1.000
Chain 1: 400 -8371.518 1.016 1.000
Chain 1: 500 -8577.804 0.818 0.922
Chain 1: 600 -8326.706 0.687 0.922
Chain 1: 700 -8332.660 0.589 0.030
Chain 1: 800 -7939.270 0.521 0.050
Chain 1: 900 -7901.978 0.464 0.030
Chain 1: 1000 -7677.947 0.420 0.030
Chain 1: 1100 -7682.362 0.320 0.029
Chain 1: 1200 -7606.749 0.109 0.024
Chain 1: 1300 -7623.367 0.017 0.023
Chain 1: 1400 -7836.468 0.018 0.024
Chain 1: 1500 -7519.223 0.020 0.027
Chain 1: 1600 -7733.382 0.019 0.027
Chain 1: 1700 -7484.146 0.023 0.028
Chain 1: 1800 -7527.080 0.018 0.027
Chain 1: 1900 -7527.669 0.018 0.027
Chain 1: 2000 -7583.076 0.016 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003463 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.63 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85814.503 1.000 1.000
Chain 1: 200 -13467.323 3.186 5.372
Chain 1: 300 -9808.793 2.248 1.000
Chain 1: 400 -10684.933 1.707 1.000
Chain 1: 500 -8791.532 1.408 0.373
Chain 1: 600 -8405.759 1.181 0.373
Chain 1: 700 -8536.518 1.015 0.215
Chain 1: 800 -8588.678 0.889 0.215
Chain 1: 900 -8689.663 0.791 0.082
Chain 1: 1000 -8434.324 0.715 0.082
Chain 1: 1100 -8589.091 0.617 0.046 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8142.834 0.085 0.046
Chain 1: 1300 -8458.382 0.052 0.037
Chain 1: 1400 -8503.669 0.044 0.030
Chain 1: 1500 -8360.841 0.024 0.018
Chain 1: 1600 -8476.004 0.021 0.017
Chain 1: 1700 -8550.542 0.020 0.017
Chain 1: 1800 -8128.918 0.025 0.018
Chain 1: 1900 -8228.490 0.025 0.018
Chain 1: 2000 -8203.146 0.022 0.017
Chain 1: 2100 -8328.435 0.022 0.015
Chain 1: 2200 -8133.189 0.019 0.015
Chain 1: 2300 -8223.612 0.016 0.014
Chain 1: 2400 -8292.570 0.016 0.014
Chain 1: 2500 -8238.742 0.015 0.012
Chain 1: 2600 -8239.842 0.014 0.011
Chain 1: 2700 -8156.711 0.014 0.011
Chain 1: 2800 -8116.945 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003858 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 38.58 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8397949.715 1.000 1.000
Chain 1: 200 -1585313.105 2.649 4.297
Chain 1: 300 -891292.738 2.025 1.000
Chain 1: 400 -457567.496 1.756 1.000
Chain 1: 500 -357853.864 1.461 0.948
Chain 1: 600 -232957.894 1.306 0.948
Chain 1: 700 -119213.977 1.256 0.948
Chain 1: 800 -86412.200 1.147 0.948
Chain 1: 900 -66759.440 1.052 0.779
Chain 1: 1000 -51561.326 0.976 0.779
Chain 1: 1100 -39036.152 0.908 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38216.113 0.481 0.380
Chain 1: 1300 -26169.610 0.449 0.380
Chain 1: 1400 -25889.680 0.355 0.321
Chain 1: 1500 -22475.309 0.342 0.321
Chain 1: 1600 -21691.481 0.292 0.295
Chain 1: 1700 -20564.842 0.203 0.294
Chain 1: 1800 -20509.133 0.165 0.152
Chain 1: 1900 -20835.330 0.137 0.055
Chain 1: 2000 -19345.997 0.115 0.055
Chain 1: 2100 -19584.523 0.084 0.036
Chain 1: 2200 -19810.972 0.083 0.036
Chain 1: 2300 -19428.183 0.039 0.020
Chain 1: 2400 -19200.218 0.039 0.020
Chain 1: 2500 -19002.177 0.025 0.016
Chain 1: 2600 -18632.335 0.024 0.016
Chain 1: 2700 -18589.325 0.018 0.012
Chain 1: 2800 -18306.035 0.020 0.015
Chain 1: 2900 -18587.402 0.020 0.015
Chain 1: 3000 -18573.635 0.012 0.012
Chain 1: 3100 -18658.585 0.011 0.012
Chain 1: 3200 -18349.227 0.012 0.015
Chain 1: 3300 -18554.002 0.011 0.012
Chain 1: 3400 -18028.778 0.013 0.015
Chain 1: 3500 -18640.819 0.015 0.015
Chain 1: 3600 -17947.357 0.017 0.015
Chain 1: 3700 -18334.226 0.019 0.017
Chain 1: 3800 -17293.643 0.023 0.021
Chain 1: 3900 -17289.768 0.022 0.021
Chain 1: 4000 -17407.105 0.022 0.021
Chain 1: 4100 -17320.783 0.022 0.021
Chain 1: 4200 -17137.012 0.022 0.021
Chain 1: 4300 -17275.441 0.021 0.021
Chain 1: 4400 -17232.214 0.019 0.011
Chain 1: 4500 -17134.738 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001972 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 19.72 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49595.396 1.000 1.000
Chain 1: 200 -19159.137 1.294 1.589
Chain 1: 300 -21153.884 0.894 1.000
Chain 1: 400 -17517.318 0.723 1.000
Chain 1: 500 -12803.073 0.652 0.368
Chain 1: 600 -12160.038 0.552 0.368
Chain 1: 700 -16689.060 0.512 0.271
Chain 1: 800 -12183.652 0.494 0.368
Chain 1: 900 -11411.720 0.447 0.271
Chain 1: 1000 -10742.081 0.408 0.271
Chain 1: 1100 -18068.745 0.349 0.271
Chain 1: 1200 -12758.276 0.232 0.271
Chain 1: 1300 -11487.192 0.233 0.271
Chain 1: 1400 -10865.347 0.218 0.271
Chain 1: 1500 -11020.768 0.183 0.111
Chain 1: 1600 -10516.645 0.182 0.111
Chain 1: 1700 -10225.345 0.158 0.068
Chain 1: 1800 -11782.332 0.134 0.068
Chain 1: 1900 -18820.159 0.165 0.111
Chain 1: 2000 -10976.252 0.230 0.132
Chain 1: 2100 -11621.070 0.195 0.111
Chain 1: 2200 -10802.546 0.161 0.076
Chain 1: 2300 -9749.883 0.161 0.076
Chain 1: 2400 -9453.789 0.158 0.076
Chain 1: 2500 -9686.140 0.159 0.076
Chain 1: 2600 -9413.251 0.157 0.076
Chain 1: 2700 -12861.918 0.181 0.108
Chain 1: 2800 -19476.489 0.202 0.108
Chain 1: 2900 -17663.365 0.175 0.103
Chain 1: 3000 -14786.036 0.123 0.103
Chain 1: 3100 -12157.867 0.139 0.108
Chain 1: 3200 -9209.951 0.163 0.195
Chain 1: 3300 -10291.395 0.163 0.195
Chain 1: 3400 -14415.979 0.189 0.216
Chain 1: 3500 -9547.413 0.237 0.268
Chain 1: 3600 -9014.218 0.240 0.268
Chain 1: 3700 -11522.750 0.235 0.218
Chain 1: 3800 -14564.889 0.222 0.216
Chain 1: 3900 -9843.440 0.260 0.218
Chain 1: 4000 -8817.840 0.252 0.218
Chain 1: 4100 -9798.428 0.240 0.218
Chain 1: 4200 -15722.790 0.246 0.218
Chain 1: 4300 -12932.291 0.257 0.218
Chain 1: 4400 -10188.855 0.255 0.218
Chain 1: 4500 -11487.248 0.216 0.216
Chain 1: 4600 -16537.527 0.240 0.218
Chain 1: 4700 -9169.102 0.299 0.269
Chain 1: 4800 -8623.271 0.284 0.269
Chain 1: 4900 -9089.886 0.241 0.216
Chain 1: 5000 -14556.667 0.267 0.269
Chain 1: 5100 -8771.941 0.323 0.305
Chain 1: 5200 -12099.176 0.313 0.275
Chain 1: 5300 -9819.289 0.315 0.275
Chain 1: 5400 -13686.382 0.316 0.283
Chain 1: 5500 -11332.314 0.326 0.283
Chain 1: 5600 -13113.881 0.309 0.275
Chain 1: 5700 -10930.503 0.248 0.232
Chain 1: 5800 -9013.866 0.263 0.232
Chain 1: 5900 -10362.709 0.271 0.232
Chain 1: 6000 -9239.052 0.246 0.213
Chain 1: 6100 -12427.083 0.205 0.213
Chain 1: 6200 -8604.875 0.222 0.213
Chain 1: 6300 -13900.356 0.237 0.213
Chain 1: 6400 -8915.976 0.265 0.213
Chain 1: 6500 -10299.107 0.258 0.213
Chain 1: 6600 -9586.530 0.251 0.213
Chain 1: 6700 -14396.297 0.265 0.257
Chain 1: 6800 -8741.202 0.308 0.334
Chain 1: 6900 -12826.338 0.327 0.334
Chain 1: 7000 -9059.693 0.356 0.381
Chain 1: 7100 -8428.289 0.338 0.381
Chain 1: 7200 -10239.430 0.312 0.334
Chain 1: 7300 -8295.452 0.297 0.318
Chain 1: 7400 -8601.444 0.245 0.234
Chain 1: 7500 -8937.592 0.235 0.234
Chain 1: 7600 -10796.488 0.245 0.234
Chain 1: 7700 -9998.969 0.219 0.177
Chain 1: 7800 -9327.703 0.162 0.172
Chain 1: 7900 -8309.574 0.142 0.123
Chain 1: 8000 -12410.171 0.134 0.123
Chain 1: 8100 -13236.964 0.132 0.123
Chain 1: 8200 -9021.148 0.161 0.123
Chain 1: 8300 -8330.445 0.146 0.083
Chain 1: 8400 -8859.630 0.149 0.083
Chain 1: 8500 -8373.129 0.151 0.083
Chain 1: 8600 -10628.528 0.155 0.083
Chain 1: 8700 -8225.208 0.176 0.123
Chain 1: 8800 -8620.327 0.173 0.123
Chain 1: 8900 -8997.398 0.165 0.083
Chain 1: 9000 -9672.607 0.139 0.070
Chain 1: 9100 -8636.464 0.145 0.083
Chain 1: 9200 -9010.631 0.102 0.070
Chain 1: 9300 -10896.962 0.111 0.070
Chain 1: 9400 -12053.347 0.115 0.096
Chain 1: 9500 -13070.144 0.117 0.096
Chain 1: 9600 -9267.770 0.137 0.096
Chain 1: 9700 -8211.321 0.120 0.096
Chain 1: 9800 -8340.084 0.117 0.096
Chain 1: 9900 -8775.451 0.118 0.096
Chain 1: 10000 -8375.628 0.116 0.096
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003269 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.69 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58654.290 1.000 1.000
Chain 1: 200 -18061.844 1.624 2.247
Chain 1: 300 -8837.398 1.430 1.044
Chain 1: 400 -8090.918 1.096 1.044
Chain 1: 500 -8710.594 0.891 1.000
Chain 1: 600 -8664.537 0.743 1.000
Chain 1: 700 -8532.531 0.639 0.092
Chain 1: 800 -8246.950 0.564 0.092
Chain 1: 900 -7952.516 0.505 0.071
Chain 1: 1000 -7707.633 0.458 0.071
Chain 1: 1100 -7843.451 0.360 0.037
Chain 1: 1200 -7818.713 0.135 0.035
Chain 1: 1300 -7582.917 0.034 0.032
Chain 1: 1400 -7609.564 0.025 0.031
Chain 1: 1500 -7548.400 0.019 0.017
Chain 1: 1600 -7735.864 0.021 0.024
Chain 1: 1700 -7652.864 0.020 0.024
Chain 1: 1800 -7548.832 0.018 0.017
Chain 1: 1900 -7581.960 0.015 0.014
Chain 1: 2000 -7667.370 0.013 0.011
Chain 1: 2100 -7624.511 0.012 0.011
Chain 1: 2200 -7732.376 0.013 0.011
Chain 1: 2300 -7534.305 0.012 0.011
Chain 1: 2400 -7530.094 0.012 0.011
Chain 1: 2500 -7615.766 0.012 0.011
Chain 1: 2600 -7519.824 0.011 0.011
Chain 1: 2700 -7506.117 0.010 0.011
Chain 1: 2800 -7500.056 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003513 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.13 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85467.591 1.000 1.000
Chain 1: 200 -13824.692 3.091 5.182
Chain 1: 300 -10060.753 2.185 1.000
Chain 1: 400 -11419.854 1.669 1.000
Chain 1: 500 -9054.485 1.387 0.374
Chain 1: 600 -9570.474 1.165 0.374
Chain 1: 700 -8380.014 1.019 0.261
Chain 1: 800 -8784.803 0.897 0.261
Chain 1: 900 -8964.074 0.800 0.142
Chain 1: 1000 -8445.419 0.726 0.142
Chain 1: 1100 -8779.173 0.630 0.119 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8368.102 0.116 0.061
Chain 1: 1300 -8676.592 0.083 0.054
Chain 1: 1400 -8472.017 0.073 0.049
Chain 1: 1500 -8528.863 0.048 0.046
Chain 1: 1600 -8630.413 0.043 0.038
Chain 1: 1700 -8681.917 0.030 0.036
Chain 1: 1800 -8226.089 0.031 0.036
Chain 1: 1900 -8336.912 0.030 0.036
Chain 1: 2000 -8344.079 0.024 0.024
Chain 1: 2100 -8285.460 0.021 0.013
Chain 1: 2200 -8262.344 0.016 0.012
Chain 1: 2300 -8442.551 0.015 0.012
Chain 1: 2400 -8232.017 0.015 0.012
Chain 1: 2500 -8305.782 0.015 0.012
Chain 1: 2600 -8221.399 0.015 0.010
Chain 1: 2700 -8254.010 0.015 0.010
Chain 1: 2800 -8205.249 0.010 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004692 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 46.92 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8381498.618 1.000 1.000
Chain 1: 200 -1581946.137 2.649 4.298
Chain 1: 300 -890591.167 2.025 1.000
Chain 1: 400 -457813.286 1.755 1.000
Chain 1: 500 -358383.053 1.459 0.945
Chain 1: 600 -233606.154 1.305 0.945
Chain 1: 700 -119752.997 1.255 0.945
Chain 1: 800 -86924.720 1.145 0.945
Chain 1: 900 -67249.292 1.050 0.776
Chain 1: 1000 -52035.332 0.974 0.776
Chain 1: 1100 -39489.279 0.906 0.534 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38673.999 0.479 0.378
Chain 1: 1300 -26592.140 0.446 0.378
Chain 1: 1400 -26312.745 0.353 0.318
Chain 1: 1500 -22888.221 0.340 0.318
Chain 1: 1600 -22102.752 0.290 0.293
Chain 1: 1700 -20970.695 0.201 0.292
Chain 1: 1800 -20914.348 0.163 0.150
Chain 1: 1900 -21241.206 0.135 0.054
Chain 1: 2000 -19748.025 0.114 0.054
Chain 1: 2100 -19986.817 0.083 0.036
Chain 1: 2200 -20214.075 0.082 0.036
Chain 1: 2300 -19830.385 0.039 0.019
Chain 1: 2400 -19602.117 0.039 0.019
Chain 1: 2500 -19404.279 0.025 0.015
Chain 1: 2600 -19033.584 0.023 0.015
Chain 1: 2700 -18990.380 0.018 0.012
Chain 1: 2800 -18706.815 0.019 0.015
Chain 1: 2900 -18988.572 0.019 0.015
Chain 1: 3000 -18974.733 0.012 0.012
Chain 1: 3100 -19059.767 0.011 0.012
Chain 1: 3200 -18749.974 0.011 0.015
Chain 1: 3300 -18955.135 0.011 0.012
Chain 1: 3400 -18429.128 0.012 0.015
Chain 1: 3500 -19042.402 0.015 0.015
Chain 1: 3600 -18347.361 0.016 0.015
Chain 1: 3700 -18735.384 0.018 0.017
Chain 1: 3800 -17692.411 0.023 0.021
Chain 1: 3900 -17688.518 0.021 0.021
Chain 1: 4000 -17805.818 0.022 0.021
Chain 1: 4100 -17719.343 0.022 0.021
Chain 1: 4200 -17535.104 0.021 0.021
Chain 1: 4300 -17673.857 0.021 0.021
Chain 1: 4400 -17630.192 0.018 0.011
Chain 1: 4500 -17532.671 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001619 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 16.19 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48778.821 1.000 1.000
Chain 1: 200 -19213.058 1.269 1.539
Chain 1: 300 -15759.684 0.919 1.000
Chain 1: 400 -18273.892 0.724 1.000
Chain 1: 500 -12126.308 0.681 0.507
Chain 1: 600 -27204.702 0.659 0.554
Chain 1: 700 -22946.719 0.592 0.507
Chain 1: 800 -14007.261 0.598 0.554
Chain 1: 900 -10672.077 0.566 0.507
Chain 1: 1000 -11099.862 0.513 0.507
Chain 1: 1100 -11003.319 0.414 0.313
Chain 1: 1200 -10292.577 0.267 0.219
Chain 1: 1300 -14211.203 0.273 0.276
Chain 1: 1400 -10497.353 0.294 0.313
Chain 1: 1500 -9801.959 0.251 0.276
Chain 1: 1600 -14787.116 0.229 0.276
Chain 1: 1700 -9902.769 0.260 0.313
Chain 1: 1800 -14856.342 0.229 0.313
Chain 1: 1900 -10238.654 0.243 0.333
Chain 1: 2000 -10946.220 0.246 0.333
Chain 1: 2100 -13437.706 0.263 0.333
Chain 1: 2200 -9498.563 0.298 0.337
Chain 1: 2300 -11533.621 0.288 0.337
Chain 1: 2400 -9362.852 0.276 0.333
Chain 1: 2500 -9314.865 0.269 0.333
Chain 1: 2600 -15544.477 0.276 0.333
Chain 1: 2700 -9246.383 0.294 0.333
Chain 1: 2800 -9468.614 0.263 0.232
Chain 1: 2900 -9188.633 0.221 0.185
Chain 1: 3000 -9141.029 0.215 0.185
Chain 1: 3100 -13239.579 0.228 0.232
Chain 1: 3200 -12650.499 0.191 0.176
Chain 1: 3300 -12472.921 0.175 0.047
Chain 1: 3400 -9044.207 0.190 0.047
Chain 1: 3500 -9042.237 0.189 0.047
Chain 1: 3600 -14843.841 0.188 0.047
Chain 1: 3700 -18840.108 0.141 0.047
Chain 1: 3800 -12089.460 0.195 0.212
Chain 1: 3900 -10685.623 0.205 0.212
Chain 1: 4000 -9976.223 0.211 0.212
Chain 1: 4100 -9469.779 0.186 0.131
Chain 1: 4200 -8725.532 0.190 0.131
Chain 1: 4300 -9157.534 0.193 0.131
Chain 1: 4400 -10890.027 0.171 0.131
Chain 1: 4500 -9032.831 0.191 0.159
Chain 1: 4600 -14507.443 0.190 0.159
Chain 1: 4700 -8704.016 0.236 0.159
Chain 1: 4800 -8543.716 0.182 0.131
Chain 1: 4900 -8896.959 0.172 0.085
Chain 1: 5000 -11017.169 0.185 0.159
Chain 1: 5100 -8569.842 0.208 0.192
Chain 1: 5200 -8707.474 0.201 0.192
Chain 1: 5300 -12966.202 0.229 0.206
Chain 1: 5400 -10503.415 0.236 0.234
Chain 1: 5500 -9065.393 0.232 0.234
Chain 1: 5600 -9239.128 0.196 0.192
Chain 1: 5700 -9128.129 0.130 0.159
Chain 1: 5800 -8590.371 0.135 0.159
Chain 1: 5900 -11110.996 0.154 0.192
Chain 1: 6000 -11837.936 0.140 0.159
Chain 1: 6100 -11769.316 0.113 0.063
Chain 1: 6200 -8426.582 0.151 0.159
Chain 1: 6300 -9049.151 0.125 0.069
Chain 1: 6400 -13321.238 0.133 0.069
Chain 1: 6500 -12287.277 0.126 0.069
Chain 1: 6600 -8606.074 0.167 0.084
Chain 1: 6700 -9102.579 0.171 0.084
Chain 1: 6800 -10166.462 0.175 0.105
Chain 1: 6900 -10444.479 0.155 0.084
Chain 1: 7000 -8544.786 0.171 0.105
Chain 1: 7100 -8552.698 0.171 0.105
Chain 1: 7200 -8840.987 0.134 0.084
Chain 1: 7300 -8724.227 0.129 0.084
Chain 1: 7400 -10223.370 0.111 0.084
Chain 1: 7500 -8961.780 0.117 0.105
Chain 1: 7600 -13155.951 0.106 0.105
Chain 1: 7700 -12014.900 0.110 0.105
Chain 1: 7800 -11848.205 0.101 0.095
Chain 1: 7900 -9670.997 0.121 0.141
Chain 1: 8000 -8563.058 0.112 0.129
Chain 1: 8100 -8305.542 0.115 0.129
Chain 1: 8200 -9824.604 0.127 0.141
Chain 1: 8300 -8347.568 0.143 0.147
Chain 1: 8400 -10309.352 0.148 0.155
Chain 1: 8500 -8177.544 0.160 0.177
Chain 1: 8600 -10810.634 0.152 0.177
Chain 1: 8700 -8631.471 0.168 0.190
Chain 1: 8800 -8279.126 0.171 0.190
Chain 1: 8900 -9317.369 0.159 0.177
Chain 1: 9000 -8422.249 0.157 0.177
Chain 1: 9100 -8219.523 0.156 0.177
Chain 1: 9200 -8480.121 0.144 0.177
Chain 1: 9300 -9078.020 0.133 0.111
Chain 1: 9400 -8583.943 0.120 0.106
Chain 1: 9500 -9536.367 0.103 0.100
Chain 1: 9600 -8994.195 0.085 0.066
Chain 1: 9700 -9588.590 0.066 0.062
Chain 1: 9800 -8357.519 0.077 0.066
Chain 1: 9900 -10502.007 0.086 0.066
Chain 1: 10000 -8217.332 0.103 0.066
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001566 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.66 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56779.860 1.000 1.000
Chain 1: 200 -17398.885 1.632 2.263
Chain 1: 300 -8744.748 1.418 1.000
Chain 1: 400 -8384.475 1.074 1.000
Chain 1: 500 -8640.689 0.865 0.990
Chain 1: 600 -8392.766 0.726 0.990
Chain 1: 700 -8183.139 0.626 0.043
Chain 1: 800 -8159.252 0.548 0.043
Chain 1: 900 -7953.967 0.490 0.030
Chain 1: 1000 -7636.267 0.445 0.042
Chain 1: 1100 -7807.690 0.347 0.030
Chain 1: 1200 -7665.574 0.123 0.030
Chain 1: 1300 -7639.639 0.024 0.026
Chain 1: 1400 -7662.000 0.020 0.026
Chain 1: 1500 -7641.525 0.017 0.022
Chain 1: 1600 -7740.686 0.016 0.019
Chain 1: 1700 -7465.687 0.017 0.019
Chain 1: 1800 -7626.839 0.019 0.021
Chain 1: 1900 -7603.148 0.016 0.019
Chain 1: 2000 -7640.632 0.013 0.013
Chain 1: 2100 -7577.839 0.011 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.005349 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 53.49 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86679.522 1.000 1.000
Chain 1: 200 -13537.238 3.202 5.403
Chain 1: 300 -9948.102 2.255 1.000
Chain 1: 400 -10710.049 1.709 1.000
Chain 1: 500 -8898.970 1.408 0.361
Chain 1: 600 -8463.953 1.182 0.361
Chain 1: 700 -8498.102 1.013 0.204
Chain 1: 800 -8709.740 0.890 0.204
Chain 1: 900 -8794.248 0.792 0.071
Chain 1: 1000 -8595.940 0.715 0.071
Chain 1: 1100 -8812.092 0.618 0.051 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8462.780 0.081 0.041
Chain 1: 1300 -8658.159 0.048 0.025
Chain 1: 1400 -8659.493 0.040 0.024
Chain 1: 1500 -8556.154 0.021 0.023
Chain 1: 1600 -8657.852 0.017 0.023
Chain 1: 1700 -8746.158 0.018 0.023
Chain 1: 1800 -8346.749 0.020 0.023
Chain 1: 1900 -8447.414 0.021 0.023
Chain 1: 2000 -8418.347 0.019 0.012
Chain 1: 2100 -8538.905 0.018 0.012
Chain 1: 2200 -8315.463 0.016 0.012
Chain 1: 2300 -8476.547 0.016 0.012
Chain 1: 2400 -8482.729 0.016 0.012
Chain 1: 2500 -8461.230 0.015 0.012
Chain 1: 2600 -8462.984 0.014 0.012
Chain 1: 2700 -8368.985 0.014 0.012
Chain 1: 2800 -8339.059 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004394 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 43.94 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8390741.885 1.000 1.000
Chain 1: 200 -1586069.231 2.645 4.290
Chain 1: 300 -891121.032 2.023 1.000
Chain 1: 400 -457189.607 1.755 1.000
Chain 1: 500 -357199.521 1.460 0.949
Chain 1: 600 -232286.050 1.306 0.949
Chain 1: 700 -118924.822 1.256 0.949
Chain 1: 800 -86174.607 1.146 0.949
Chain 1: 900 -66602.627 1.052 0.780
Chain 1: 1000 -51455.696 0.976 0.780
Chain 1: 1100 -38979.854 0.908 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38163.337 0.481 0.380
Chain 1: 1300 -26176.993 0.449 0.380
Chain 1: 1400 -25900.420 0.355 0.320
Chain 1: 1500 -22500.739 0.342 0.320
Chain 1: 1600 -21720.513 0.292 0.294
Chain 1: 1700 -20601.717 0.202 0.294
Chain 1: 1800 -20547.450 0.164 0.151
Chain 1: 1900 -20873.287 0.136 0.054
Chain 1: 2000 -19388.419 0.115 0.054
Chain 1: 2100 -19626.905 0.084 0.036
Chain 1: 2200 -19852.276 0.083 0.036
Chain 1: 2300 -19470.458 0.039 0.020
Chain 1: 2400 -19242.709 0.039 0.020
Chain 1: 2500 -19044.242 0.025 0.016
Chain 1: 2600 -18675.260 0.023 0.016
Chain 1: 2700 -18632.509 0.018 0.012
Chain 1: 2800 -18349.239 0.020 0.015
Chain 1: 2900 -18630.268 0.019 0.015
Chain 1: 3000 -18616.694 0.012 0.012
Chain 1: 3100 -18701.558 0.011 0.012
Chain 1: 3200 -18392.578 0.012 0.015
Chain 1: 3300 -18597.045 0.011 0.012
Chain 1: 3400 -18072.339 0.013 0.015
Chain 1: 3500 -18683.524 0.015 0.015
Chain 1: 3600 -17991.141 0.017 0.015
Chain 1: 3700 -18377.138 0.018 0.017
Chain 1: 3800 -17338.190 0.023 0.021
Chain 1: 3900 -17334.306 0.021 0.021
Chain 1: 4000 -17451.679 0.022 0.021
Chain 1: 4100 -17365.416 0.022 0.021
Chain 1: 4200 -17182.024 0.021 0.021
Chain 1: 4300 -17320.257 0.021 0.021
Chain 1: 4400 -17277.341 0.019 0.011
Chain 1: 4500 -17179.841 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001394 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.94 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48824.901 1.000 1.000
Chain 1: 200 -23049.514 1.059 1.118
Chain 1: 300 -18956.438 0.778 1.000
Chain 1: 400 -12010.122 0.728 1.000
Chain 1: 500 -14713.546 0.619 0.578
Chain 1: 600 -18966.242 0.553 0.578
Chain 1: 700 -12491.648 0.548 0.518
Chain 1: 800 -11500.327 0.491 0.518
Chain 1: 900 -15597.393 0.465 0.263
Chain 1: 1000 -10436.924 0.468 0.494
Chain 1: 1100 -11145.506 0.375 0.263
Chain 1: 1200 -10669.990 0.267 0.224
Chain 1: 1300 -13107.958 0.264 0.224
Chain 1: 1400 -15938.495 0.224 0.186
Chain 1: 1500 -13898.121 0.220 0.186
Chain 1: 1600 -12343.456 0.211 0.178
Chain 1: 1700 -10908.170 0.172 0.147
Chain 1: 1800 -9890.300 0.174 0.147
Chain 1: 1900 -10151.395 0.150 0.132
Chain 1: 2000 -12900.073 0.122 0.132
Chain 1: 2100 -10889.356 0.134 0.147
Chain 1: 2200 -9761.601 0.141 0.147
Chain 1: 2300 -9474.672 0.125 0.132
Chain 1: 2400 -10286.254 0.116 0.126
Chain 1: 2500 -12486.518 0.118 0.126
Chain 1: 2600 -14049.940 0.117 0.116
Chain 1: 2700 -13894.291 0.105 0.111
Chain 1: 2800 -11071.723 0.120 0.116
Chain 1: 2900 -9372.186 0.136 0.176
Chain 1: 3000 -18508.642 0.164 0.176
Chain 1: 3100 -9318.379 0.244 0.176
Chain 1: 3200 -9358.266 0.233 0.176
Chain 1: 3300 -16256.303 0.272 0.181
Chain 1: 3400 -9227.255 0.341 0.255
Chain 1: 3500 -9075.401 0.325 0.255
Chain 1: 3600 -11130.816 0.332 0.255
Chain 1: 3700 -9694.955 0.346 0.255
Chain 1: 3800 -9825.274 0.321 0.185
Chain 1: 3900 -12345.768 0.324 0.204
Chain 1: 4000 -15301.127 0.294 0.193
Chain 1: 4100 -9362.923 0.258 0.193
Chain 1: 4200 -12168.243 0.281 0.204
Chain 1: 4300 -9688.547 0.264 0.204
Chain 1: 4400 -9444.248 0.191 0.193
Chain 1: 4500 -9546.312 0.190 0.193
Chain 1: 4600 -13396.907 0.200 0.204
Chain 1: 4700 -9003.485 0.234 0.231
Chain 1: 4800 -10772.731 0.249 0.231
Chain 1: 4900 -9151.652 0.247 0.231
Chain 1: 5000 -8728.646 0.232 0.231
Chain 1: 5100 -17294.345 0.218 0.231
Chain 1: 5200 -14106.171 0.218 0.226
Chain 1: 5300 -14416.182 0.194 0.177
Chain 1: 5400 -12122.495 0.211 0.189
Chain 1: 5500 -8677.623 0.249 0.226
Chain 1: 5600 -8599.775 0.222 0.189
Chain 1: 5700 -11958.380 0.201 0.189
Chain 1: 5800 -8587.350 0.224 0.226
Chain 1: 5900 -8575.651 0.206 0.226
Chain 1: 6000 -8692.639 0.203 0.226
Chain 1: 6100 -10587.918 0.171 0.189
Chain 1: 6200 -8958.132 0.167 0.182
Chain 1: 6300 -9452.431 0.170 0.182
Chain 1: 6400 -11859.292 0.171 0.182
Chain 1: 6500 -9263.771 0.159 0.182
Chain 1: 6600 -12125.819 0.182 0.203
Chain 1: 6700 -8515.274 0.196 0.203
Chain 1: 6800 -8612.266 0.158 0.182
Chain 1: 6900 -12390.773 0.189 0.203
Chain 1: 7000 -8382.773 0.235 0.236
Chain 1: 7100 -9350.083 0.228 0.236
Chain 1: 7200 -8442.540 0.220 0.236
Chain 1: 7300 -11421.017 0.241 0.261
Chain 1: 7400 -8586.950 0.254 0.280
Chain 1: 7500 -9889.319 0.239 0.261
Chain 1: 7600 -9375.657 0.221 0.261
Chain 1: 7700 -8581.886 0.188 0.132
Chain 1: 7800 -8681.276 0.188 0.132
Chain 1: 7900 -8312.400 0.161 0.107
Chain 1: 8000 -9874.850 0.129 0.107
Chain 1: 8100 -8557.013 0.135 0.132
Chain 1: 8200 -9044.973 0.129 0.132
Chain 1: 8300 -8242.747 0.113 0.097
Chain 1: 8400 -9206.494 0.090 0.097
Chain 1: 8500 -8441.656 0.086 0.092
Chain 1: 8600 -8281.947 0.083 0.092
Chain 1: 8700 -8708.223 0.078 0.091
Chain 1: 8800 -9288.976 0.083 0.091
Chain 1: 8900 -11452.777 0.098 0.097
Chain 1: 9000 -10525.027 0.091 0.091
Chain 1: 9100 -10860.875 0.079 0.088
Chain 1: 9200 -8788.124 0.097 0.091
Chain 1: 9300 -8349.265 0.092 0.088
Chain 1: 9400 -8245.534 0.083 0.063
Chain 1: 9500 -8240.925 0.074 0.053
Chain 1: 9600 -8297.276 0.073 0.053
Chain 1: 9700 -8622.896 0.072 0.053
Chain 1: 9800 -10448.383 0.083 0.053
Chain 1: 9900 -10029.813 0.068 0.042
Chain 1: 10000 -8826.292 0.073 0.042
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001593 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.93 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56870.675 1.000 1.000
Chain 1: 200 -17338.894 1.640 2.280
Chain 1: 300 -8730.204 1.422 1.000
Chain 1: 400 -8424.440 1.076 1.000
Chain 1: 500 -8192.699 0.866 0.986
Chain 1: 600 -8414.627 0.726 0.986
Chain 1: 700 -7950.368 0.631 0.058
Chain 1: 800 -8066.496 0.554 0.058
Chain 1: 900 -8043.006 0.493 0.036
Chain 1: 1000 -7858.683 0.446 0.036
Chain 1: 1100 -7678.960 0.348 0.028
Chain 1: 1200 -7660.637 0.120 0.026
Chain 1: 1300 -7721.741 0.022 0.023
Chain 1: 1400 -7859.550 0.021 0.023
Chain 1: 1500 -7666.682 0.020 0.023
Chain 1: 1600 -7779.086 0.019 0.018
Chain 1: 1700 -7551.289 0.016 0.018
Chain 1: 1800 -7594.481 0.015 0.018
Chain 1: 1900 -7600.805 0.015 0.018
Chain 1: 2000 -7671.358 0.014 0.014
Chain 1: 2100 -7633.153 0.012 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003223 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.23 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85826.547 1.000 1.000
Chain 1: 200 -13444.107 3.192 5.384
Chain 1: 300 -9896.684 2.247 1.000
Chain 1: 400 -10631.494 1.703 1.000
Chain 1: 500 -8833.635 1.403 0.358
Chain 1: 600 -8400.291 1.178 0.358
Chain 1: 700 -8802.688 1.016 0.204
Chain 1: 800 -9257.217 0.895 0.204
Chain 1: 900 -8702.169 0.803 0.069
Chain 1: 1000 -8476.917 0.725 0.069
Chain 1: 1100 -8624.053 0.627 0.064 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8493.383 0.090 0.052
Chain 1: 1300 -8613.883 0.056 0.049
Chain 1: 1400 -8623.424 0.049 0.046
Chain 1: 1500 -8521.194 0.030 0.027
Chain 1: 1600 -8619.227 0.026 0.017
Chain 1: 1700 -8711.458 0.022 0.015
Chain 1: 1800 -8323.029 0.022 0.015
Chain 1: 1900 -8425.548 0.017 0.014
Chain 1: 2000 -8395.488 0.014 0.012
Chain 1: 2100 -8527.446 0.014 0.012
Chain 1: 2200 -8313.198 0.015 0.012
Chain 1: 2300 -8454.843 0.016 0.012
Chain 1: 2400 -8467.168 0.016 0.012
Chain 1: 2500 -8435.265 0.015 0.012
Chain 1: 2600 -8434.893 0.014 0.012
Chain 1: 2700 -8343.188 0.014 0.012
Chain 1: 2800 -8319.372 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.005534 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 55.34 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8381422.476 1.000 1.000
Chain 1: 200 -1582640.562 2.648 4.296
Chain 1: 300 -890791.788 2.024 1.000
Chain 1: 400 -457450.724 1.755 1.000
Chain 1: 500 -357743.644 1.460 0.947
Chain 1: 600 -232914.592 1.306 0.947
Chain 1: 700 -119196.732 1.255 0.947
Chain 1: 800 -86375.281 1.146 0.947
Chain 1: 900 -66722.978 1.051 0.777
Chain 1: 1000 -51513.272 0.976 0.777
Chain 1: 1100 -38980.059 0.908 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38155.811 0.481 0.380
Chain 1: 1300 -26112.285 0.449 0.380
Chain 1: 1400 -25829.897 0.355 0.322
Chain 1: 1500 -22416.379 0.343 0.322
Chain 1: 1600 -21631.941 0.293 0.295
Chain 1: 1700 -20506.290 0.203 0.295
Chain 1: 1800 -20450.432 0.165 0.152
Chain 1: 1900 -20776.028 0.137 0.055
Chain 1: 2000 -19288.398 0.115 0.055
Chain 1: 2100 -19526.780 0.084 0.036
Chain 1: 2200 -19752.674 0.083 0.036
Chain 1: 2300 -19370.534 0.039 0.020
Chain 1: 2400 -19142.828 0.039 0.020
Chain 1: 2500 -18944.761 0.025 0.016
Chain 1: 2600 -18575.549 0.024 0.016
Chain 1: 2700 -18532.767 0.018 0.012
Chain 1: 2800 -18249.738 0.020 0.016
Chain 1: 2900 -18530.834 0.020 0.015
Chain 1: 3000 -18517.127 0.012 0.012
Chain 1: 3100 -18601.959 0.011 0.012
Chain 1: 3200 -18293.018 0.012 0.015
Chain 1: 3300 -18497.501 0.011 0.012
Chain 1: 3400 -17973.013 0.013 0.015
Chain 1: 3500 -18583.893 0.015 0.016
Chain 1: 3600 -17892.010 0.017 0.016
Chain 1: 3700 -18277.703 0.019 0.017
Chain 1: 3800 -17239.496 0.023 0.021
Chain 1: 3900 -17235.713 0.022 0.021
Chain 1: 4000 -17353.039 0.022 0.021
Chain 1: 4100 -17266.788 0.022 0.021
Chain 1: 4200 -17083.605 0.022 0.021
Chain 1: 4300 -17221.636 0.021 0.021
Chain 1: 4400 -17178.848 0.019 0.011
Chain 1: 4500 -17081.463 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001583 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.83 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13193.614 1.000 1.000
Chain 1: 200 -9784.922 0.674 1.000
Chain 1: 300 -7953.889 0.526 0.348
Chain 1: 400 -8148.259 0.401 0.348
Chain 1: 500 -8084.773 0.322 0.230
Chain 1: 600 -7877.819 0.273 0.230
Chain 1: 700 -7864.561 0.234 0.026
Chain 1: 800 -7938.023 0.206 0.026
Chain 1: 900 -8009.352 0.184 0.024
Chain 1: 1000 -7939.563 0.167 0.024
Chain 1: 1100 -8054.671 0.068 0.014
Chain 1: 1200 -7906.389 0.035 0.014
Chain 1: 1300 -7815.463 0.013 0.012
Chain 1: 1400 -7838.246 0.011 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001475 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.75 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58099.481 1.000 1.000
Chain 1: 200 -17830.526 1.629 2.258
Chain 1: 300 -8729.382 1.434 1.043
Chain 1: 400 -8103.753 1.095 1.043
Chain 1: 500 -8632.623 0.888 1.000
Chain 1: 600 -8772.946 0.743 1.000
Chain 1: 700 -8215.271 0.646 0.077
Chain 1: 800 -8267.719 0.566 0.077
Chain 1: 900 -7840.123 0.509 0.068
Chain 1: 1000 -7825.525 0.459 0.068
Chain 1: 1100 -7788.614 0.359 0.061
Chain 1: 1200 -7609.163 0.136 0.055
Chain 1: 1300 -7578.052 0.032 0.024
Chain 1: 1400 -7892.830 0.028 0.024
Chain 1: 1500 -7544.645 0.027 0.024
Chain 1: 1600 -7732.333 0.027 0.024
Chain 1: 1700 -7730.154 0.021 0.024
Chain 1: 1800 -7646.301 0.021 0.024
Chain 1: 1900 -7605.156 0.016 0.011
Chain 1: 2000 -7630.498 0.016 0.011
Chain 1: 2100 -7545.921 0.017 0.011
Chain 1: 2200 -7696.787 0.017 0.011
Chain 1: 2300 -7563.252 0.018 0.018
Chain 1: 2400 -7634.263 0.015 0.011
Chain 1: 2500 -7541.266 0.011 0.011
Chain 1: 2600 -7498.872 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003193 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.93 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86263.700 1.000 1.000
Chain 1: 200 -13622.737 3.166 5.332
Chain 1: 300 -9924.977 2.235 1.000
Chain 1: 400 -11006.931 1.701 1.000
Chain 1: 500 -8927.454 1.407 0.373
Chain 1: 600 -8570.384 1.180 0.373
Chain 1: 700 -8525.651 1.012 0.233
Chain 1: 800 -9405.995 0.897 0.233
Chain 1: 900 -8763.817 0.806 0.098
Chain 1: 1000 -8588.651 0.727 0.098
Chain 1: 1100 -8732.385 0.629 0.094 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8177.891 0.102 0.073
Chain 1: 1300 -8609.073 0.070 0.068
Chain 1: 1400 -8556.352 0.061 0.050
Chain 1: 1500 -8456.939 0.039 0.042
Chain 1: 1600 -8562.677 0.036 0.020
Chain 1: 1700 -8627.812 0.036 0.020
Chain 1: 1800 -8194.438 0.032 0.020
Chain 1: 1900 -8298.850 0.026 0.016
Chain 1: 2000 -8274.250 0.024 0.013
Chain 1: 2100 -8251.884 0.023 0.012
Chain 1: 2200 -8216.921 0.016 0.012
Chain 1: 2300 -8346.118 0.013 0.012
Chain 1: 2400 -8201.117 0.014 0.012
Chain 1: 2500 -8269.878 0.014 0.012
Chain 1: 2600 -8188.959 0.013 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004574 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 45.74 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8398110.376 1.000 1.000
Chain 1: 200 -1584715.187 2.650 4.299
Chain 1: 300 -890689.975 2.026 1.000
Chain 1: 400 -457606.575 1.756 1.000
Chain 1: 500 -357883.270 1.461 0.946
Chain 1: 600 -233039.367 1.307 0.946
Chain 1: 700 -119351.345 1.256 0.946
Chain 1: 800 -86532.398 1.146 0.946
Chain 1: 900 -66896.894 1.052 0.779
Chain 1: 1000 -51708.008 0.976 0.779
Chain 1: 1100 -39189.780 0.908 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38370.583 0.480 0.379
Chain 1: 1300 -26332.352 0.448 0.379
Chain 1: 1400 -26053.236 0.354 0.319
Chain 1: 1500 -22640.146 0.341 0.319
Chain 1: 1600 -21856.685 0.291 0.294
Chain 1: 1700 -20731.134 0.202 0.294
Chain 1: 1800 -20675.531 0.164 0.151
Chain 1: 1900 -21002.009 0.136 0.054
Chain 1: 2000 -19512.464 0.114 0.054
Chain 1: 2100 -19751.187 0.084 0.036
Chain 1: 2200 -19977.600 0.083 0.036
Chain 1: 2300 -19594.696 0.039 0.020
Chain 1: 2400 -19366.659 0.039 0.020
Chain 1: 2500 -19168.438 0.025 0.016
Chain 1: 2600 -18798.608 0.023 0.016
Chain 1: 2700 -18755.527 0.018 0.012
Chain 1: 2800 -18472.120 0.019 0.015
Chain 1: 2900 -18753.496 0.019 0.015
Chain 1: 3000 -18739.823 0.012 0.012
Chain 1: 3100 -18824.807 0.011 0.012
Chain 1: 3200 -18515.349 0.012 0.015
Chain 1: 3300 -18720.156 0.011 0.012
Chain 1: 3400 -18194.709 0.012 0.015
Chain 1: 3500 -18807.096 0.015 0.015
Chain 1: 3600 -18113.112 0.017 0.015
Chain 1: 3700 -18500.385 0.018 0.017
Chain 1: 3800 -17458.991 0.023 0.021
Chain 1: 3900 -17455.051 0.021 0.021
Chain 1: 4000 -17572.419 0.022 0.021
Chain 1: 4100 -17486.087 0.022 0.021
Chain 1: 4200 -17302.084 0.021 0.021
Chain 1: 4300 -17440.712 0.021 0.021
Chain 1: 4400 -17397.364 0.018 0.011
Chain 1: 4500 -17299.795 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001364 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.64 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49259.891 1.000 1.000
Chain 1: 200 -15525.056 1.586 2.173
Chain 1: 300 -29312.744 1.214 1.000
Chain 1: 400 -14979.928 1.150 1.000
Chain 1: 500 -16487.700 0.938 0.957
Chain 1: 600 -18224.103 0.798 0.957
Chain 1: 700 -12646.273 0.747 0.470
Chain 1: 800 -15694.896 0.678 0.470
Chain 1: 900 -13871.620 0.617 0.441
Chain 1: 1000 -13479.099 0.558 0.441
Chain 1: 1100 -11655.846 0.474 0.194
Chain 1: 1200 -18675.941 0.294 0.194
Chain 1: 1300 -11732.140 0.306 0.194
Chain 1: 1400 -12355.656 0.216 0.156
Chain 1: 1500 -10392.406 0.225 0.189
Chain 1: 1600 -11751.668 0.228 0.189
Chain 1: 1700 -17397.396 0.216 0.189
Chain 1: 1800 -13574.772 0.225 0.189
Chain 1: 1900 -11265.653 0.232 0.205
Chain 1: 2000 -15911.936 0.258 0.282
Chain 1: 2100 -11025.951 0.287 0.292
Chain 1: 2200 -10138.345 0.258 0.282
Chain 1: 2300 -12387.876 0.217 0.205
Chain 1: 2400 -9667.268 0.240 0.281
Chain 1: 2500 -17876.720 0.267 0.282
Chain 1: 2600 -9701.167 0.340 0.292
Chain 1: 2700 -10111.621 0.311 0.282
Chain 1: 2800 -11395.146 0.295 0.281
Chain 1: 2900 -9898.398 0.289 0.281
Chain 1: 3000 -10391.253 0.265 0.182
Chain 1: 3100 -10274.603 0.222 0.151
Chain 1: 3200 -18783.129 0.258 0.182
Chain 1: 3300 -16539.243 0.254 0.151
Chain 1: 3400 -9904.050 0.292 0.151
Chain 1: 3500 -10552.174 0.253 0.136
Chain 1: 3600 -17459.575 0.208 0.136
Chain 1: 3700 -9816.926 0.282 0.151
Chain 1: 3800 -14075.854 0.301 0.303
Chain 1: 3900 -10380.933 0.321 0.356
Chain 1: 4000 -9979.403 0.320 0.356
Chain 1: 4100 -9803.373 0.321 0.356
Chain 1: 4200 -10708.433 0.284 0.303
Chain 1: 4300 -10142.795 0.276 0.303
Chain 1: 4400 -10085.673 0.210 0.085
Chain 1: 4500 -9261.671 0.213 0.089
Chain 1: 4600 -9168.507 0.174 0.085
Chain 1: 4700 -9220.379 0.097 0.056
Chain 1: 4800 -8927.352 0.070 0.040
Chain 1: 4900 -9074.989 0.036 0.033
Chain 1: 5000 -14142.985 0.068 0.033
Chain 1: 5100 -8874.207 0.125 0.056
Chain 1: 5200 -12375.428 0.145 0.056
Chain 1: 5300 -12082.784 0.142 0.033
Chain 1: 5400 -8931.967 0.177 0.089
Chain 1: 5500 -9040.922 0.169 0.033
Chain 1: 5600 -8987.072 0.168 0.033
Chain 1: 5700 -9110.103 0.169 0.033
Chain 1: 5800 -9246.899 0.167 0.024
Chain 1: 5900 -15874.793 0.208 0.283
Chain 1: 6000 -12662.357 0.197 0.254
Chain 1: 6100 -9067.201 0.177 0.254
Chain 1: 6200 -9935.308 0.158 0.087
Chain 1: 6300 -8891.178 0.167 0.117
Chain 1: 6400 -9190.684 0.135 0.087
Chain 1: 6500 -8875.220 0.137 0.087
Chain 1: 6600 -8950.874 0.138 0.087
Chain 1: 6700 -9994.878 0.147 0.104
Chain 1: 6800 -8854.767 0.158 0.117
Chain 1: 6900 -8993.508 0.118 0.104
Chain 1: 7000 -9763.371 0.101 0.087
Chain 1: 7100 -8709.595 0.073 0.087
Chain 1: 7200 -9381.129 0.071 0.079
Chain 1: 7300 -10762.651 0.073 0.079
Chain 1: 7400 -9478.768 0.083 0.104
Chain 1: 7500 -11193.782 0.095 0.121
Chain 1: 7600 -9845.188 0.107 0.128
Chain 1: 7700 -9036.503 0.106 0.128
Chain 1: 7800 -9090.079 0.094 0.121
Chain 1: 7900 -8914.562 0.094 0.121
Chain 1: 8000 -12338.027 0.114 0.128
Chain 1: 8100 -9656.510 0.130 0.135
Chain 1: 8200 -11466.564 0.138 0.137
Chain 1: 8300 -8624.436 0.158 0.153
Chain 1: 8400 -8648.644 0.145 0.153
Chain 1: 8500 -8659.231 0.130 0.137
Chain 1: 8600 -8589.219 0.117 0.089
Chain 1: 8700 -9064.842 0.113 0.052
Chain 1: 8800 -8809.684 0.116 0.052
Chain 1: 8900 -9324.765 0.119 0.055
Chain 1: 9000 -9929.257 0.097 0.055
Chain 1: 9100 -8916.797 0.081 0.055
Chain 1: 9200 -8762.317 0.067 0.052
Chain 1: 9300 -13076.874 0.067 0.052
Chain 1: 9400 -8679.625 0.117 0.055
Chain 1: 9500 -11170.419 0.140 0.061
Chain 1: 9600 -10776.992 0.142 0.061
Chain 1: 9700 -8839.136 0.159 0.114
Chain 1: 9800 -8823.746 0.156 0.114
Chain 1: 9900 -10245.082 0.165 0.139
Chain 1: 10000 -8888.402 0.174 0.153
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001895 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 18.95 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57679.285 1.000 1.000
Chain 1: 200 -18052.499 1.598 2.195
Chain 1: 300 -9042.248 1.397 1.000
Chain 1: 400 -8462.690 1.065 1.000
Chain 1: 500 -8637.818 0.856 0.996
Chain 1: 600 -8670.708 0.714 0.996
Chain 1: 700 -8189.784 0.620 0.068
Chain 1: 800 -8455.108 0.547 0.068
Chain 1: 900 -8076.702 0.491 0.059
Chain 1: 1000 -7900.531 0.444 0.059
Chain 1: 1100 -7986.475 0.345 0.047
Chain 1: 1200 -7789.167 0.128 0.031
Chain 1: 1300 -8022.458 0.032 0.029
Chain 1: 1400 -7890.615 0.027 0.025
Chain 1: 1500 -7675.198 0.027 0.028
Chain 1: 1600 -7847.852 0.029 0.028
Chain 1: 1700 -7712.604 0.025 0.025
Chain 1: 1800 -7829.630 0.023 0.022
Chain 1: 1900 -7691.992 0.020 0.022
Chain 1: 2000 -7801.288 0.020 0.018
Chain 1: 2100 -7683.289 0.020 0.018
Chain 1: 2200 -7875.523 0.020 0.018
Chain 1: 2300 -7652.029 0.020 0.018
Chain 1: 2400 -7730.784 0.019 0.018
Chain 1: 2500 -7701.707 0.017 0.018
Chain 1: 2600 -7621.440 0.016 0.015
Chain 1: 2700 -7609.482 0.014 0.015
Chain 1: 2800 -7613.945 0.013 0.014
Chain 1: 2900 -7473.951 0.013 0.014
Chain 1: 3000 -7625.116 0.013 0.015
Chain 1: 3100 -7622.700 0.012 0.011
Chain 1: 3200 -7839.413 0.012 0.011
Chain 1: 3300 -7562.023 0.013 0.011
Chain 1: 3400 -7793.905 0.015 0.019
Chain 1: 3500 -7533.859 0.018 0.020
Chain 1: 3600 -7601.412 0.018 0.020
Chain 1: 3700 -7550.820 0.018 0.020
Chain 1: 3800 -7551.836 0.018 0.020
Chain 1: 3900 -7510.636 0.017 0.020
Chain 1: 4000 -7502.653 0.015 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003286 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.86 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87195.448 1.000 1.000
Chain 1: 200 -14028.318 3.108 5.216
Chain 1: 300 -10320.308 2.192 1.000
Chain 1: 400 -11491.074 1.669 1.000
Chain 1: 500 -9327.530 1.382 0.359
Chain 1: 600 -8837.521 1.161 0.359
Chain 1: 700 -9253.716 1.001 0.232
Chain 1: 800 -9450.338 0.879 0.232
Chain 1: 900 -9097.597 0.785 0.102
Chain 1: 1000 -9074.135 0.707 0.102
Chain 1: 1100 -8922.046 0.609 0.055 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8714.396 0.090 0.045
Chain 1: 1300 -9005.749 0.057 0.039
Chain 1: 1400 -8959.802 0.047 0.032
Chain 1: 1500 -8853.611 0.025 0.024
Chain 1: 1600 -8959.697 0.021 0.021
Chain 1: 1700 -9035.113 0.017 0.017
Chain 1: 1800 -8603.068 0.020 0.017
Chain 1: 1900 -8707.155 0.018 0.012
Chain 1: 2000 -8682.584 0.018 0.012
Chain 1: 2100 -8653.278 0.016 0.012
Chain 1: 2200 -8625.246 0.014 0.012
Chain 1: 2300 -8753.629 0.012 0.012
Chain 1: 2400 -8610.599 0.014 0.012
Chain 1: 2500 -8678.116 0.013 0.012
Chain 1: 2600 -8598.484 0.013 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004794 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 47.94 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8391493.286 1.000 1.000
Chain 1: 200 -1582783.377 2.651 4.302
Chain 1: 300 -891464.389 2.026 1.000
Chain 1: 400 -458702.093 1.755 1.000
Chain 1: 500 -358954.863 1.460 0.943
Chain 1: 600 -233836.098 1.306 0.943
Chain 1: 700 -119915.482 1.255 0.943
Chain 1: 800 -87068.641 1.145 0.943
Chain 1: 900 -67393.213 1.050 0.775
Chain 1: 1000 -52171.957 0.974 0.775
Chain 1: 1100 -39630.836 0.906 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38806.893 0.478 0.377
Chain 1: 1300 -26744.183 0.446 0.377
Chain 1: 1400 -26462.489 0.352 0.316
Chain 1: 1500 -23044.009 0.339 0.316
Chain 1: 1600 -22259.044 0.289 0.292
Chain 1: 1700 -21130.389 0.200 0.292
Chain 1: 1800 -21074.075 0.162 0.148
Chain 1: 1900 -21400.485 0.135 0.053
Chain 1: 2000 -19909.613 0.113 0.053
Chain 1: 2100 -20148.268 0.082 0.035
Chain 1: 2200 -20375.062 0.081 0.035
Chain 1: 2300 -19991.856 0.038 0.019
Chain 1: 2400 -19763.805 0.038 0.019
Chain 1: 2500 -19565.812 0.025 0.015
Chain 1: 2600 -19195.784 0.023 0.015
Chain 1: 2700 -19152.629 0.018 0.012
Chain 1: 2800 -18869.386 0.019 0.015
Chain 1: 2900 -19150.752 0.019 0.015
Chain 1: 3000 -19136.958 0.012 0.012
Chain 1: 3100 -19222.000 0.011 0.012
Chain 1: 3200 -18912.512 0.011 0.015
Chain 1: 3300 -19117.336 0.011 0.012
Chain 1: 3400 -18591.956 0.012 0.015
Chain 1: 3500 -19204.347 0.014 0.015
Chain 1: 3600 -18510.330 0.016 0.015
Chain 1: 3700 -18897.669 0.018 0.016
Chain 1: 3800 -17856.334 0.022 0.020
Chain 1: 3900 -17852.433 0.021 0.020
Chain 1: 4000 -17969.744 0.021 0.020
Chain 1: 4100 -17883.479 0.022 0.020
Chain 1: 4200 -17699.445 0.021 0.020
Chain 1: 4300 -17838.049 0.021 0.020
Chain 1: 4400 -17794.696 0.018 0.010
Chain 1: 4500 -17697.156 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00136 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.6 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12658.746 1.000 1.000
Chain 1: 200 -9589.283 0.660 1.000
Chain 1: 300 -8355.508 0.489 0.320
Chain 1: 400 -8561.564 0.373 0.320
Chain 1: 500 -8454.833 0.301 0.148
Chain 1: 600 -8303.055 0.254 0.148
Chain 1: 700 -8225.205 0.219 0.024
Chain 1: 800 -8233.517 0.192 0.024
Chain 1: 900 -8140.559 0.172 0.018
Chain 1: 1000 -8247.373 0.156 0.018
Chain 1: 1100 -8368.056 0.057 0.014
Chain 1: 1200 -8260.040 0.026 0.013
Chain 1: 1300 -8197.797 0.012 0.013
Chain 1: 1400 -8220.935 0.010 0.013
Chain 1: 1500 -8308.675 0.010 0.011
Chain 1: 1600 -8280.461 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002983 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.83 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58188.110 1.000 1.000
Chain 1: 200 -17895.120 1.626 2.252
Chain 1: 300 -8800.260 1.428 1.033
Chain 1: 400 -8256.469 1.088 1.033
Chain 1: 500 -8553.247 0.877 1.000
Chain 1: 600 -8314.519 0.736 1.000
Chain 1: 700 -8352.584 0.631 0.066
Chain 1: 800 -8458.866 0.554 0.066
Chain 1: 900 -8033.692 0.498 0.053
Chain 1: 1000 -8064.744 0.449 0.053
Chain 1: 1100 -7751.484 0.353 0.040
Chain 1: 1200 -7823.479 0.129 0.035
Chain 1: 1300 -7762.866 0.026 0.029
Chain 1: 1400 -7905.781 0.021 0.018
Chain 1: 1500 -7691.496 0.021 0.018
Chain 1: 1600 -7828.176 0.019 0.017
Chain 1: 1700 -7577.741 0.022 0.018
Chain 1: 1800 -7687.783 0.022 0.018
Chain 1: 1900 -7620.216 0.018 0.017
Chain 1: 2000 -7699.341 0.019 0.017
Chain 1: 2100 -7653.177 0.015 0.014
Chain 1: 2200 -7773.431 0.016 0.015
Chain 1: 2300 -7669.237 0.016 0.015
Chain 1: 2400 -7712.465 0.015 0.014
Chain 1: 2500 -7624.683 0.014 0.014
Chain 1: 2600 -7587.252 0.012 0.012
Chain 1: 2700 -7574.975 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003327 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.27 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86898.573 1.000 1.000
Chain 1: 200 -13718.172 3.167 5.335
Chain 1: 300 -10130.319 2.230 1.000
Chain 1: 400 -10858.668 1.689 1.000
Chain 1: 500 -9084.034 1.390 0.354
Chain 1: 600 -8882.328 1.162 0.354
Chain 1: 700 -8620.186 1.001 0.195
Chain 1: 800 -9141.984 0.883 0.195
Chain 1: 900 -8963.781 0.787 0.067
Chain 1: 1000 -8735.932 0.711 0.067
Chain 1: 1100 -8999.629 0.614 0.057 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8618.374 0.085 0.044
Chain 1: 1300 -8857.126 0.052 0.030
Chain 1: 1400 -8834.020 0.045 0.029
Chain 1: 1500 -8738.580 0.027 0.027
Chain 1: 1600 -8840.860 0.026 0.027
Chain 1: 1700 -8927.545 0.024 0.026
Chain 1: 1800 -8531.070 0.023 0.026
Chain 1: 1900 -8632.291 0.022 0.026
Chain 1: 2000 -8603.085 0.020 0.012
Chain 1: 2100 -8724.649 0.018 0.012
Chain 1: 2200 -8504.187 0.016 0.012
Chain 1: 2300 -8661.093 0.015 0.012
Chain 1: 2400 -8674.506 0.015 0.012
Chain 1: 2500 -8644.706 0.015 0.012
Chain 1: 2600 -8647.386 0.013 0.012
Chain 1: 2700 -8553.596 0.014 0.012
Chain 1: 2800 -8524.310 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.005679 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 56.79 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8414515.243 1.000 1.000
Chain 1: 200 -1586103.253 2.653 4.305
Chain 1: 300 -891434.179 2.028 1.000
Chain 1: 400 -458475.831 1.757 1.000
Chain 1: 500 -358589.291 1.461 0.944
Chain 1: 600 -233448.421 1.307 0.944
Chain 1: 700 -119544.210 1.257 0.944
Chain 1: 800 -86700.050 1.147 0.944
Chain 1: 900 -67022.666 1.052 0.779
Chain 1: 1000 -51797.436 0.976 0.779
Chain 1: 1100 -39261.215 0.908 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38432.054 0.480 0.379
Chain 1: 1300 -26387.376 0.448 0.379
Chain 1: 1400 -26104.179 0.354 0.319
Chain 1: 1500 -22691.307 0.341 0.319
Chain 1: 1600 -21906.835 0.291 0.294
Chain 1: 1700 -20781.160 0.201 0.294
Chain 1: 1800 -20725.058 0.164 0.150
Chain 1: 1900 -21050.841 0.136 0.054
Chain 1: 2000 -19562.799 0.114 0.054
Chain 1: 2100 -19801.188 0.084 0.036
Chain 1: 2200 -20027.334 0.083 0.036
Chain 1: 2300 -19644.888 0.039 0.019
Chain 1: 2400 -19417.098 0.039 0.019
Chain 1: 2500 -19219.070 0.025 0.015
Chain 1: 2600 -18849.752 0.023 0.015
Chain 1: 2700 -18806.786 0.018 0.012
Chain 1: 2800 -18523.820 0.019 0.015
Chain 1: 2900 -18804.896 0.019 0.015
Chain 1: 3000 -18791.120 0.012 0.012
Chain 1: 3100 -18876.063 0.011 0.012
Chain 1: 3200 -18567.020 0.012 0.015
Chain 1: 3300 -18771.489 0.011 0.012
Chain 1: 3400 -18246.901 0.012 0.015
Chain 1: 3500 -18858.062 0.015 0.015
Chain 1: 3600 -18165.650 0.016 0.015
Chain 1: 3700 -18551.817 0.018 0.017
Chain 1: 3800 -17512.918 0.023 0.021
Chain 1: 3900 -17509.070 0.021 0.021
Chain 1: 4000 -17626.396 0.022 0.021
Chain 1: 4100 -17540.246 0.022 0.021
Chain 1: 4200 -17356.739 0.021 0.021
Chain 1: 4300 -17494.965 0.021 0.021
Chain 1: 4400 -17452.057 0.018 0.011
Chain 1: 4500 -17354.598 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001199 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 11.99 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48786.038 1.000 1.000
Chain 1: 200 -20090.218 1.214 1.428
Chain 1: 300 -22690.208 0.848 1.000
Chain 1: 400 -34947.279 0.723 1.000
Chain 1: 500 -12851.678 0.923 1.000
Chain 1: 600 -14902.381 0.792 1.000
Chain 1: 700 -14723.711 0.680 0.351
Chain 1: 800 -13325.617 0.608 0.351
Chain 1: 900 -16284.176 0.561 0.182
Chain 1: 1000 -13485.276 0.526 0.208
Chain 1: 1100 -15021.756 0.436 0.182
Chain 1: 1200 -10771.500 0.333 0.182
Chain 1: 1300 -11854.833 0.330 0.182
Chain 1: 1400 -11555.752 0.298 0.138
Chain 1: 1500 -12347.887 0.132 0.105
Chain 1: 1600 -10871.673 0.132 0.105
Chain 1: 1700 -10396.226 0.135 0.105
Chain 1: 1800 -11406.254 0.134 0.102
Chain 1: 1900 -9632.918 0.134 0.102
Chain 1: 2000 -11334.062 0.128 0.102
Chain 1: 2100 -9335.453 0.139 0.136
Chain 1: 2200 -9010.176 0.104 0.091
Chain 1: 2300 -9504.580 0.100 0.089
Chain 1: 2400 -8988.548 0.103 0.089
Chain 1: 2500 -8986.671 0.096 0.089
Chain 1: 2600 -9616.373 0.089 0.065
Chain 1: 2700 -15832.986 0.124 0.089
Chain 1: 2800 -10049.126 0.173 0.150
Chain 1: 2900 -10999.653 0.163 0.086
Chain 1: 3000 -10198.448 0.156 0.079
Chain 1: 3100 -8938.712 0.149 0.079
Chain 1: 3200 -8866.064 0.146 0.079
Chain 1: 3300 -8966.368 0.142 0.079
Chain 1: 3400 -9107.722 0.137 0.079
Chain 1: 3500 -10624.654 0.152 0.086
Chain 1: 3600 -12333.063 0.159 0.139
Chain 1: 3700 -9703.250 0.147 0.139
Chain 1: 3800 -15902.621 0.128 0.139
Chain 1: 3900 -11752.103 0.155 0.141
Chain 1: 4000 -8555.030 0.184 0.143
Chain 1: 4100 -9062.539 0.176 0.143
Chain 1: 4200 -12553.671 0.203 0.271
Chain 1: 4300 -8763.675 0.245 0.278
Chain 1: 4400 -9415.100 0.250 0.278
Chain 1: 4500 -9330.455 0.237 0.278
Chain 1: 4600 -8705.177 0.230 0.278
Chain 1: 4700 -10525.095 0.221 0.278
Chain 1: 4800 -8355.731 0.208 0.260
Chain 1: 4900 -8369.894 0.172 0.173
Chain 1: 5000 -11839.785 0.164 0.173
Chain 1: 5100 -8311.712 0.201 0.260
Chain 1: 5200 -8906.255 0.180 0.173
Chain 1: 5300 -9285.399 0.141 0.072
Chain 1: 5400 -8578.687 0.142 0.082
Chain 1: 5500 -13036.241 0.176 0.173
Chain 1: 5600 -9002.494 0.213 0.260
Chain 1: 5700 -13077.263 0.227 0.293
Chain 1: 5800 -9106.406 0.245 0.312
Chain 1: 5900 -8427.046 0.253 0.312
Chain 1: 6000 -8955.942 0.229 0.312
Chain 1: 6100 -9114.745 0.188 0.082
Chain 1: 6200 -8537.420 0.189 0.082
Chain 1: 6300 -8342.696 0.187 0.082
Chain 1: 6400 -10588.275 0.200 0.212
Chain 1: 6500 -8490.637 0.190 0.212
Chain 1: 6600 -8297.543 0.148 0.081
Chain 1: 6700 -12555.743 0.151 0.081
Chain 1: 6800 -8538.776 0.154 0.081
Chain 1: 6900 -8190.047 0.150 0.068
Chain 1: 7000 -8820.998 0.151 0.072
Chain 1: 7100 -8534.749 0.153 0.072
Chain 1: 7200 -12086.528 0.176 0.212
Chain 1: 7300 -8736.631 0.212 0.247
Chain 1: 7400 -12194.633 0.219 0.284
Chain 1: 7500 -10344.776 0.212 0.284
Chain 1: 7600 -8662.729 0.229 0.284
Chain 1: 7700 -8597.490 0.196 0.194
Chain 1: 7800 -10230.299 0.165 0.179
Chain 1: 7900 -8696.921 0.178 0.179
Chain 1: 8000 -9857.956 0.183 0.179
Chain 1: 8100 -8825.829 0.191 0.179
Chain 1: 8200 -10993.489 0.182 0.179
Chain 1: 8300 -8174.486 0.178 0.179
Chain 1: 8400 -8113.369 0.150 0.176
Chain 1: 8500 -12468.470 0.167 0.176
Chain 1: 8600 -8352.522 0.197 0.176
Chain 1: 8700 -10849.374 0.219 0.197
Chain 1: 8800 -8562.531 0.230 0.230
Chain 1: 8900 -8144.635 0.217 0.230
Chain 1: 9000 -9741.826 0.222 0.230
Chain 1: 9100 -11207.115 0.223 0.230
Chain 1: 9200 -8429.849 0.237 0.267
Chain 1: 9300 -8102.012 0.206 0.230
Chain 1: 9400 -8229.274 0.207 0.230
Chain 1: 9500 -8200.776 0.172 0.164
Chain 1: 9600 -9600.395 0.138 0.146
Chain 1: 9700 -10300.352 0.122 0.131
Chain 1: 9800 -8071.023 0.122 0.131
Chain 1: 9900 -8667.326 0.124 0.131
Chain 1: 10000 -8033.595 0.116 0.079
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001422 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.22 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56539.293 1.000 1.000
Chain 1: 200 -17085.357 1.655 2.309
Chain 1: 300 -8594.462 1.432 1.000
Chain 1: 400 -7909.664 1.096 1.000
Chain 1: 500 -8306.587 0.886 0.988
Chain 1: 600 -8237.822 0.740 0.988
Chain 1: 700 -7687.695 0.644 0.087
Chain 1: 800 -8075.138 0.570 0.087
Chain 1: 900 -7757.051 0.511 0.072
Chain 1: 1000 -7670.068 0.461 0.072
Chain 1: 1100 -7620.653 0.362 0.048
Chain 1: 1200 -7566.663 0.132 0.048
Chain 1: 1300 -7583.910 0.033 0.041
Chain 1: 1400 -7817.254 0.027 0.030
Chain 1: 1500 -7584.615 0.026 0.030
Chain 1: 1600 -7485.931 0.026 0.030
Chain 1: 1700 -7482.105 0.019 0.013
Chain 1: 1800 -7550.450 0.015 0.011
Chain 1: 1900 -7562.184 0.011 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003294 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.94 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85637.148 1.000 1.000
Chain 1: 200 -13225.286 3.238 5.475
Chain 1: 300 -9665.274 2.281 1.000
Chain 1: 400 -10450.845 1.730 1.000
Chain 1: 500 -8577.051 1.427 0.368
Chain 1: 600 -8210.250 1.197 0.368
Chain 1: 700 -8363.551 1.029 0.218
Chain 1: 800 -8748.714 0.906 0.218
Chain 1: 900 -8517.610 0.808 0.075
Chain 1: 1000 -8266.201 0.730 0.075
Chain 1: 1100 -8458.533 0.632 0.045 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8175.980 0.088 0.044
Chain 1: 1300 -8234.208 0.052 0.035
Chain 1: 1400 -8222.504 0.045 0.030
Chain 1: 1500 -8260.301 0.023 0.027
Chain 1: 1600 -8270.199 0.019 0.023
Chain 1: 1700 -8195.164 0.018 0.023
Chain 1: 1800 -8081.641 0.015 0.014
Chain 1: 1900 -8200.284 0.014 0.014
Chain 1: 2000 -8160.215 0.011 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.006036 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 60.36 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8387636.068 1.000 1.000
Chain 1: 200 -1578809.311 2.656 4.313
Chain 1: 300 -889687.540 2.029 1.000
Chain 1: 400 -456848.691 1.759 1.000
Chain 1: 500 -357390.412 1.463 0.947
Chain 1: 600 -232542.229 1.308 0.947
Chain 1: 700 -118919.067 1.258 0.947
Chain 1: 800 -86112.395 1.148 0.947
Chain 1: 900 -66473.169 1.054 0.775
Chain 1: 1000 -51268.783 0.978 0.775
Chain 1: 1100 -38748.778 0.910 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37925.216 0.481 0.381
Chain 1: 1300 -25892.338 0.450 0.381
Chain 1: 1400 -25610.624 0.356 0.323
Chain 1: 1500 -22199.931 0.344 0.323
Chain 1: 1600 -21416.594 0.294 0.297
Chain 1: 1700 -20292.134 0.204 0.295
Chain 1: 1800 -20236.701 0.166 0.154
Chain 1: 1900 -20562.317 0.138 0.055
Chain 1: 2000 -19075.426 0.116 0.055
Chain 1: 2100 -19313.629 0.085 0.037
Chain 1: 2200 -19539.483 0.084 0.037
Chain 1: 2300 -19157.404 0.040 0.020
Chain 1: 2400 -18929.716 0.040 0.020
Chain 1: 2500 -18731.591 0.025 0.016
Chain 1: 2600 -18362.277 0.024 0.016
Chain 1: 2700 -18319.573 0.019 0.012
Chain 1: 2800 -18036.505 0.020 0.016
Chain 1: 2900 -18317.587 0.020 0.015
Chain 1: 3000 -18303.847 0.012 0.012
Chain 1: 3100 -18388.681 0.011 0.012
Chain 1: 3200 -18079.736 0.012 0.015
Chain 1: 3300 -18284.249 0.011 0.012
Chain 1: 3400 -17759.690 0.013 0.015
Chain 1: 3500 -18370.673 0.015 0.016
Chain 1: 3600 -17678.665 0.017 0.016
Chain 1: 3700 -18064.402 0.019 0.017
Chain 1: 3800 -17026.020 0.023 0.021
Chain 1: 3900 -17022.257 0.022 0.021
Chain 1: 4000 -17139.554 0.022 0.021
Chain 1: 4100 -17053.281 0.022 0.021
Chain 1: 4200 -16870.112 0.022 0.021
Chain 1: 4300 -17008.141 0.022 0.021
Chain 1: 4400 -16965.312 0.019 0.011
Chain 1: 4500 -16867.943 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001235 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.35 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48870.028 1.000 1.000
Chain 1: 200 -14896.216 1.640 2.281
Chain 1: 300 -12913.505 1.145 1.000
Chain 1: 400 -41814.033 1.031 1.000
Chain 1: 500 -12129.784 1.315 1.000
Chain 1: 600 -16328.674 1.138 1.000
Chain 1: 700 -12666.339 1.017 0.691
Chain 1: 800 -14187.502 0.903 0.691
Chain 1: 900 -13372.848 0.810 0.289
Chain 1: 1000 -11979.768 0.740 0.289
Chain 1: 1100 -10919.919 0.650 0.257 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -10502.840 0.426 0.154
Chain 1: 1300 -19360.971 0.456 0.257
Chain 1: 1400 -13276.497 0.433 0.257
Chain 1: 1500 -12983.166 0.191 0.116
Chain 1: 1600 -10241.365 0.192 0.116
Chain 1: 1700 -10326.719 0.164 0.107
Chain 1: 1800 -11733.511 0.165 0.116
Chain 1: 1900 -10338.515 0.172 0.120
Chain 1: 2000 -9920.317 0.165 0.120
Chain 1: 2100 -10937.460 0.164 0.120
Chain 1: 2200 -10702.161 0.163 0.120
Chain 1: 2300 -13485.482 0.138 0.120
Chain 1: 2400 -9296.614 0.137 0.120
Chain 1: 2500 -17192.982 0.180 0.135
Chain 1: 2600 -16181.735 0.160 0.120
Chain 1: 2700 -8991.123 0.239 0.135
Chain 1: 2800 -10154.838 0.239 0.135
Chain 1: 2900 -15315.643 0.259 0.206
Chain 1: 3000 -9046.511 0.324 0.337
Chain 1: 3100 -11325.154 0.335 0.337
Chain 1: 3200 -8813.450 0.361 0.337
Chain 1: 3300 -10158.787 0.354 0.337
Chain 1: 3400 -16415.895 0.347 0.337
Chain 1: 3500 -9542.108 0.373 0.337
Chain 1: 3600 -13497.225 0.396 0.337
Chain 1: 3700 -8722.068 0.371 0.337
Chain 1: 3800 -8720.060 0.359 0.337
Chain 1: 3900 -10176.486 0.340 0.293
Chain 1: 4000 -8846.185 0.285 0.285
Chain 1: 4100 -9866.848 0.276 0.285
Chain 1: 4200 -14749.312 0.280 0.293
Chain 1: 4300 -9267.584 0.326 0.331
Chain 1: 4400 -8967.654 0.291 0.293
Chain 1: 4500 -8860.196 0.221 0.150
Chain 1: 4600 -13119.295 0.224 0.150
Chain 1: 4700 -8595.554 0.222 0.150
Chain 1: 4800 -13114.443 0.256 0.325
Chain 1: 4900 -8935.682 0.289 0.331
Chain 1: 5000 -9469.175 0.279 0.331
Chain 1: 5100 -8731.646 0.277 0.331
Chain 1: 5200 -16576.792 0.291 0.345
Chain 1: 5300 -10192.443 0.295 0.345
Chain 1: 5400 -8860.277 0.307 0.345
Chain 1: 5500 -13017.459 0.337 0.345
Chain 1: 5600 -13164.137 0.306 0.345
Chain 1: 5700 -13319.581 0.255 0.319
Chain 1: 5800 -9202.164 0.265 0.319
Chain 1: 5900 -9598.721 0.222 0.150
Chain 1: 6000 -8955.653 0.224 0.150
Chain 1: 6100 -9103.824 0.217 0.150
Chain 1: 6200 -13421.881 0.202 0.150
Chain 1: 6300 -8670.069 0.194 0.150
Chain 1: 6400 -8318.504 0.183 0.072
Chain 1: 6500 -9326.573 0.162 0.072
Chain 1: 6600 -8431.238 0.171 0.106
Chain 1: 6700 -8424.166 0.170 0.106
Chain 1: 6800 -8420.856 0.126 0.072
Chain 1: 6900 -12056.033 0.152 0.106
Chain 1: 7000 -12938.937 0.151 0.106
Chain 1: 7100 -13309.028 0.153 0.106
Chain 1: 7200 -8492.095 0.177 0.106
Chain 1: 7300 -10920.442 0.144 0.106
Chain 1: 7400 -14172.256 0.163 0.108
Chain 1: 7500 -8232.645 0.225 0.222
Chain 1: 7600 -8488.203 0.217 0.222
Chain 1: 7700 -8829.217 0.221 0.222
Chain 1: 7800 -9037.713 0.223 0.222
Chain 1: 7900 -8316.207 0.202 0.087
Chain 1: 8000 -8466.997 0.196 0.087
Chain 1: 8100 -9717.236 0.207 0.129
Chain 1: 8200 -8291.774 0.167 0.129
Chain 1: 8300 -8064.356 0.148 0.087
Chain 1: 8400 -13580.440 0.165 0.087
Chain 1: 8500 -8674.731 0.150 0.087
Chain 1: 8600 -11528.593 0.171 0.129
Chain 1: 8700 -8337.643 0.206 0.172
Chain 1: 8800 -8449.415 0.205 0.172
Chain 1: 8900 -10490.403 0.216 0.195
Chain 1: 9000 -8255.270 0.241 0.248
Chain 1: 9100 -8634.650 0.232 0.248
Chain 1: 9200 -8471.143 0.217 0.248
Chain 1: 9300 -8402.519 0.215 0.248
Chain 1: 9400 -8052.071 0.179 0.195
Chain 1: 9500 -7928.095 0.124 0.044
Chain 1: 9600 -8351.336 0.104 0.044
Chain 1: 9700 -9371.983 0.077 0.044
Chain 1: 9800 -11133.254 0.091 0.051
Chain 1: 9900 -8286.585 0.106 0.051
Chain 1: 10000 -8311.475 0.079 0.044
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001448 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.48 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -46553.884 1.000 1.000
Chain 1: 200 -15634.378 1.489 1.978
Chain 1: 300 -8754.569 1.255 1.000
Chain 1: 400 -8503.910 0.948 1.000
Chain 1: 500 -8760.661 0.764 0.786
Chain 1: 600 -8774.876 0.637 0.786
Chain 1: 700 -8116.958 0.558 0.081
Chain 1: 800 -8249.598 0.490 0.081
Chain 1: 900 -7965.662 0.440 0.036
Chain 1: 1000 -7858.761 0.397 0.036
Chain 1: 1100 -7805.073 0.298 0.029
Chain 1: 1200 -7614.124 0.102 0.029
Chain 1: 1300 -7785.959 0.026 0.025
Chain 1: 1400 -7974.398 0.025 0.024
Chain 1: 1500 -7617.535 0.027 0.024
Chain 1: 1600 -7830.550 0.030 0.025
Chain 1: 1700 -7555.841 0.025 0.025
Chain 1: 1800 -7620.635 0.025 0.025
Chain 1: 1900 -7638.468 0.021 0.024
Chain 1: 2000 -7666.471 0.020 0.024
Chain 1: 2100 -7633.314 0.020 0.024
Chain 1: 2200 -7739.767 0.019 0.022
Chain 1: 2300 -7606.292 0.018 0.018
Chain 1: 2400 -7679.578 0.017 0.014
Chain 1: 2500 -7657.706 0.013 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003821 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 38.21 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86200.737 1.000 1.000
Chain 1: 200 -13564.662 3.177 5.355
Chain 1: 300 -9851.628 2.244 1.000
Chain 1: 400 -11113.353 1.711 1.000
Chain 1: 500 -8852.156 1.420 0.377
Chain 1: 600 -8964.038 1.186 0.377
Chain 1: 700 -8637.079 1.022 0.255
Chain 1: 800 -8174.201 0.901 0.255
Chain 1: 900 -8262.174 0.802 0.114
Chain 1: 1000 -8511.132 0.725 0.114
Chain 1: 1100 -8647.118 0.626 0.057 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8222.339 0.096 0.052
Chain 1: 1300 -8534.479 0.062 0.038
Chain 1: 1400 -8454.331 0.052 0.037
Chain 1: 1500 -8369.072 0.027 0.029
Chain 1: 1600 -8479.597 0.027 0.029
Chain 1: 1700 -8546.598 0.024 0.016
Chain 1: 1800 -8110.675 0.024 0.016
Chain 1: 1900 -8215.214 0.024 0.016
Chain 1: 2000 -8191.141 0.021 0.013
Chain 1: 2100 -8335.807 0.022 0.013
Chain 1: 2200 -8122.855 0.019 0.013
Chain 1: 2300 -8278.403 0.017 0.013
Chain 1: 2400 -8118.417 0.018 0.017
Chain 1: 2500 -8189.242 0.018 0.017
Chain 1: 2600 -8101.529 0.018 0.017
Chain 1: 2700 -8135.481 0.018 0.017
Chain 1: 2800 -8095.704 0.013 0.013
Chain 1: 2900 -8188.773 0.012 0.011
Chain 1: 3000 -8020.375 0.014 0.017
Chain 1: 3100 -8178.328 0.014 0.019
Chain 1: 3200 -8050.403 0.013 0.016
Chain 1: 3300 -8058.152 0.012 0.011
Chain 1: 3400 -8216.185 0.012 0.011
Chain 1: 3500 -8221.052 0.011 0.011
Chain 1: 3600 -8007.038 0.012 0.016
Chain 1: 3700 -8152.413 0.014 0.018
Chain 1: 3800 -8013.647 0.015 0.018
Chain 1: 3900 -7948.338 0.015 0.018
Chain 1: 4000 -8023.319 0.014 0.017
Chain 1: 4100 -8014.220 0.012 0.016
Chain 1: 4200 -8003.903 0.010 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00378 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 37.8 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8380957.220 1.000 1.000
Chain 1: 200 -1581251.602 2.650 4.300
Chain 1: 300 -890377.455 2.025 1.000
Chain 1: 400 -457508.870 1.756 1.000
Chain 1: 500 -358267.624 1.460 0.946
Chain 1: 600 -233348.673 1.306 0.946
Chain 1: 700 -119458.637 1.255 0.946
Chain 1: 800 -86623.695 1.146 0.946
Chain 1: 900 -66947.758 1.051 0.776
Chain 1: 1000 -51730.264 0.976 0.776
Chain 1: 1100 -39186.210 0.908 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38363.570 0.480 0.379
Chain 1: 1300 -26294.892 0.448 0.379
Chain 1: 1400 -26013.154 0.354 0.320
Chain 1: 1500 -22593.051 0.342 0.320
Chain 1: 1600 -21807.584 0.292 0.294
Chain 1: 1700 -20678.119 0.202 0.294
Chain 1: 1800 -20621.702 0.164 0.151
Chain 1: 1900 -20948.215 0.137 0.055
Chain 1: 2000 -19456.748 0.115 0.055
Chain 1: 2100 -19695.451 0.084 0.036
Chain 1: 2200 -19922.340 0.083 0.036
Chain 1: 2300 -19539.022 0.039 0.020
Chain 1: 2400 -19310.933 0.039 0.020
Chain 1: 2500 -19112.934 0.025 0.016
Chain 1: 2600 -18742.822 0.023 0.016
Chain 1: 2700 -18699.632 0.018 0.012
Chain 1: 2800 -18416.342 0.019 0.015
Chain 1: 2900 -18697.773 0.019 0.015
Chain 1: 3000 -18683.992 0.012 0.012
Chain 1: 3100 -18769.030 0.011 0.012
Chain 1: 3200 -18459.460 0.012 0.015
Chain 1: 3300 -18664.350 0.011 0.012
Chain 1: 3400 -18138.841 0.012 0.015
Chain 1: 3500 -18751.397 0.015 0.015
Chain 1: 3600 -18057.170 0.017 0.015
Chain 1: 3700 -18444.678 0.018 0.017
Chain 1: 3800 -17402.981 0.023 0.021
Chain 1: 3900 -17399.064 0.021 0.021
Chain 1: 4000 -17516.390 0.022 0.021
Chain 1: 4100 -17430.091 0.022 0.021
Chain 1: 4200 -17245.997 0.021 0.021
Chain 1: 4300 -17384.652 0.021 0.021
Chain 1: 4400 -17341.229 0.019 0.011
Chain 1: 4500 -17243.694 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001426 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.26 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48351.930 1.000 1.000
Chain 1: 200 -43027.722 0.562 1.000
Chain 1: 300 -20691.889 0.734 1.000
Chain 1: 400 -17347.467 0.599 1.000
Chain 1: 500 -21822.449 0.520 0.205
Chain 1: 600 -11470.150 0.584 0.903
Chain 1: 700 -11266.237 0.503 0.205
Chain 1: 800 -12440.994 0.452 0.205
Chain 1: 900 -10936.381 0.417 0.193
Chain 1: 1000 -10348.467 0.381 0.193
Chain 1: 1100 -15576.963 0.315 0.193
Chain 1: 1200 -18315.750 0.317 0.193
Chain 1: 1300 -11075.652 0.275 0.193
Chain 1: 1400 -13744.987 0.275 0.194
Chain 1: 1500 -23610.488 0.296 0.194
Chain 1: 1600 -10229.342 0.337 0.194
Chain 1: 1700 -10707.855 0.339 0.194
Chain 1: 1800 -9529.535 0.342 0.194
Chain 1: 1900 -10389.513 0.337 0.194
Chain 1: 2000 -15229.982 0.363 0.318
Chain 1: 2100 -9571.743 0.388 0.318
Chain 1: 2200 -9230.260 0.377 0.318
Chain 1: 2300 -8691.408 0.318 0.194
Chain 1: 2400 -8669.194 0.299 0.124
Chain 1: 2500 -9852.523 0.269 0.120
Chain 1: 2600 -9582.615 0.141 0.083
Chain 1: 2700 -19416.587 0.187 0.120
Chain 1: 2800 -9029.266 0.290 0.120
Chain 1: 2900 -8842.865 0.284 0.120
Chain 1: 3000 -9005.285 0.254 0.062
Chain 1: 3100 -8661.930 0.199 0.040
Chain 1: 3200 -8229.785 0.200 0.053
Chain 1: 3300 -10793.029 0.218 0.053
Chain 1: 3400 -14617.300 0.244 0.120
Chain 1: 3500 -8863.221 0.296 0.237
Chain 1: 3600 -10356.256 0.308 0.237
Chain 1: 3700 -9555.109 0.266 0.144
Chain 1: 3800 -9633.908 0.152 0.084
Chain 1: 3900 -8903.364 0.158 0.084
Chain 1: 4000 -9033.290 0.157 0.084
Chain 1: 4100 -9666.933 0.160 0.084
Chain 1: 4200 -12803.697 0.179 0.144
Chain 1: 4300 -9053.957 0.197 0.144
Chain 1: 4400 -8506.915 0.177 0.084
Chain 1: 4500 -9077.991 0.118 0.082
Chain 1: 4600 -11786.940 0.127 0.082
Chain 1: 4700 -10521.977 0.131 0.082
Chain 1: 4800 -8438.248 0.155 0.120
Chain 1: 4900 -12948.632 0.181 0.230
Chain 1: 5000 -9207.842 0.220 0.245
Chain 1: 5100 -8288.838 0.225 0.245
Chain 1: 5200 -9007.805 0.208 0.230
Chain 1: 5300 -8737.572 0.170 0.120
Chain 1: 5400 -8110.126 0.171 0.120
Chain 1: 5500 -9053.215 0.175 0.120
Chain 1: 5600 -9472.370 0.157 0.111
Chain 1: 5700 -8251.744 0.160 0.111
Chain 1: 5800 -8820.531 0.141 0.104
Chain 1: 5900 -10908.379 0.126 0.104
Chain 1: 6000 -8836.256 0.109 0.104
Chain 1: 6100 -9396.152 0.103 0.080
Chain 1: 6200 -8490.815 0.106 0.104
Chain 1: 6300 -10958.434 0.126 0.107
Chain 1: 6400 -10211.942 0.125 0.107
Chain 1: 6500 -12172.829 0.131 0.148
Chain 1: 6600 -8064.784 0.177 0.161
Chain 1: 6700 -10381.757 0.185 0.191
Chain 1: 6800 -11721.009 0.190 0.191
Chain 1: 6900 -8505.755 0.208 0.223
Chain 1: 7000 -14191.052 0.225 0.223
Chain 1: 7100 -8416.194 0.288 0.225
Chain 1: 7200 -8144.727 0.280 0.225
Chain 1: 7300 -9247.417 0.270 0.223
Chain 1: 7400 -8220.582 0.275 0.223
Chain 1: 7500 -10297.179 0.279 0.223
Chain 1: 7600 -8605.472 0.248 0.202
Chain 1: 7700 -7920.290 0.234 0.197
Chain 1: 7800 -8208.057 0.226 0.197
Chain 1: 7900 -8162.264 0.189 0.125
Chain 1: 8000 -7944.445 0.152 0.119
Chain 1: 8100 -10648.763 0.108 0.119
Chain 1: 8200 -8711.500 0.127 0.125
Chain 1: 8300 -8053.205 0.124 0.125
Chain 1: 8400 -8135.763 0.112 0.087
Chain 1: 8500 -8353.188 0.095 0.082
Chain 1: 8600 -9166.569 0.084 0.082
Chain 1: 8700 -10846.816 0.091 0.082
Chain 1: 8800 -9267.880 0.104 0.089
Chain 1: 8900 -8677.082 0.110 0.089
Chain 1: 9000 -9971.584 0.121 0.130
Chain 1: 9100 -7884.552 0.122 0.130
Chain 1: 9200 -10075.442 0.121 0.130
Chain 1: 9300 -7996.348 0.139 0.155
Chain 1: 9400 -10206.435 0.160 0.170
Chain 1: 9500 -11835.450 0.171 0.170
Chain 1: 9600 -8053.848 0.209 0.217
Chain 1: 9700 -8590.341 0.200 0.217
Chain 1: 9800 -10725.868 0.203 0.217
Chain 1: 9900 -8303.339 0.225 0.217
Chain 1: 10000 -8305.204 0.212 0.217
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001483 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.83 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57456.147 1.000 1.000
Chain 1: 200 -17145.384 1.676 2.351
Chain 1: 300 -8480.785 1.458 1.022
Chain 1: 400 -7869.778 1.113 1.022
Chain 1: 500 -8324.894 0.901 1.000
Chain 1: 600 -8412.185 0.753 1.000
Chain 1: 700 -7961.530 0.653 0.078
Chain 1: 800 -7960.093 0.572 0.078
Chain 1: 900 -7698.383 0.512 0.057
Chain 1: 1000 -7806.456 0.462 0.057
Chain 1: 1100 -7643.353 0.364 0.055
Chain 1: 1200 -7604.737 0.130 0.034
Chain 1: 1300 -7684.334 0.028 0.021
Chain 1: 1400 -7899.485 0.023 0.021
Chain 1: 1500 -7642.294 0.021 0.021
Chain 1: 1600 -7582.589 0.021 0.021
Chain 1: 1700 -7525.196 0.016 0.014
Chain 1: 1800 -7555.552 0.017 0.014
Chain 1: 1900 -7565.868 0.013 0.010
Chain 1: 2000 -7607.804 0.012 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003225 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.25 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86381.799 1.000 1.000
Chain 1: 200 -12978.212 3.328 5.656
Chain 1: 300 -9461.888 2.343 1.000
Chain 1: 400 -10306.444 1.777 1.000
Chain 1: 500 -8344.207 1.469 0.372
Chain 1: 600 -8060.243 1.230 0.372
Chain 1: 700 -8375.933 1.060 0.235
Chain 1: 800 -8603.577 0.931 0.235
Chain 1: 900 -8366.796 0.830 0.082
Chain 1: 1000 -8105.574 0.750 0.082
Chain 1: 1100 -8393.081 0.654 0.038 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8141.866 0.091 0.035
Chain 1: 1300 -8071.102 0.055 0.034
Chain 1: 1400 -8104.312 0.047 0.032
Chain 1: 1500 -8110.149 0.024 0.031
Chain 1: 1600 -8114.332 0.020 0.028
Chain 1: 1700 -8056.561 0.017 0.026
Chain 1: 1800 -7933.677 0.016 0.015
Chain 1: 1900 -8047.897 0.015 0.014
Chain 1: 2000 -8009.049 0.012 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004145 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 41.45 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8422919.672 1.000 1.000
Chain 1: 200 -1589821.425 2.649 4.298
Chain 1: 300 -891393.273 2.027 1.000
Chain 1: 400 -457470.417 1.758 1.000
Chain 1: 500 -357328.866 1.462 0.949
Chain 1: 600 -232095.338 1.308 0.949
Chain 1: 700 -118470.403 1.258 0.949
Chain 1: 800 -85712.043 1.149 0.949
Chain 1: 900 -66090.168 1.054 0.784
Chain 1: 1000 -50910.015 0.979 0.784
Chain 1: 1100 -38418.513 0.911 0.540 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37589.157 0.484 0.382
Chain 1: 1300 -25596.016 0.452 0.382
Chain 1: 1400 -25314.800 0.358 0.325
Chain 1: 1500 -21916.097 0.346 0.325
Chain 1: 1600 -21135.166 0.296 0.298
Chain 1: 1700 -20016.326 0.205 0.297
Chain 1: 1800 -19961.454 0.167 0.155
Chain 1: 1900 -20286.837 0.139 0.056
Chain 1: 2000 -18803.072 0.117 0.056
Chain 1: 2100 -19041.157 0.086 0.037
Chain 1: 2200 -19266.510 0.085 0.037
Chain 1: 2300 -18884.859 0.040 0.020
Chain 1: 2400 -18657.321 0.040 0.020
Chain 1: 2500 -18459.059 0.026 0.016
Chain 1: 2600 -18090.404 0.024 0.016
Chain 1: 2700 -18047.627 0.019 0.013
Chain 1: 2800 -17764.794 0.020 0.016
Chain 1: 2900 -18045.549 0.020 0.016
Chain 1: 3000 -18031.847 0.012 0.013
Chain 1: 3100 -18116.732 0.011 0.012
Chain 1: 3200 -17808.016 0.012 0.016
Chain 1: 3300 -18012.220 0.011 0.012
Chain 1: 3400 -17488.153 0.013 0.016
Chain 1: 3500 -18098.479 0.015 0.016
Chain 1: 3600 -17407.123 0.017 0.016
Chain 1: 3700 -17792.463 0.019 0.017
Chain 1: 3800 -16755.194 0.024 0.022
Chain 1: 3900 -16751.364 0.022 0.022
Chain 1: 4000 -16868.697 0.023 0.022
Chain 1: 4100 -16782.654 0.023 0.022
Chain 1: 4200 -16599.495 0.022 0.022
Chain 1: 4300 -16737.490 0.022 0.022
Chain 1: 4400 -16694.867 0.019 0.011
Chain 1: 4500 -16597.451 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001478 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.78 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12316.205 1.000 1.000
Chain 1: 200 -9209.914 0.669 1.000
Chain 1: 300 -7986.440 0.497 0.337
Chain 1: 400 -8112.587 0.377 0.337
Chain 1: 500 -8104.425 0.301 0.153
Chain 1: 600 -7917.792 0.255 0.153
Chain 1: 700 -7853.414 0.220 0.024
Chain 1: 800 -7868.525 0.193 0.024
Chain 1: 900 -7732.209 0.173 0.018
Chain 1: 1000 -7872.385 0.158 0.018
Chain 1: 1100 -7857.895 0.058 0.018
Chain 1: 1200 -7850.992 0.024 0.016
Chain 1: 1300 -7800.917 0.009 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.0018 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 18 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58371.495 1.000 1.000
Chain 1: 200 -17669.819 1.652 2.303
Chain 1: 300 -8632.818 1.450 1.047
Chain 1: 400 -8175.249 1.102 1.047
Chain 1: 500 -8302.283 0.884 1.000
Chain 1: 600 -8685.920 0.744 1.000
Chain 1: 700 -7701.758 0.656 0.128
Chain 1: 800 -7972.107 0.578 0.128
Chain 1: 900 -7811.319 0.516 0.056
Chain 1: 1000 -7987.220 0.467 0.056
Chain 1: 1100 -7609.320 0.372 0.050
Chain 1: 1200 -7527.951 0.143 0.044
Chain 1: 1300 -7581.680 0.039 0.034
Chain 1: 1400 -7768.652 0.036 0.024
Chain 1: 1500 -7548.404 0.037 0.029
Chain 1: 1600 -7689.434 0.034 0.024
Chain 1: 1700 -7443.262 0.025 0.024
Chain 1: 1800 -7520.954 0.023 0.022
Chain 1: 1900 -7503.408 0.021 0.022
Chain 1: 2000 -7544.765 0.019 0.018
Chain 1: 2100 -7515.255 0.014 0.011
Chain 1: 2200 -7638.504 0.015 0.016
Chain 1: 2300 -7539.222 0.016 0.016
Chain 1: 2400 -7579.274 0.014 0.013
Chain 1: 2500 -7520.055 0.012 0.010
Chain 1: 2600 -7492.884 0.010 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003308 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.08 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85843.838 1.000 1.000
Chain 1: 200 -13429.831 3.196 5.392
Chain 1: 300 -9808.169 2.254 1.000
Chain 1: 400 -10630.889 1.710 1.000
Chain 1: 500 -8784.523 1.410 0.369
Chain 1: 600 -8236.351 1.186 0.369
Chain 1: 700 -8469.536 1.020 0.210
Chain 1: 800 -9237.269 0.903 0.210
Chain 1: 900 -8613.285 0.811 0.083
Chain 1: 1000 -8457.650 0.732 0.083
Chain 1: 1100 -8582.896 0.633 0.077 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8160.003 0.099 0.072
Chain 1: 1300 -8491.676 0.066 0.067
Chain 1: 1400 -8501.054 0.058 0.052
Chain 1: 1500 -8385.123 0.039 0.039
Chain 1: 1600 -8494.578 0.033 0.028
Chain 1: 1700 -8570.793 0.032 0.018
Chain 1: 1800 -8159.474 0.028 0.018
Chain 1: 1900 -8255.304 0.022 0.015
Chain 1: 2000 -8228.445 0.021 0.014
Chain 1: 2100 -8350.980 0.021 0.014
Chain 1: 2200 -8170.856 0.018 0.014
Chain 1: 2300 -8250.438 0.015 0.013
Chain 1: 2400 -8320.042 0.016 0.013
Chain 1: 2500 -8265.356 0.015 0.012
Chain 1: 2600 -8264.662 0.014 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003477 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.77 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8422774.670 1.000 1.000
Chain 1: 200 -1583010.236 2.660 4.321
Chain 1: 300 -889331.865 2.034 1.000
Chain 1: 400 -457003.367 1.762 1.000
Chain 1: 500 -357453.058 1.465 0.946
Chain 1: 600 -232624.365 1.310 0.946
Chain 1: 700 -119012.823 1.259 0.946
Chain 1: 800 -86294.090 1.149 0.946
Chain 1: 900 -66657.662 1.054 0.780
Chain 1: 1000 -51473.459 0.979 0.780
Chain 1: 1100 -38972.911 0.911 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38148.201 0.481 0.379
Chain 1: 1300 -26121.943 0.449 0.379
Chain 1: 1400 -25841.981 0.355 0.321
Chain 1: 1500 -22435.111 0.343 0.321
Chain 1: 1600 -21653.712 0.292 0.295
Chain 1: 1700 -20529.222 0.203 0.295
Chain 1: 1800 -20473.849 0.165 0.152
Chain 1: 1900 -20800.002 0.137 0.055
Chain 1: 2000 -19312.470 0.115 0.055
Chain 1: 2100 -19550.479 0.084 0.036
Chain 1: 2200 -19777.027 0.083 0.036
Chain 1: 2300 -19394.216 0.039 0.020
Chain 1: 2400 -19166.360 0.039 0.020
Chain 1: 2500 -18968.477 0.025 0.016
Chain 1: 2600 -18598.586 0.024 0.016
Chain 1: 2700 -18555.523 0.018 0.012
Chain 1: 2800 -18272.506 0.020 0.015
Chain 1: 2900 -18553.679 0.020 0.015
Chain 1: 3000 -18539.787 0.012 0.012
Chain 1: 3100 -18624.821 0.011 0.012
Chain 1: 3200 -18315.498 0.012 0.015
Chain 1: 3300 -18520.213 0.011 0.012
Chain 1: 3400 -17995.206 0.013 0.015
Chain 1: 3500 -18607.014 0.015 0.015
Chain 1: 3600 -17913.729 0.017 0.015
Chain 1: 3700 -18300.533 0.019 0.017
Chain 1: 3800 -17260.359 0.023 0.021
Chain 1: 3900 -17256.525 0.022 0.021
Chain 1: 4000 -17373.801 0.022 0.021
Chain 1: 4100 -17287.630 0.022 0.021
Chain 1: 4200 -17103.864 0.022 0.021
Chain 1: 4300 -17242.248 0.021 0.021
Chain 1: 4400 -17199.080 0.019 0.011
Chain 1: 4500 -17101.634 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001528 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.28 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12745.895 1.000 1.000
Chain 1: 200 -9592.634 0.664 1.000
Chain 1: 300 -8331.691 0.493 0.329
Chain 1: 400 -8517.204 0.375 0.329
Chain 1: 500 -8372.723 0.304 0.151
Chain 1: 600 -8229.997 0.256 0.151
Chain 1: 700 -8308.898 0.221 0.022
Chain 1: 800 -8161.735 0.195 0.022
Chain 1: 900 -8218.357 0.175 0.018
Chain 1: 1000 -8090.535 0.159 0.018
Chain 1: 1100 -8219.011 0.060 0.017
Chain 1: 1200 -8141.079 0.028 0.017
Chain 1: 1300 -8074.361 0.014 0.016
Chain 1: 1400 -8104.040 0.012 0.016
Chain 1: 1500 -8202.291 0.012 0.012
Chain 1: 1600 -8131.042 0.011 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004155 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 41.55 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -62161.127 1.000 1.000
Chain 1: 200 -18381.767 1.691 2.382
Chain 1: 300 -9081.251 1.469 1.024
Chain 1: 400 -8854.741 1.108 1.024
Chain 1: 500 -8533.995 0.894 1.000
Chain 1: 600 -8488.206 0.746 1.000
Chain 1: 700 -8352.597 0.642 0.038
Chain 1: 800 -8112.599 0.565 0.038
Chain 1: 900 -7505.835 0.511 0.038
Chain 1: 1000 -8096.611 0.467 0.073
Chain 1: 1100 -7812.485 0.371 0.038
Chain 1: 1200 -7848.246 0.133 0.036
Chain 1: 1300 -7547.944 0.035 0.036
Chain 1: 1400 -7893.026 0.037 0.038
Chain 1: 1500 -7493.766 0.038 0.040
Chain 1: 1600 -7564.373 0.039 0.040
Chain 1: 1700 -7541.978 0.037 0.040
Chain 1: 1800 -7618.917 0.035 0.040
Chain 1: 1900 -7553.770 0.028 0.036
Chain 1: 2000 -7668.965 0.022 0.015
Chain 1: 2100 -7558.301 0.020 0.015
Chain 1: 2200 -7732.712 0.022 0.015
Chain 1: 2300 -7515.888 0.021 0.015
Chain 1: 2400 -7549.660 0.017 0.015
Chain 1: 2500 -7550.370 0.012 0.010
Chain 1: 2600 -7499.223 0.011 0.010
Chain 1: 2700 -7466.613 0.012 0.010
Chain 1: 2800 -7466.238 0.011 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003165 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.65 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86247.768 1.000 1.000
Chain 1: 200 -13980.403 3.085 5.169
Chain 1: 300 -10234.740 2.178 1.000
Chain 1: 400 -11976.837 1.670 1.000
Chain 1: 500 -8882.266 1.406 0.366
Chain 1: 600 -8581.259 1.177 0.366
Chain 1: 700 -8973.401 1.015 0.348
Chain 1: 800 -8853.529 0.890 0.348
Chain 1: 900 -8951.791 0.792 0.145
Chain 1: 1000 -8898.391 0.714 0.145
Chain 1: 1100 -9067.447 0.616 0.044 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8549.141 0.105 0.044
Chain 1: 1300 -8778.643 0.071 0.035
Chain 1: 1400 -8887.783 0.058 0.026
Chain 1: 1500 -8727.679 0.025 0.019
Chain 1: 1600 -8840.469 0.022 0.018
Chain 1: 1700 -8900.117 0.019 0.014
Chain 1: 1800 -8453.821 0.023 0.018
Chain 1: 1900 -8563.071 0.023 0.018
Chain 1: 2000 -8548.094 0.022 0.018
Chain 1: 2100 -8669.679 0.022 0.014
Chain 1: 2200 -8459.081 0.018 0.014
Chain 1: 2300 -8558.125 0.017 0.013
Chain 1: 2400 -8622.455 0.016 0.013
Chain 1: 2500 -8572.370 0.015 0.013
Chain 1: 2600 -8585.835 0.014 0.012
Chain 1: 2700 -8492.811 0.014 0.012
Chain 1: 2800 -8439.150 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004238 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 42.38 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8390855.070 1.000 1.000
Chain 1: 200 -1584809.471 2.647 4.295
Chain 1: 300 -892997.883 2.023 1.000
Chain 1: 400 -459523.150 1.753 1.000
Chain 1: 500 -359789.243 1.458 0.943
Chain 1: 600 -234416.074 1.304 0.943
Chain 1: 700 -120151.802 1.254 0.943
Chain 1: 800 -87268.864 1.144 0.943
Chain 1: 900 -67528.083 1.049 0.775
Chain 1: 1000 -52270.641 0.974 0.775
Chain 1: 1100 -39693.404 0.905 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38867.913 0.478 0.377
Chain 1: 1300 -26755.732 0.446 0.377
Chain 1: 1400 -26471.721 0.353 0.317
Chain 1: 1500 -23041.182 0.340 0.317
Chain 1: 1600 -22253.556 0.290 0.292
Chain 1: 1700 -21118.113 0.200 0.292
Chain 1: 1800 -21060.595 0.163 0.149
Chain 1: 1900 -21387.293 0.135 0.054
Chain 1: 2000 -19892.820 0.113 0.054
Chain 1: 2100 -20131.415 0.083 0.035
Chain 1: 2200 -20359.153 0.082 0.035
Chain 1: 2300 -19975.096 0.038 0.019
Chain 1: 2400 -19746.886 0.039 0.019
Chain 1: 2500 -19549.307 0.025 0.015
Chain 1: 2600 -19178.524 0.023 0.015
Chain 1: 2700 -19135.194 0.018 0.012
Chain 1: 2800 -18851.962 0.019 0.015
Chain 1: 2900 -19133.573 0.019 0.015
Chain 1: 3000 -19119.629 0.012 0.012
Chain 1: 3100 -19204.733 0.011 0.012
Chain 1: 3200 -18894.941 0.011 0.015
Chain 1: 3300 -19100.041 0.011 0.012
Chain 1: 3400 -18574.263 0.012 0.015
Chain 1: 3500 -19187.339 0.014 0.015
Chain 1: 3600 -18492.489 0.016 0.015
Chain 1: 3700 -18880.466 0.018 0.016
Chain 1: 3800 -17837.913 0.022 0.021
Chain 1: 3900 -17834.056 0.021 0.021
Chain 1: 4000 -17951.301 0.022 0.021
Chain 1: 4100 -17864.994 0.022 0.021
Chain 1: 4200 -17680.724 0.021 0.021
Chain 1: 4300 -17819.419 0.021 0.021
Chain 1: 4400 -17775.825 0.018 0.010
Chain 1: 4500 -17678.350 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001321 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.21 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12421.068 1.000 1.000
Chain 1: 200 -9422.632 0.659 1.000
Chain 1: 300 -8256.780 0.486 0.318
Chain 1: 400 -8370.771 0.368 0.318
Chain 1: 500 -8340.227 0.295 0.141
Chain 1: 600 -8314.816 0.247 0.141
Chain 1: 700 -8094.531 0.215 0.027
Chain 1: 800 -8107.939 0.189 0.027
Chain 1: 900 -8241.407 0.169 0.016
Chain 1: 1000 -8116.510 0.154 0.016
Chain 1: 1100 -8157.998 0.055 0.015
Chain 1: 1200 -8106.508 0.023 0.014
Chain 1: 1300 -8040.354 0.010 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001447 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.47 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -62060.883 1.000 1.000
Chain 1: 200 -17994.235 1.724 2.449
Chain 1: 300 -8882.323 1.492 1.026
Chain 1: 400 -9586.158 1.137 1.026
Chain 1: 500 -8129.285 0.945 1.000
Chain 1: 600 -8265.462 0.791 1.000
Chain 1: 700 -8102.605 0.681 0.179
Chain 1: 800 -7775.888 0.601 0.179
Chain 1: 900 -7977.740 0.537 0.073
Chain 1: 1000 -7898.018 0.484 0.073
Chain 1: 1100 -7589.653 0.388 0.042
Chain 1: 1200 -7789.334 0.146 0.041
Chain 1: 1300 -7540.959 0.047 0.033
Chain 1: 1400 -7587.881 0.040 0.026
Chain 1: 1500 -7492.974 0.023 0.025
Chain 1: 1600 -7660.059 0.024 0.025
Chain 1: 1700 -7484.878 0.024 0.025
Chain 1: 1800 -7561.468 0.021 0.023
Chain 1: 1900 -7552.440 0.018 0.022
Chain 1: 2000 -7537.810 0.018 0.022
Chain 1: 2100 -7523.465 0.014 0.013
Chain 1: 2200 -7659.242 0.013 0.013
Chain 1: 2300 -7540.569 0.011 0.013
Chain 1: 2400 -7569.465 0.011 0.013
Chain 1: 2500 -7500.619 0.011 0.010
Chain 1: 2600 -7468.093 0.009 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003537 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.37 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85862.503 1.000 1.000
Chain 1: 200 -13620.880 3.152 5.304
Chain 1: 300 -10028.350 2.221 1.000
Chain 1: 400 -10776.388 1.683 1.000
Chain 1: 500 -8953.516 1.387 0.358
Chain 1: 600 -8667.231 1.161 0.358
Chain 1: 700 -8857.058 0.998 0.204
Chain 1: 800 -9369.562 0.881 0.204
Chain 1: 900 -8875.471 0.789 0.069
Chain 1: 1000 -8536.082 0.714 0.069
Chain 1: 1100 -8873.026 0.618 0.056 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8551.515 0.091 0.055
Chain 1: 1300 -8741.159 0.057 0.040
Chain 1: 1400 -8741.378 0.051 0.038
Chain 1: 1500 -8605.453 0.032 0.038
Chain 1: 1600 -8715.139 0.030 0.038
Chain 1: 1700 -8803.666 0.029 0.038
Chain 1: 1800 -8400.359 0.028 0.038
Chain 1: 1900 -8497.869 0.023 0.022
Chain 1: 2000 -8469.637 0.020 0.016
Chain 1: 2100 -8589.517 0.017 0.014
Chain 1: 2200 -8393.077 0.016 0.014
Chain 1: 2300 -8533.533 0.016 0.014
Chain 1: 2400 -8536.842 0.016 0.014
Chain 1: 2500 -8512.100 0.014 0.013
Chain 1: 2600 -8511.617 0.013 0.011
Chain 1: 2700 -8422.096 0.013 0.011
Chain 1: 2800 -8388.826 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003298 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.98 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8369384.943 1.000 1.000
Chain 1: 200 -1577790.545 2.652 4.304
Chain 1: 300 -890167.535 2.026 1.000
Chain 1: 400 -457503.552 1.756 1.000
Chain 1: 500 -358632.689 1.460 0.946
Chain 1: 600 -233745.180 1.305 0.946
Chain 1: 700 -119714.689 1.255 0.946
Chain 1: 800 -86853.102 1.145 0.946
Chain 1: 900 -67126.437 1.051 0.772
Chain 1: 1000 -51862.759 0.975 0.772
Chain 1: 1100 -39280.884 0.907 0.534 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38448.802 0.479 0.378
Chain 1: 1300 -26344.521 0.448 0.378
Chain 1: 1400 -26057.015 0.354 0.320
Chain 1: 1500 -22628.963 0.342 0.320
Chain 1: 1600 -21840.805 0.292 0.294
Chain 1: 1700 -20707.229 0.202 0.294
Chain 1: 1800 -20649.653 0.165 0.151
Chain 1: 1900 -20975.622 0.137 0.055
Chain 1: 2000 -19483.466 0.115 0.055
Chain 1: 2100 -19721.855 0.084 0.036
Chain 1: 2200 -19948.902 0.083 0.036
Chain 1: 2300 -19565.675 0.039 0.020
Chain 1: 2400 -19337.775 0.039 0.020
Chain 1: 2500 -19140.131 0.025 0.016
Chain 1: 2600 -18770.191 0.023 0.016
Chain 1: 2700 -18727.132 0.018 0.012
Chain 1: 2800 -18444.220 0.019 0.015
Chain 1: 2900 -18725.500 0.019 0.015
Chain 1: 3000 -18711.581 0.012 0.012
Chain 1: 3100 -18796.546 0.011 0.012
Chain 1: 3200 -18487.302 0.012 0.015
Chain 1: 3300 -18691.993 0.011 0.012
Chain 1: 3400 -18167.151 0.012 0.015
Chain 1: 3500 -18778.795 0.015 0.015
Chain 1: 3600 -18085.858 0.017 0.015
Chain 1: 3700 -18472.431 0.018 0.017
Chain 1: 3800 -17432.784 0.023 0.021
Chain 1: 3900 -17429.016 0.021 0.021
Chain 1: 4000 -17546.249 0.022 0.021
Chain 1: 4100 -17460.063 0.022 0.021
Chain 1: 4200 -17276.462 0.021 0.021
Chain 1: 4300 -17414.706 0.021 0.021
Chain 1: 4400 -17371.652 0.018 0.011
Chain 1: 4500 -17274.263 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00133 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.3 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12663.427 1.000 1.000
Chain 1: 200 -9546.448 0.663 1.000
Chain 1: 300 -8103.980 0.502 0.327
Chain 1: 400 -8326.028 0.383 0.327
Chain 1: 500 -8185.996 0.310 0.178
Chain 1: 600 -8041.239 0.261 0.178
Chain 1: 700 -8164.650 0.226 0.027
Chain 1: 800 -7981.242 0.201 0.027
Chain 1: 900 -7892.427 0.180 0.023
Chain 1: 1000 -7955.182 0.162 0.023
Chain 1: 1100 -8075.344 0.064 0.018
Chain 1: 1200 -7963.141 0.033 0.017
Chain 1: 1300 -7908.946 0.015 0.015
Chain 1: 1400 -7916.358 0.013 0.015
Chain 1: 1500 -8019.109 0.012 0.014
Chain 1: 1600 -7919.150 0.012 0.013
Chain 1: 1700 -7893.366 0.011 0.013
Chain 1: 1800 -7866.050 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001435 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.35 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -46589.255 1.000 1.000
Chain 1: 200 -15878.983 1.467 1.934
Chain 1: 300 -8827.590 1.244 1.000
Chain 1: 400 -8400.816 0.946 1.000
Chain 1: 500 -8313.766 0.759 0.799
Chain 1: 600 -8412.196 0.634 0.799
Chain 1: 700 -8572.246 0.546 0.051
Chain 1: 800 -8293.666 0.482 0.051
Chain 1: 900 -8047.205 0.432 0.034
Chain 1: 1000 -7889.721 0.391 0.034
Chain 1: 1100 -7645.315 0.294 0.032
Chain 1: 1200 -8138.771 0.107 0.032
Chain 1: 1300 -7572.299 0.034 0.032
Chain 1: 1400 -7649.436 0.030 0.031
Chain 1: 1500 -7525.429 0.031 0.031
Chain 1: 1600 -7757.975 0.033 0.031
Chain 1: 1700 -7518.757 0.034 0.032
Chain 1: 1800 -7560.284 0.031 0.031
Chain 1: 1900 -7543.738 0.028 0.030
Chain 1: 2000 -7612.620 0.027 0.030
Chain 1: 2100 -7541.681 0.025 0.016
Chain 1: 2200 -7729.041 0.021 0.016
Chain 1: 2300 -7494.559 0.017 0.016
Chain 1: 2400 -7536.568 0.017 0.016
Chain 1: 2500 -7578.865 0.015 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003285 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.85 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86253.611 1.000 1.000
Chain 1: 200 -13831.965 3.118 5.236
Chain 1: 300 -10069.228 2.203 1.000
Chain 1: 400 -11745.339 1.688 1.000
Chain 1: 500 -8703.204 1.420 0.374
Chain 1: 600 -8677.620 1.184 0.374
Chain 1: 700 -8427.437 1.019 0.350
Chain 1: 800 -9039.869 0.900 0.350
Chain 1: 900 -8812.632 0.803 0.143
Chain 1: 1000 -8669.818 0.724 0.143
Chain 1: 1100 -8719.723 0.625 0.068 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8358.085 0.106 0.043
Chain 1: 1300 -8817.927 0.074 0.043
Chain 1: 1400 -8558.824 0.062 0.030
Chain 1: 1500 -8557.434 0.027 0.030
Chain 1: 1600 -8650.249 0.028 0.030
Chain 1: 1700 -8710.083 0.026 0.026
Chain 1: 1800 -8267.556 0.024 0.026
Chain 1: 1900 -8377.044 0.023 0.016
Chain 1: 2000 -8366.232 0.022 0.013
Chain 1: 2100 -8540.564 0.023 0.020
Chain 1: 2200 -8272.348 0.022 0.020
Chain 1: 2300 -8456.093 0.019 0.020
Chain 1: 2400 -8272.672 0.018 0.020
Chain 1: 2500 -8350.215 0.019 0.020
Chain 1: 2600 -8266.810 0.019 0.020
Chain 1: 2700 -8294.277 0.019 0.020
Chain 1: 2800 -8247.587 0.014 0.013
Chain 1: 2900 -8355.358 0.014 0.013
Chain 1: 3000 -8306.169 0.014 0.013
Chain 1: 3100 -8239.012 0.013 0.010
Chain 1: 3200 -8212.124 0.010 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00362 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 36.2 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8389048.532 1.000 1.000
Chain 1: 200 -1585354.500 2.646 4.292
Chain 1: 300 -892362.522 2.023 1.000
Chain 1: 400 -458280.387 1.754 1.000
Chain 1: 500 -358798.652 1.459 0.947
Chain 1: 600 -233700.249 1.305 0.947
Chain 1: 700 -119778.146 1.254 0.947
Chain 1: 800 -86919.796 1.145 0.947
Chain 1: 900 -67247.402 1.050 0.777
Chain 1: 1000 -52035.326 0.974 0.777
Chain 1: 1100 -39487.032 0.906 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38668.628 0.479 0.378
Chain 1: 1300 -26589.648 0.447 0.378
Chain 1: 1400 -26308.802 0.353 0.318
Chain 1: 1500 -22885.564 0.340 0.318
Chain 1: 1600 -22099.550 0.290 0.293
Chain 1: 1700 -20968.731 0.201 0.292
Chain 1: 1800 -20912.144 0.163 0.150
Chain 1: 1900 -21238.944 0.135 0.054
Chain 1: 2000 -19746.070 0.114 0.054
Chain 1: 2100 -19984.901 0.083 0.036
Chain 1: 2200 -20212.115 0.082 0.036
Chain 1: 2300 -19828.469 0.039 0.019
Chain 1: 2400 -19600.251 0.039 0.019
Chain 1: 2500 -19402.275 0.025 0.015
Chain 1: 2600 -19031.857 0.023 0.015
Chain 1: 2700 -18988.605 0.018 0.012
Chain 1: 2800 -18705.146 0.019 0.015
Chain 1: 2900 -18986.759 0.019 0.015
Chain 1: 3000 -18972.960 0.012 0.012
Chain 1: 3100 -19058.026 0.011 0.012
Chain 1: 3200 -18748.271 0.011 0.015
Chain 1: 3300 -18953.325 0.011 0.012
Chain 1: 3400 -18427.441 0.012 0.015
Chain 1: 3500 -19040.553 0.014 0.015
Chain 1: 3600 -18345.671 0.016 0.015
Chain 1: 3700 -18733.647 0.018 0.017
Chain 1: 3800 -17690.897 0.023 0.021
Chain 1: 3900 -17686.971 0.021 0.021
Chain 1: 4000 -17804.294 0.022 0.021
Chain 1: 4100 -17717.917 0.022 0.021
Chain 1: 4200 -17533.623 0.021 0.021
Chain 1: 4300 -17672.411 0.021 0.021
Chain 1: 4400 -17628.804 0.018 0.011
Chain 1: 4500 -17531.245 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001587 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.87 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48999.815 1.000 1.000
Chain 1: 200 -20155.331 1.216 1.431
Chain 1: 300 -16571.538 0.882 1.000
Chain 1: 400 -19039.101 0.694 1.000
Chain 1: 500 -21951.159 0.582 0.216
Chain 1: 600 -12223.576 0.618 0.796
Chain 1: 700 -14819.099 0.554 0.216
Chain 1: 800 -12507.126 0.508 0.216
Chain 1: 900 -10856.175 0.469 0.185
Chain 1: 1000 -10719.453 0.423 0.185
Chain 1: 1100 -10414.338 0.326 0.175
Chain 1: 1200 -16383.001 0.219 0.175
Chain 1: 1300 -13606.690 0.218 0.175
Chain 1: 1400 -25324.920 0.251 0.185
Chain 1: 1500 -10179.522 0.387 0.204
Chain 1: 1600 -11168.818 0.316 0.185
Chain 1: 1700 -9555.196 0.316 0.185
Chain 1: 1800 -19286.016 0.348 0.204
Chain 1: 1900 -10562.546 0.415 0.364
Chain 1: 2000 -9942.268 0.420 0.364
Chain 1: 2100 -9584.972 0.421 0.364
Chain 1: 2200 -10547.470 0.393 0.204
Chain 1: 2300 -14731.173 0.401 0.284
Chain 1: 2400 -8904.341 0.421 0.284
Chain 1: 2500 -10435.324 0.286 0.169
Chain 1: 2600 -9419.000 0.288 0.169
Chain 1: 2700 -14211.806 0.305 0.284
Chain 1: 2800 -10421.257 0.291 0.284
Chain 1: 2900 -9825.400 0.215 0.147
Chain 1: 3000 -8687.861 0.221 0.147
Chain 1: 3100 -9446.811 0.226 0.147
Chain 1: 3200 -15900.689 0.257 0.284
Chain 1: 3300 -16102.849 0.230 0.147
Chain 1: 3400 -10051.407 0.225 0.147
Chain 1: 3500 -9646.728 0.214 0.131
Chain 1: 3600 -9631.004 0.204 0.131
Chain 1: 3700 -9472.746 0.172 0.080
Chain 1: 3800 -9261.228 0.138 0.061
Chain 1: 3900 -15493.040 0.172 0.080
Chain 1: 4000 -8566.018 0.239 0.080
Chain 1: 4100 -8754.720 0.234 0.042
Chain 1: 4200 -10629.186 0.211 0.042
Chain 1: 4300 -10180.792 0.214 0.044
Chain 1: 4400 -9004.176 0.167 0.044
Chain 1: 4500 -10490.483 0.177 0.131
Chain 1: 4600 -9027.167 0.193 0.142
Chain 1: 4700 -9802.257 0.199 0.142
Chain 1: 4800 -8500.176 0.212 0.153
Chain 1: 4900 -10102.438 0.188 0.153
Chain 1: 5000 -13807.380 0.134 0.153
Chain 1: 5100 -11085.932 0.156 0.159
Chain 1: 5200 -9216.082 0.159 0.159
Chain 1: 5300 -11347.898 0.173 0.162
Chain 1: 5400 -8367.318 0.196 0.188
Chain 1: 5500 -14006.834 0.222 0.203
Chain 1: 5600 -10518.438 0.239 0.245
Chain 1: 5700 -8400.887 0.256 0.252
Chain 1: 5800 -10523.930 0.261 0.252
Chain 1: 5900 -10425.158 0.246 0.252
Chain 1: 6000 -11262.977 0.226 0.245
Chain 1: 6100 -10337.073 0.211 0.203
Chain 1: 6200 -8499.686 0.212 0.216
Chain 1: 6300 -10141.378 0.210 0.216
Chain 1: 6400 -10041.560 0.175 0.202
Chain 1: 6500 -9131.653 0.145 0.162
Chain 1: 6600 -8427.690 0.120 0.100
Chain 1: 6700 -10044.278 0.111 0.100
Chain 1: 6800 -8902.611 0.103 0.100
Chain 1: 6900 -9396.554 0.108 0.100
Chain 1: 7000 -11893.727 0.121 0.128
Chain 1: 7100 -12736.235 0.119 0.128
Chain 1: 7200 -8508.308 0.147 0.128
Chain 1: 7300 -9354.045 0.140 0.100
Chain 1: 7400 -8459.704 0.149 0.106
Chain 1: 7500 -8966.508 0.145 0.106
Chain 1: 7600 -9229.020 0.140 0.106
Chain 1: 7700 -8441.920 0.133 0.093
Chain 1: 7800 -9089.222 0.127 0.090
Chain 1: 7900 -8594.304 0.128 0.090
Chain 1: 8000 -8175.388 0.112 0.071
Chain 1: 8100 -8377.597 0.108 0.071
Chain 1: 8200 -10925.910 0.081 0.071
Chain 1: 8300 -8178.343 0.106 0.071
Chain 1: 8400 -8026.834 0.097 0.058
Chain 1: 8500 -8255.601 0.094 0.058
Chain 1: 8600 -10226.589 0.111 0.071
Chain 1: 8700 -8272.353 0.125 0.071
Chain 1: 8800 -8489.418 0.120 0.058
Chain 1: 8900 -10894.193 0.137 0.193
Chain 1: 9000 -8053.500 0.167 0.221
Chain 1: 9100 -10492.733 0.188 0.232
Chain 1: 9200 -9915.011 0.170 0.221
Chain 1: 9300 -9292.824 0.143 0.193
Chain 1: 9400 -8349.315 0.153 0.193
Chain 1: 9500 -8094.163 0.153 0.193
Chain 1: 9600 -8229.503 0.135 0.113
Chain 1: 9700 -8382.433 0.114 0.067
Chain 1: 9800 -9278.279 0.121 0.097
Chain 1: 9900 -10149.964 0.107 0.086
Chain 1: 10000 -8139.440 0.097 0.086
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00195 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 19.5 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -63152.200 1.000 1.000
Chain 1: 200 -18077.278 1.747 2.493
Chain 1: 300 -8724.161 1.522 1.072
Chain 1: 400 -8408.803 1.151 1.072
Chain 1: 500 -8706.216 0.927 1.000
Chain 1: 600 -8639.582 0.774 1.000
Chain 1: 700 -7943.913 0.676 0.088
Chain 1: 800 -7979.047 0.592 0.088
Chain 1: 900 -7892.730 0.528 0.038
Chain 1: 1000 -7769.419 0.476 0.038
Chain 1: 1100 -7683.821 0.377 0.034
Chain 1: 1200 -7608.927 0.129 0.016
Chain 1: 1300 -7793.369 0.024 0.016
Chain 1: 1400 -7877.834 0.022 0.011
Chain 1: 1500 -7617.471 0.022 0.011
Chain 1: 1600 -7744.540 0.022 0.016
Chain 1: 1700 -7512.553 0.017 0.016
Chain 1: 1800 -7558.918 0.017 0.016
Chain 1: 1900 -7560.394 0.016 0.016
Chain 1: 2000 -7593.125 0.015 0.011
Chain 1: 2100 -7595.006 0.014 0.011
Chain 1: 2200 -7692.731 0.014 0.013
Chain 1: 2300 -7606.206 0.013 0.011
Chain 1: 2400 -7637.847 0.012 0.011
Chain 1: 2500 -7561.751 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003765 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 37.65 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86315.182 1.000 1.000
Chain 1: 200 -13350.945 3.233 5.465
Chain 1: 300 -9782.098 2.277 1.000
Chain 1: 400 -10593.201 1.727 1.000
Chain 1: 500 -8684.380 1.425 0.365
Chain 1: 600 -8425.371 1.193 0.365
Chain 1: 700 -8492.044 1.024 0.220
Chain 1: 800 -8750.440 0.899 0.220
Chain 1: 900 -8657.824 0.801 0.077
Chain 1: 1000 -8327.070 0.724 0.077
Chain 1: 1100 -8675.926 0.629 0.040 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8394.987 0.085 0.040
Chain 1: 1300 -8524.006 0.050 0.033
Chain 1: 1400 -8515.548 0.043 0.031
Chain 1: 1500 -8390.647 0.022 0.030
Chain 1: 1600 -8497.720 0.021 0.015
Chain 1: 1700 -8583.414 0.021 0.015
Chain 1: 1800 -8191.074 0.023 0.015
Chain 1: 1900 -8292.693 0.023 0.015
Chain 1: 2000 -8263.114 0.019 0.015
Chain 1: 2100 -8387.797 0.017 0.015
Chain 1: 2200 -8172.160 0.016 0.015
Chain 1: 2300 -8321.419 0.016 0.015
Chain 1: 2400 -8336.462 0.016 0.015
Chain 1: 2500 -8304.018 0.015 0.013
Chain 1: 2600 -8306.198 0.014 0.012
Chain 1: 2700 -8212.800 0.014 0.012
Chain 1: 2800 -8185.101 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003774 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 37.74 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8426765.073 1.000 1.000
Chain 1: 200 -1587546.039 2.654 4.308
Chain 1: 300 -891009.034 2.030 1.000
Chain 1: 400 -457641.000 1.759 1.000
Chain 1: 500 -357674.166 1.463 0.947
Chain 1: 600 -232578.173 1.309 0.947
Chain 1: 700 -118917.118 1.259 0.947
Chain 1: 800 -86166.317 1.149 0.947
Chain 1: 900 -66528.647 1.054 0.782
Chain 1: 1000 -51343.316 0.978 0.782
Chain 1: 1100 -38844.099 0.910 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38017.901 0.482 0.380
Chain 1: 1300 -26006.146 0.450 0.380
Chain 1: 1400 -25725.576 0.356 0.322
Chain 1: 1500 -22322.362 0.343 0.322
Chain 1: 1600 -21541.164 0.293 0.296
Chain 1: 1700 -20419.201 0.203 0.295
Chain 1: 1800 -20364.112 0.165 0.152
Chain 1: 1900 -20689.856 0.137 0.055
Chain 1: 2000 -19204.261 0.116 0.055
Chain 1: 2100 -19442.207 0.085 0.036
Chain 1: 2200 -19668.159 0.084 0.036
Chain 1: 2300 -19285.978 0.039 0.020
Chain 1: 2400 -19058.287 0.039 0.020
Chain 1: 2500 -18860.227 0.025 0.016
Chain 1: 2600 -18490.849 0.024 0.016
Chain 1: 2700 -18447.976 0.018 0.012
Chain 1: 2800 -18164.991 0.020 0.016
Chain 1: 2900 -18446.037 0.020 0.015
Chain 1: 3000 -18432.222 0.012 0.012
Chain 1: 3100 -18517.142 0.011 0.012
Chain 1: 3200 -18208.107 0.012 0.015
Chain 1: 3300 -18412.631 0.011 0.012
Chain 1: 3400 -17888.016 0.013 0.015
Chain 1: 3500 -18499.114 0.015 0.016
Chain 1: 3600 -17806.844 0.017 0.016
Chain 1: 3700 -18192.844 0.019 0.017
Chain 1: 3800 -17154.111 0.023 0.021
Chain 1: 3900 -17150.306 0.022 0.021
Chain 1: 4000 -17267.618 0.022 0.021
Chain 1: 4100 -17181.441 0.022 0.021
Chain 1: 4200 -16998.046 0.022 0.021
Chain 1: 4300 -17136.187 0.021 0.021
Chain 1: 4400 -17093.288 0.019 0.011
Chain 1: 4500 -16995.889 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001656 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 16.56 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49077.044 1.000 1.000
Chain 1: 200 -19172.134 1.280 1.560
Chain 1: 300 -20020.665 0.867 1.000
Chain 1: 400 -13279.185 0.777 1.000
Chain 1: 500 -18745.905 0.680 0.508
Chain 1: 600 -17341.980 0.580 0.508
Chain 1: 700 -11764.803 0.565 0.474
Chain 1: 800 -23116.290 0.556 0.491
Chain 1: 900 -12943.619 0.581 0.491
Chain 1: 1000 -12233.526 0.529 0.491
Chain 1: 1100 -14045.021 0.442 0.474
Chain 1: 1200 -11070.771 0.313 0.292
Chain 1: 1300 -13508.143 0.327 0.292
Chain 1: 1400 -16646.454 0.295 0.269
Chain 1: 1500 -9989.267 0.332 0.269
Chain 1: 1600 -10030.408 0.325 0.269
Chain 1: 1700 -17576.006 0.320 0.269
Chain 1: 1800 -19922.060 0.283 0.189
Chain 1: 1900 -11002.752 0.285 0.189
Chain 1: 2000 -12400.733 0.291 0.189
Chain 1: 2100 -11245.045 0.288 0.189
Chain 1: 2200 -9883.839 0.275 0.180
Chain 1: 2300 -9404.289 0.262 0.138
Chain 1: 2400 -17998.799 0.291 0.138
Chain 1: 2500 -9554.754 0.313 0.138
Chain 1: 2600 -9738.051 0.314 0.138
Chain 1: 2700 -9826.724 0.272 0.118
Chain 1: 2800 -9254.829 0.267 0.113
Chain 1: 2900 -9241.856 0.186 0.103
Chain 1: 3000 -9050.913 0.176 0.062
Chain 1: 3100 -11749.930 0.189 0.062
Chain 1: 3200 -20208.394 0.217 0.062
Chain 1: 3300 -16430.099 0.235 0.230
Chain 1: 3400 -15985.342 0.190 0.062
Chain 1: 3500 -9560.861 0.169 0.062
Chain 1: 3600 -9161.231 0.171 0.062
Chain 1: 3700 -9823.711 0.177 0.067
Chain 1: 3800 -9184.712 0.178 0.070
Chain 1: 3900 -14859.673 0.216 0.230
Chain 1: 4000 -9096.789 0.277 0.230
Chain 1: 4100 -11904.575 0.278 0.236
Chain 1: 4200 -8938.323 0.269 0.236
Chain 1: 4300 -12366.820 0.274 0.277
Chain 1: 4400 -8980.186 0.309 0.332
Chain 1: 4500 -15003.969 0.282 0.332
Chain 1: 4600 -10237.118 0.324 0.377
Chain 1: 4700 -9942.468 0.320 0.377
Chain 1: 4800 -13499.844 0.340 0.377
Chain 1: 4900 -9678.561 0.341 0.377
Chain 1: 5000 -10156.493 0.282 0.332
Chain 1: 5100 -8736.926 0.275 0.332
Chain 1: 5200 -16323.914 0.288 0.377
Chain 1: 5300 -14390.448 0.274 0.377
Chain 1: 5400 -15346.702 0.243 0.264
Chain 1: 5500 -8667.596 0.280 0.264
Chain 1: 5600 -9550.219 0.242 0.162
Chain 1: 5700 -12309.389 0.262 0.224
Chain 1: 5800 -8872.179 0.274 0.224
Chain 1: 5900 -14472.086 0.273 0.224
Chain 1: 6000 -9279.657 0.324 0.387
Chain 1: 6100 -9633.699 0.312 0.387
Chain 1: 6200 -8550.901 0.278 0.224
Chain 1: 6300 -9586.430 0.275 0.224
Chain 1: 6400 -9051.488 0.275 0.224
Chain 1: 6500 -12918.766 0.228 0.224
Chain 1: 6600 -11093.482 0.235 0.224
Chain 1: 6700 -8476.733 0.244 0.299
Chain 1: 6800 -9042.523 0.211 0.165
Chain 1: 6900 -10406.344 0.186 0.131
Chain 1: 7000 -8564.730 0.151 0.131
Chain 1: 7100 -8457.639 0.149 0.131
Chain 1: 7200 -9600.564 0.148 0.131
Chain 1: 7300 -10864.263 0.149 0.131
Chain 1: 7400 -8820.661 0.166 0.165
Chain 1: 7500 -8344.465 0.142 0.131
Chain 1: 7600 -8548.960 0.128 0.119
Chain 1: 7700 -11346.372 0.122 0.119
Chain 1: 7800 -11260.742 0.116 0.119
Chain 1: 7900 -8816.962 0.131 0.119
Chain 1: 8000 -8436.190 0.114 0.116
Chain 1: 8100 -8561.716 0.114 0.116
Chain 1: 8200 -9100.038 0.108 0.059
Chain 1: 8300 -10735.069 0.112 0.059
Chain 1: 8400 -8871.939 0.109 0.059
Chain 1: 8500 -8529.166 0.108 0.059
Chain 1: 8600 -10161.562 0.121 0.152
Chain 1: 8700 -8418.265 0.117 0.152
Chain 1: 8800 -8295.527 0.118 0.152
Chain 1: 8900 -8681.013 0.095 0.059
Chain 1: 9000 -11712.341 0.116 0.152
Chain 1: 9100 -8280.405 0.156 0.161
Chain 1: 9200 -8837.418 0.157 0.161
Chain 1: 9300 -8390.051 0.147 0.161
Chain 1: 9400 -8375.554 0.126 0.063
Chain 1: 9500 -8413.684 0.122 0.063
Chain 1: 9600 -8504.720 0.107 0.053
Chain 1: 9700 -8296.031 0.089 0.044
Chain 1: 9800 -9906.181 0.104 0.053
Chain 1: 9900 -11141.468 0.111 0.063
Chain 1: 10000 -8247.649 0.120 0.063
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001399 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.99 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61860.791 1.000 1.000
Chain 1: 200 -18099.470 1.709 2.418
Chain 1: 300 -8964.902 1.479 1.019
Chain 1: 400 -9343.124 1.119 1.019
Chain 1: 500 -7955.357 0.930 1.000
Chain 1: 600 -8704.289 0.790 1.000
Chain 1: 700 -7883.390 0.692 0.174
Chain 1: 800 -8526.447 0.615 0.174
Chain 1: 900 -7871.285 0.556 0.104
Chain 1: 1000 -7893.039 0.500 0.104
Chain 1: 1100 -7945.231 0.401 0.086
Chain 1: 1200 -7699.525 0.162 0.083
Chain 1: 1300 -7747.943 0.061 0.075
Chain 1: 1400 -7697.211 0.058 0.075
Chain 1: 1500 -7638.110 0.041 0.032
Chain 1: 1600 -7821.808 0.035 0.023
Chain 1: 1700 -7716.681 0.026 0.014
Chain 1: 1800 -7713.899 0.018 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003394 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.94 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85111.378 1.000 1.000
Chain 1: 200 -13684.879 3.110 5.219
Chain 1: 300 -10068.350 2.193 1.000
Chain 1: 400 -10875.089 1.663 1.000
Chain 1: 500 -9042.007 1.371 0.359
Chain 1: 600 -8518.968 1.153 0.359
Chain 1: 700 -8593.093 0.989 0.203
Chain 1: 800 -8762.653 0.868 0.203
Chain 1: 900 -8892.091 0.773 0.074
Chain 1: 1000 -8698.496 0.698 0.074
Chain 1: 1100 -8862.615 0.600 0.061 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8545.639 0.082 0.037
Chain 1: 1300 -8775.547 0.048 0.026
Chain 1: 1400 -8761.258 0.041 0.022
Chain 1: 1500 -8624.930 0.023 0.019
Chain 1: 1600 -8735.506 0.018 0.019
Chain 1: 1700 -8820.368 0.018 0.019
Chain 1: 1800 -8407.306 0.021 0.019
Chain 1: 1900 -8503.445 0.020 0.019
Chain 1: 2000 -8476.732 0.019 0.016
Chain 1: 2100 -8599.285 0.018 0.014
Chain 1: 2200 -8419.454 0.017 0.014
Chain 1: 2300 -8498.399 0.015 0.013
Chain 1: 2400 -8568.104 0.015 0.013
Chain 1: 2500 -8513.521 0.015 0.011
Chain 1: 2600 -8512.925 0.013 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.007243 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 72.43 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8376387.049 1.000 1.000
Chain 1: 200 -1579658.030 2.651 4.303
Chain 1: 300 -891058.256 2.025 1.000
Chain 1: 400 -458162.589 1.755 1.000
Chain 1: 500 -359025.974 1.459 0.945
Chain 1: 600 -234098.149 1.305 0.945
Chain 1: 700 -119882.300 1.255 0.945
Chain 1: 800 -86990.167 1.145 0.945
Chain 1: 900 -67243.880 1.051 0.773
Chain 1: 1000 -51971.217 0.975 0.773
Chain 1: 1100 -39380.018 0.907 0.534 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38546.791 0.479 0.378
Chain 1: 1300 -26431.180 0.447 0.378
Chain 1: 1400 -26143.689 0.354 0.320
Chain 1: 1500 -22712.369 0.341 0.320
Chain 1: 1600 -21923.318 0.292 0.294
Chain 1: 1700 -20788.340 0.202 0.294
Chain 1: 1800 -20730.416 0.164 0.151
Chain 1: 1900 -21056.513 0.136 0.055
Chain 1: 2000 -19563.164 0.115 0.055
Chain 1: 2100 -19801.711 0.084 0.036
Chain 1: 2200 -20028.984 0.083 0.036
Chain 1: 2300 -19645.491 0.039 0.020
Chain 1: 2400 -19417.502 0.039 0.020
Chain 1: 2500 -19219.896 0.025 0.015
Chain 1: 2600 -18849.792 0.023 0.015
Chain 1: 2700 -18806.623 0.018 0.012
Chain 1: 2800 -18523.687 0.019 0.015
Chain 1: 2900 -18804.974 0.019 0.015
Chain 1: 3000 -18791.134 0.012 0.012
Chain 1: 3100 -18876.140 0.011 0.012
Chain 1: 3200 -18566.748 0.012 0.015
Chain 1: 3300 -18771.495 0.011 0.012
Chain 1: 3400 -18246.467 0.012 0.015
Chain 1: 3500 -18858.439 0.015 0.015
Chain 1: 3600 -18165.009 0.016 0.015
Chain 1: 3700 -18551.961 0.018 0.017
Chain 1: 3800 -17511.620 0.023 0.021
Chain 1: 3900 -17507.813 0.021 0.021
Chain 1: 4000 -17625.051 0.022 0.021
Chain 1: 4100 -17538.871 0.022 0.021
Chain 1: 4200 -17355.077 0.021 0.021
Chain 1: 4300 -17493.455 0.021 0.021
Chain 1: 4400 -17450.260 0.018 0.011
Chain 1: 4500 -17352.846 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001327 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.27 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48821.292 1.000 1.000
Chain 1: 200 -39581.388 0.617 1.000
Chain 1: 300 -33164.898 0.476 0.233
Chain 1: 400 -14036.190 0.697 1.000
Chain 1: 500 -27361.771 0.655 0.487
Chain 1: 600 -11469.620 0.777 1.000
Chain 1: 700 -14819.420 0.698 0.487
Chain 1: 800 -13576.723 0.622 0.487
Chain 1: 900 -14524.139 0.561 0.233
Chain 1: 1000 -10913.879 0.538 0.331
Chain 1: 1100 -20066.718 0.483 0.331
Chain 1: 1200 -17841.541 0.472 0.331
Chain 1: 1300 -11139.183 0.513 0.456 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1400 -13457.016 0.394 0.331
Chain 1: 1500 -9996.866 0.380 0.331
Chain 1: 1600 -19644.618 0.291 0.331
Chain 1: 1700 -9572.267 0.373 0.346
Chain 1: 1800 -17210.319 0.408 0.444
Chain 1: 1900 -10157.246 0.471 0.456
Chain 1: 2000 -10703.548 0.443 0.456
Chain 1: 2100 -13806.818 0.420 0.444
Chain 1: 2200 -10974.932 0.434 0.444
Chain 1: 2300 -16472.729 0.407 0.346
Chain 1: 2400 -9691.004 0.460 0.444
Chain 1: 2500 -9328.988 0.429 0.444
Chain 1: 2600 -9450.108 0.381 0.334
Chain 1: 2700 -10146.443 0.283 0.258
Chain 1: 2800 -14526.059 0.268 0.258
Chain 1: 2900 -9771.516 0.248 0.258
Chain 1: 3000 -9137.054 0.249 0.258
Chain 1: 3100 -9298.410 0.229 0.258
Chain 1: 3200 -10107.760 0.211 0.080
Chain 1: 3300 -9589.879 0.183 0.069
Chain 1: 3400 -9833.678 0.115 0.069
Chain 1: 3500 -9299.051 0.117 0.069
Chain 1: 3600 -10365.441 0.126 0.069
Chain 1: 3700 -8860.879 0.136 0.080
Chain 1: 3800 -8958.906 0.107 0.069
Chain 1: 3900 -9884.560 0.068 0.069
Chain 1: 4000 -9795.436 0.062 0.057
Chain 1: 4100 -9182.781 0.067 0.067
Chain 1: 4200 -11775.410 0.081 0.067
Chain 1: 4300 -9660.105 0.097 0.094
Chain 1: 4400 -9524.604 0.096 0.094
Chain 1: 4500 -9550.015 0.091 0.094
Chain 1: 4600 -8686.787 0.091 0.094
Chain 1: 4700 -13800.600 0.111 0.094
Chain 1: 4800 -8666.553 0.169 0.099
Chain 1: 4900 -13069.650 0.193 0.219
Chain 1: 5000 -11204.425 0.209 0.219
Chain 1: 5100 -9068.249 0.226 0.220
Chain 1: 5200 -15257.475 0.244 0.236
Chain 1: 5300 -14303.670 0.229 0.236
Chain 1: 5400 -13537.789 0.233 0.236
Chain 1: 5500 -12294.570 0.243 0.236
Chain 1: 5600 -12387.711 0.234 0.236
Chain 1: 5700 -8744.685 0.239 0.236
Chain 1: 5800 -17538.198 0.229 0.236
Chain 1: 5900 -14976.399 0.213 0.171
Chain 1: 6000 -9281.923 0.258 0.236
Chain 1: 6100 -11456.151 0.253 0.190
Chain 1: 6200 -10095.679 0.226 0.171
Chain 1: 6300 -8997.197 0.231 0.171
Chain 1: 6400 -10098.220 0.237 0.171
Chain 1: 6500 -13282.523 0.251 0.190
Chain 1: 6600 -9395.027 0.291 0.240
Chain 1: 6700 -8404.025 0.261 0.190
Chain 1: 6800 -10834.852 0.234 0.190
Chain 1: 6900 -9098.640 0.236 0.191
Chain 1: 7000 -13794.365 0.208 0.191
Chain 1: 7100 -8276.803 0.256 0.224
Chain 1: 7200 -11161.848 0.268 0.240
Chain 1: 7300 -10349.082 0.264 0.240
Chain 1: 7400 -12028.803 0.267 0.240
Chain 1: 7500 -8789.544 0.280 0.258
Chain 1: 7600 -9744.365 0.248 0.224
Chain 1: 7700 -8374.184 0.253 0.224
Chain 1: 7800 -10245.241 0.249 0.191
Chain 1: 7900 -10066.423 0.231 0.183
Chain 1: 8000 -9314.077 0.205 0.164
Chain 1: 8100 -8427.142 0.149 0.140
Chain 1: 8200 -8374.399 0.124 0.105
Chain 1: 8300 -8888.663 0.122 0.105
Chain 1: 8400 -8294.568 0.115 0.098
Chain 1: 8500 -8228.273 0.079 0.081
Chain 1: 8600 -8783.641 0.076 0.072
Chain 1: 8700 -8256.607 0.066 0.064
Chain 1: 8800 -8366.207 0.049 0.063
Chain 1: 8900 -8997.440 0.054 0.064
Chain 1: 9000 -9100.050 0.047 0.063
Chain 1: 9100 -8196.839 0.048 0.063
Chain 1: 9200 -9469.363 0.060 0.064
Chain 1: 9300 -11491.522 0.072 0.070
Chain 1: 9400 -9474.795 0.086 0.070
Chain 1: 9500 -8192.747 0.101 0.110
Chain 1: 9600 -8485.454 0.098 0.110
Chain 1: 9700 -10224.911 0.109 0.134
Chain 1: 9800 -8625.706 0.126 0.156
Chain 1: 9900 -10439.638 0.136 0.170
Chain 1: 10000 -8212.970 0.162 0.174
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00148 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.8 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57061.758 1.000 1.000
Chain 1: 200 -17551.638 1.626 2.251
Chain 1: 300 -8790.524 1.416 1.000
Chain 1: 400 -8435.653 1.072 1.000
Chain 1: 500 -7883.044 0.872 0.997
Chain 1: 600 -8221.105 0.734 0.997
Chain 1: 700 -8529.890 0.634 0.070
Chain 1: 800 -8203.487 0.560 0.070
Chain 1: 900 -7867.771 0.502 0.043
Chain 1: 1000 -7669.756 0.455 0.043
Chain 1: 1100 -7745.549 0.356 0.042
Chain 1: 1200 -7791.927 0.131 0.041
Chain 1: 1300 -7734.418 0.032 0.040
Chain 1: 1400 -7747.357 0.028 0.036
Chain 1: 1500 -7625.081 0.023 0.026
Chain 1: 1600 -7819.710 0.021 0.025
Chain 1: 1700 -7576.292 0.021 0.025
Chain 1: 1800 -7693.236 0.018 0.016
Chain 1: 1900 -7602.149 0.015 0.015
Chain 1: 2000 -7697.096 0.014 0.012
Chain 1: 2100 -7701.890 0.013 0.012
Chain 1: 2200 -7736.349 0.013 0.012
Chain 1: 2300 -7661.847 0.013 0.012
Chain 1: 2400 -7705.909 0.013 0.012
Chain 1: 2500 -7632.163 0.013 0.012
Chain 1: 2600 -7576.980 0.011 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003774 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 37.74 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86280.689 1.000 1.000
Chain 1: 200 -13577.889 3.177 5.354
Chain 1: 300 -9933.346 2.240 1.000
Chain 1: 400 -10941.635 1.703 1.000
Chain 1: 500 -8773.819 1.412 0.367
Chain 1: 600 -8376.323 1.185 0.367
Chain 1: 700 -8411.977 1.016 0.247
Chain 1: 800 -9084.143 0.898 0.247
Chain 1: 900 -8785.854 0.802 0.092
Chain 1: 1000 -8599.672 0.724 0.092
Chain 1: 1100 -8658.429 0.625 0.074 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8451.607 0.092 0.047
Chain 1: 1300 -8618.985 0.057 0.034
Chain 1: 1400 -8626.294 0.048 0.024
Chain 1: 1500 -8492.672 0.025 0.022
Chain 1: 1600 -8604.096 0.021 0.019
Chain 1: 1700 -8685.989 0.022 0.019
Chain 1: 1800 -8268.670 0.020 0.019
Chain 1: 1900 -8366.680 0.017 0.016
Chain 1: 2000 -8340.547 0.015 0.013
Chain 1: 2100 -8464.371 0.016 0.015
Chain 1: 2200 -8279.353 0.016 0.015
Chain 1: 2300 -8361.296 0.015 0.013
Chain 1: 2400 -8430.866 0.016 0.013
Chain 1: 2500 -8376.761 0.015 0.012
Chain 1: 2600 -8376.972 0.014 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003626 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 36.26 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8397198.385 1.000 1.000
Chain 1: 200 -1587771.559 2.644 4.289
Chain 1: 300 -891957.847 2.023 1.000
Chain 1: 400 -457633.337 1.754 1.000
Chain 1: 500 -357787.426 1.459 0.949
Chain 1: 600 -232692.846 1.306 0.949
Chain 1: 700 -119102.098 1.255 0.949
Chain 1: 800 -86341.892 1.146 0.949
Chain 1: 900 -66735.799 1.051 0.780
Chain 1: 1000 -51578.519 0.976 0.780
Chain 1: 1100 -39088.325 0.907 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38270.842 0.481 0.379
Chain 1: 1300 -26258.726 0.448 0.379
Chain 1: 1400 -25981.175 0.355 0.320
Chain 1: 1500 -22575.716 0.342 0.320
Chain 1: 1600 -21794.320 0.292 0.294
Chain 1: 1700 -20671.878 0.202 0.294
Chain 1: 1800 -20616.877 0.164 0.151
Chain 1: 1900 -20943.125 0.136 0.054
Chain 1: 2000 -19455.809 0.114 0.054
Chain 1: 2100 -19694.293 0.084 0.036
Chain 1: 2200 -19920.417 0.083 0.036
Chain 1: 2300 -19537.869 0.039 0.020
Chain 1: 2400 -19309.965 0.039 0.020
Chain 1: 2500 -19111.767 0.025 0.016
Chain 1: 2600 -18742.204 0.023 0.016
Chain 1: 2700 -18699.222 0.018 0.012
Chain 1: 2800 -18415.961 0.019 0.015
Chain 1: 2900 -18697.163 0.019 0.015
Chain 1: 3000 -18683.468 0.012 0.012
Chain 1: 3100 -18768.439 0.011 0.012
Chain 1: 3200 -18459.164 0.012 0.015
Chain 1: 3300 -18663.844 0.011 0.012
Chain 1: 3400 -18138.731 0.012 0.015
Chain 1: 3500 -18750.619 0.015 0.015
Chain 1: 3600 -18057.294 0.017 0.015
Chain 1: 3700 -18444.052 0.018 0.017
Chain 1: 3800 -17403.725 0.023 0.021
Chain 1: 3900 -17399.830 0.021 0.021
Chain 1: 4000 -17517.172 0.022 0.021
Chain 1: 4100 -17430.918 0.022 0.021
Chain 1: 4200 -17247.152 0.021 0.021
Chain 1: 4300 -17385.592 0.021 0.021
Chain 1: 4400 -17342.419 0.018 0.011
Chain 1: 4500 -17244.905 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001365 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.65 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12552.223 1.000 1.000
Chain 1: 200 -9246.058 0.679 1.000
Chain 1: 300 -8110.645 0.499 0.358
Chain 1: 400 -8243.772 0.378 0.358
Chain 1: 500 -8203.020 0.304 0.140
Chain 1: 600 -8004.256 0.257 0.140
Chain 1: 700 -7899.759 0.222 0.025
Chain 1: 800 -7902.832 0.195 0.025
Chain 1: 900 -7843.175 0.174 0.016
Chain 1: 1000 -8032.020 0.159 0.024
Chain 1: 1100 -8045.282 0.059 0.016
Chain 1: 1200 -7918.942 0.025 0.016
Chain 1: 1300 -7886.307 0.011 0.013
Chain 1: 1400 -7894.498 0.010 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001562 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.62 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61284.086 1.000 1.000
Chain 1: 200 -18185.314 1.685 2.370
Chain 1: 300 -8849.014 1.475 1.055
Chain 1: 400 -8114.455 1.129 1.055
Chain 1: 500 -7892.527 0.909 1.000
Chain 1: 600 -8153.682 0.763 1.000
Chain 1: 700 -8259.418 0.656 0.091
Chain 1: 800 -8238.003 0.574 0.091
Chain 1: 900 -8028.932 0.513 0.032
Chain 1: 1000 -7911.399 0.463 0.032
Chain 1: 1100 -7763.978 0.365 0.028
Chain 1: 1200 -7589.370 0.130 0.026
Chain 1: 1300 -7747.881 0.027 0.023
Chain 1: 1400 -7833.473 0.019 0.020
Chain 1: 1500 -7664.664 0.018 0.020
Chain 1: 1600 -7553.692 0.017 0.019
Chain 1: 1700 -7586.840 0.016 0.019
Chain 1: 1800 -7658.332 0.016 0.019
Chain 1: 1900 -7650.842 0.014 0.015
Chain 1: 2000 -7695.901 0.013 0.015
Chain 1: 2100 -7640.567 0.012 0.011
Chain 1: 2200 -7766.318 0.011 0.011
Chain 1: 2300 -7602.511 0.011 0.011
Chain 1: 2400 -7589.541 0.010 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003786 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 37.86 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86137.168 1.000 1.000
Chain 1: 200 -13697.097 3.144 5.289
Chain 1: 300 -10002.260 2.219 1.000
Chain 1: 400 -11291.441 1.693 1.000
Chain 1: 500 -8790.678 1.411 0.369
Chain 1: 600 -8693.517 1.178 0.369
Chain 1: 700 -8417.138 1.014 0.284
Chain 1: 800 -8794.183 0.893 0.284
Chain 1: 900 -8715.998 0.795 0.114
Chain 1: 1000 -8383.606 0.719 0.114
Chain 1: 1100 -8756.940 0.623 0.043 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8336.269 0.100 0.043
Chain 1: 1300 -8606.054 0.066 0.043
Chain 1: 1400 -8624.033 0.055 0.040
Chain 1: 1500 -8502.256 0.028 0.033
Chain 1: 1600 -8609.347 0.028 0.033
Chain 1: 1700 -8677.992 0.025 0.031
Chain 1: 1800 -8241.077 0.026 0.031
Chain 1: 1900 -8346.243 0.027 0.031
Chain 1: 2000 -8322.769 0.023 0.014
Chain 1: 2100 -8464.973 0.020 0.014
Chain 1: 2200 -8252.276 0.018 0.014
Chain 1: 2300 -8411.693 0.017 0.014
Chain 1: 2400 -8248.454 0.018 0.017
Chain 1: 2500 -8320.446 0.018 0.017
Chain 1: 2600 -8232.115 0.018 0.017
Chain 1: 2700 -8266.167 0.017 0.017
Chain 1: 2800 -8225.842 0.013 0.013
Chain 1: 2900 -8319.720 0.012 0.011
Chain 1: 3000 -8154.611 0.014 0.017
Chain 1: 3100 -8308.749 0.014 0.019
Chain 1: 3200 -8180.324 0.013 0.016
Chain 1: 3300 -8189.094 0.012 0.011
Chain 1: 3400 -8351.768 0.011 0.011
Chain 1: 3500 -8363.568 0.011 0.011
Chain 1: 3600 -8136.686 0.012 0.016
Chain 1: 3700 -8283.577 0.014 0.018
Chain 1: 3800 -8142.976 0.015 0.018
Chain 1: 3900 -8077.230 0.015 0.018
Chain 1: 4000 -8154.537 0.014 0.017
Chain 1: 4100 -8148.580 0.012 0.016
Chain 1: 4200 -8132.976 0.011 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00469 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 46.9 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8422487.890 1.000 1.000
Chain 1: 200 -1588060.493 2.652 4.304
Chain 1: 300 -891263.096 2.028 1.000
Chain 1: 400 -458440.764 1.757 1.000
Chain 1: 500 -358360.735 1.462 0.944
Chain 1: 600 -233178.911 1.308 0.944
Chain 1: 700 -119347.110 1.257 0.944
Chain 1: 800 -86616.206 1.147 0.944
Chain 1: 900 -66961.804 1.052 0.782
Chain 1: 1000 -51777.920 0.976 0.782
Chain 1: 1100 -39275.413 0.908 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38455.318 0.480 0.378
Chain 1: 1300 -26420.088 0.447 0.378
Chain 1: 1400 -26142.731 0.354 0.318
Chain 1: 1500 -22733.142 0.341 0.318
Chain 1: 1600 -21951.706 0.291 0.294
Chain 1: 1700 -20825.503 0.201 0.293
Chain 1: 1800 -20770.229 0.163 0.150
Chain 1: 1900 -21096.738 0.136 0.054
Chain 1: 2000 -19607.525 0.114 0.054
Chain 1: 2100 -19845.803 0.083 0.036
Chain 1: 2200 -20072.689 0.082 0.036
Chain 1: 2300 -19689.429 0.039 0.019
Chain 1: 2400 -19461.345 0.039 0.019
Chain 1: 2500 -19263.539 0.025 0.015
Chain 1: 2600 -18893.151 0.023 0.015
Chain 1: 2700 -18849.924 0.018 0.012
Chain 1: 2800 -18566.682 0.019 0.015
Chain 1: 2900 -18848.094 0.019 0.015
Chain 1: 3000 -18834.189 0.012 0.012
Chain 1: 3100 -18919.300 0.011 0.012
Chain 1: 3200 -18609.649 0.012 0.015
Chain 1: 3300 -18814.625 0.011 0.012
Chain 1: 3400 -18289.056 0.012 0.015
Chain 1: 3500 -18901.712 0.015 0.015
Chain 1: 3600 -18207.280 0.016 0.015
Chain 1: 3700 -18594.905 0.018 0.017
Chain 1: 3800 -17553.025 0.023 0.021
Chain 1: 3900 -17549.128 0.021 0.021
Chain 1: 4000 -17666.416 0.022 0.021
Chain 1: 4100 -17580.140 0.022 0.021
Chain 1: 4200 -17395.999 0.021 0.021
Chain 1: 4300 -17534.638 0.021 0.021
Chain 1: 4400 -17491.133 0.018 0.011
Chain 1: 4500 -17393.638 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00152 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.2 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12191.565 1.000 1.000
Chain 1: 200 -9157.276 0.666 1.000
Chain 1: 300 -7864.162 0.499 0.331
Chain 1: 400 -8019.718 0.379 0.331
Chain 1: 500 -7887.424 0.306 0.164
Chain 1: 600 -7819.204 0.257 0.164
Chain 1: 700 -7734.005 0.222 0.019
Chain 1: 800 -7775.288 0.195 0.019
Chain 1: 900 -7891.163 0.175 0.017
Chain 1: 1000 -7800.429 0.158 0.017
Chain 1: 1100 -7811.081 0.058 0.015
Chain 1: 1200 -7742.494 0.026 0.012
Chain 1: 1300 -7710.550 0.010 0.011
Chain 1: 1400 -7724.722 0.008 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001568 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.68 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57546.694 1.000 1.000
Chain 1: 200 -17428.022 1.651 2.302
Chain 1: 300 -8569.674 1.445 1.034
Chain 1: 400 -8173.171 1.096 1.034
Chain 1: 500 -7928.928 0.883 1.000
Chain 1: 600 -8470.502 0.746 1.000
Chain 1: 700 -7790.778 0.652 0.087
Chain 1: 800 -8115.562 0.576 0.087
Chain 1: 900 -7945.636 0.514 0.064
Chain 1: 1000 -7724.843 0.466 0.064
Chain 1: 1100 -7786.884 0.366 0.049
Chain 1: 1200 -7567.655 0.139 0.040
Chain 1: 1300 -7780.454 0.038 0.031
Chain 1: 1400 -7909.755 0.035 0.029
Chain 1: 1500 -7568.299 0.037 0.029
Chain 1: 1600 -7579.530 0.030 0.029
Chain 1: 1700 -7512.232 0.023 0.027
Chain 1: 1800 -7553.101 0.019 0.021
Chain 1: 1900 -7611.623 0.018 0.016
Chain 1: 2000 -7574.723 0.015 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003547 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.47 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86547.392 1.000 1.000
Chain 1: 200 -13263.618 3.263 5.525
Chain 1: 300 -9665.403 2.299 1.000
Chain 1: 400 -10388.670 1.742 1.000
Chain 1: 500 -8618.114 1.435 0.372
Chain 1: 600 -8190.292 1.204 0.372
Chain 1: 700 -8145.075 1.033 0.205
Chain 1: 800 -8647.253 0.911 0.205
Chain 1: 900 -8457.769 0.812 0.070
Chain 1: 1000 -8302.433 0.733 0.070
Chain 1: 1100 -8488.195 0.635 0.058 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8058.347 0.088 0.053
Chain 1: 1300 -8366.317 0.054 0.052
Chain 1: 1400 -8375.850 0.048 0.037
Chain 1: 1500 -8257.098 0.028 0.022
Chain 1: 1600 -8362.569 0.024 0.022
Chain 1: 1700 -8448.979 0.025 0.022
Chain 1: 1800 -8049.019 0.024 0.022
Chain 1: 1900 -8148.874 0.023 0.019
Chain 1: 2000 -8119.966 0.022 0.014
Chain 1: 2100 -8239.974 0.021 0.014
Chain 1: 2200 -8026.581 0.018 0.014
Chain 1: 2300 -8179.725 0.016 0.014
Chain 1: 2400 -8061.314 0.018 0.015
Chain 1: 2500 -8124.337 0.017 0.015
Chain 1: 2600 -8145.558 0.016 0.015
Chain 1: 2700 -8064.723 0.016 0.015
Chain 1: 2800 -8038.881 0.011 0.012
Chain 1: 2900 -8094.265 0.011 0.010
Chain 1: 3000 -7978.590 0.012 0.014
Chain 1: 3100 -8116.171 0.012 0.014
Chain 1: 3200 -7996.248 0.011 0.014
Chain 1: 3300 -8017.620 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003586 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.86 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8395329.879 1.000 1.000
Chain 1: 200 -1583206.273 2.651 4.303
Chain 1: 300 -890314.156 2.027 1.000
Chain 1: 400 -457392.847 1.757 1.000
Chain 1: 500 -357994.273 1.461 0.946
Chain 1: 600 -233001.741 1.307 0.946
Chain 1: 700 -119131.127 1.257 0.946
Chain 1: 800 -86275.250 1.147 0.946
Chain 1: 900 -66599.842 1.053 0.778
Chain 1: 1000 -51369.396 0.977 0.778
Chain 1: 1100 -38823.677 0.909 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37995.103 0.481 0.381
Chain 1: 1300 -25939.001 0.450 0.381
Chain 1: 1400 -25654.694 0.356 0.323
Chain 1: 1500 -22238.723 0.344 0.323
Chain 1: 1600 -21453.466 0.294 0.296
Chain 1: 1700 -20326.452 0.204 0.295
Chain 1: 1800 -20270.051 0.166 0.154
Chain 1: 1900 -20595.955 0.138 0.055
Chain 1: 2000 -19107.048 0.116 0.055
Chain 1: 2100 -19345.487 0.085 0.037
Chain 1: 2200 -19571.761 0.084 0.037
Chain 1: 2300 -19189.160 0.040 0.020
Chain 1: 2400 -18961.347 0.040 0.020
Chain 1: 2500 -18763.322 0.025 0.016
Chain 1: 2600 -18393.882 0.024 0.016
Chain 1: 2700 -18350.921 0.019 0.012
Chain 1: 2800 -18067.903 0.020 0.016
Chain 1: 2900 -18349.048 0.020 0.015
Chain 1: 3000 -18335.261 0.012 0.012
Chain 1: 3100 -18420.198 0.011 0.012
Chain 1: 3200 -18111.098 0.012 0.015
Chain 1: 3300 -18315.637 0.011 0.012
Chain 1: 3400 -17790.917 0.013 0.015
Chain 1: 3500 -18402.265 0.015 0.016
Chain 1: 3600 -17709.649 0.017 0.016
Chain 1: 3700 -18095.957 0.019 0.017
Chain 1: 3800 -17056.718 0.023 0.021
Chain 1: 3900 -17052.879 0.022 0.021
Chain 1: 4000 -17170.189 0.022 0.021
Chain 1: 4100 -17084.002 0.022 0.021
Chain 1: 4200 -16900.462 0.022 0.021
Chain 1: 4300 -17038.717 0.022 0.021
Chain 1: 4400 -16995.756 0.019 0.011
Chain 1: 4500 -16898.304 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001214 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.14 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48709.335 1.000 1.000
Chain 1: 200 -19936.501 1.222 1.443
Chain 1: 300 -22602.431 0.854 1.000
Chain 1: 400 -13033.231 0.824 1.000
Chain 1: 500 -16157.296 0.698 0.734
Chain 1: 600 -12547.288 0.629 0.734
Chain 1: 700 -14409.647 0.558 0.288
Chain 1: 800 -11349.027 0.522 0.288
Chain 1: 900 -13613.702 0.482 0.270
Chain 1: 1000 -35815.712 0.496 0.288
Chain 1: 1100 -10293.162 0.644 0.288 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -10075.262 0.502 0.270 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1300 -10187.844 0.491 0.270
Chain 1: 1400 -11022.474 0.425 0.193
Chain 1: 1500 -10426.936 0.412 0.166
Chain 1: 1600 -11847.739 0.395 0.129
Chain 1: 1700 -17664.924 0.415 0.166
Chain 1: 1800 -10480.985 0.457 0.166
Chain 1: 1900 -12698.695 0.457 0.175
Chain 1: 2000 -11497.931 0.406 0.120
Chain 1: 2100 -9118.725 0.184 0.120
Chain 1: 2200 -11130.044 0.200 0.175
Chain 1: 2300 -8977.569 0.223 0.181
Chain 1: 2400 -9677.850 0.222 0.181
Chain 1: 2500 -15851.495 0.256 0.240
Chain 1: 2600 -9441.931 0.312 0.261
Chain 1: 2700 -18800.744 0.328 0.261
Chain 1: 2800 -9336.884 0.361 0.261
Chain 1: 2900 -9310.217 0.344 0.261
Chain 1: 3000 -10705.072 0.347 0.261
Chain 1: 3100 -15465.129 0.351 0.308
Chain 1: 3200 -10010.859 0.388 0.389
Chain 1: 3300 -9699.057 0.367 0.389
Chain 1: 3400 -9110.956 0.366 0.389
Chain 1: 3500 -9030.492 0.328 0.308
Chain 1: 3600 -9841.130 0.269 0.130
Chain 1: 3700 -17267.523 0.262 0.130
Chain 1: 3800 -9017.478 0.252 0.130
Chain 1: 3900 -10139.586 0.263 0.130
Chain 1: 4000 -9406.533 0.257 0.111
Chain 1: 4100 -8748.767 0.234 0.082
Chain 1: 4200 -14030.056 0.217 0.082
Chain 1: 4300 -9594.096 0.260 0.111
Chain 1: 4400 -9295.561 0.257 0.111
Chain 1: 4500 -8733.596 0.263 0.111
Chain 1: 4600 -9399.488 0.261 0.111
Chain 1: 4700 -14046.657 0.252 0.111
Chain 1: 4800 -8374.012 0.228 0.111
Chain 1: 4900 -8666.908 0.220 0.078
Chain 1: 5000 -14384.959 0.252 0.331
Chain 1: 5100 -8398.483 0.316 0.376
Chain 1: 5200 -8638.921 0.281 0.331
Chain 1: 5300 -12682.527 0.267 0.319
Chain 1: 5400 -8155.472 0.319 0.331
Chain 1: 5500 -8338.894 0.315 0.331
Chain 1: 5600 -10275.165 0.326 0.331
Chain 1: 5700 -13658.874 0.318 0.319
Chain 1: 5800 -11696.395 0.267 0.248
Chain 1: 5900 -9442.899 0.288 0.248
Chain 1: 6000 -8827.468 0.255 0.239
Chain 1: 6100 -8927.050 0.185 0.188
Chain 1: 6200 -9003.677 0.183 0.188
Chain 1: 6300 -8500.388 0.157 0.168
Chain 1: 6400 -9048.336 0.107 0.070
Chain 1: 6500 -11341.448 0.125 0.168
Chain 1: 6600 -12006.704 0.112 0.070
Chain 1: 6700 -8238.956 0.133 0.070
Chain 1: 6800 -8967.259 0.124 0.070
Chain 1: 6900 -9243.107 0.104 0.061
Chain 1: 7000 -8042.443 0.111 0.061
Chain 1: 7100 -12165.961 0.144 0.081
Chain 1: 7200 -8113.430 0.193 0.149
Chain 1: 7300 -10942.517 0.213 0.202
Chain 1: 7400 -11763.125 0.214 0.202
Chain 1: 7500 -11581.136 0.196 0.149
Chain 1: 7600 -8071.482 0.233 0.259
Chain 1: 7700 -8150.910 0.189 0.149
Chain 1: 7800 -13256.790 0.219 0.259
Chain 1: 7900 -8270.891 0.276 0.339
Chain 1: 8000 -8695.544 0.266 0.339
Chain 1: 8100 -10412.584 0.249 0.259
Chain 1: 8200 -9438.803 0.209 0.165
Chain 1: 8300 -8171.384 0.199 0.155
Chain 1: 8400 -11849.258 0.223 0.165
Chain 1: 8500 -8810.118 0.256 0.310
Chain 1: 8600 -8398.173 0.217 0.165
Chain 1: 8700 -11100.249 0.241 0.243
Chain 1: 8800 -8728.842 0.229 0.243
Chain 1: 8900 -12003.454 0.196 0.243
Chain 1: 9000 -9824.770 0.214 0.243
Chain 1: 9100 -9103.545 0.205 0.243
Chain 1: 9200 -8317.799 0.204 0.243
Chain 1: 9300 -8441.153 0.190 0.243
Chain 1: 9400 -8207.047 0.162 0.222
Chain 1: 9500 -8229.870 0.128 0.094
Chain 1: 9600 -8066.273 0.125 0.094
Chain 1: 9700 -10605.394 0.125 0.094
Chain 1: 9800 -9109.768 0.114 0.094
Chain 1: 9900 -8840.574 0.090 0.079
Chain 1: 10000 -8206.053 0.075 0.077
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001518 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.18 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57814.594 1.000 1.000
Chain 1: 200 -17437.600 1.658 2.316
Chain 1: 300 -8556.337 1.451 1.038
Chain 1: 400 -8151.287 1.101 1.038
Chain 1: 500 -8024.201 0.884 1.000
Chain 1: 600 -8791.660 0.751 1.000
Chain 1: 700 -8099.906 0.656 0.087
Chain 1: 800 -8129.136 0.574 0.087
Chain 1: 900 -7752.219 0.516 0.085
Chain 1: 1000 -7722.681 0.465 0.085
Chain 1: 1100 -7645.676 0.366 0.050
Chain 1: 1200 -7693.908 0.135 0.049
Chain 1: 1300 -7516.688 0.033 0.024
Chain 1: 1400 -7607.361 0.030 0.016
Chain 1: 1500 -7551.250 0.029 0.012
Chain 1: 1600 -7715.529 0.022 0.012
Chain 1: 1700 -7454.002 0.017 0.012
Chain 1: 1800 -7522.976 0.018 0.012
Chain 1: 1900 -7529.198 0.013 0.010
Chain 1: 2000 -7549.859 0.013 0.010
Chain 1: 2100 -7542.074 0.012 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.007321 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 73.21 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86199.903 1.000 1.000
Chain 1: 200 -13288.435 3.243 5.487
Chain 1: 300 -9672.472 2.287 1.000
Chain 1: 400 -10419.686 1.733 1.000
Chain 1: 500 -8645.612 1.428 0.374
Chain 1: 600 -8161.201 1.199 0.374
Chain 1: 700 -8375.368 1.032 0.205
Chain 1: 800 -8949.834 0.911 0.205
Chain 1: 900 -8501.053 0.815 0.072
Chain 1: 1000 -8317.443 0.736 0.072
Chain 1: 1100 -8467.220 0.638 0.064 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8201.542 0.092 0.059
Chain 1: 1300 -8347.582 0.057 0.053
Chain 1: 1400 -8360.189 0.050 0.032
Chain 1: 1500 -8252.471 0.031 0.026
Chain 1: 1600 -8359.854 0.026 0.022
Chain 1: 1700 -8445.460 0.024 0.018
Chain 1: 1800 -8039.786 0.023 0.018
Chain 1: 1900 -8136.404 0.019 0.017
Chain 1: 2000 -8108.487 0.017 0.013
Chain 1: 2100 -8229.329 0.017 0.013
Chain 1: 2200 -8046.965 0.016 0.013
Chain 1: 2300 -8175.859 0.016 0.013
Chain 1: 2400 -8186.016 0.016 0.013
Chain 1: 2500 -8148.034 0.015 0.013
Chain 1: 2600 -8146.891 0.014 0.012
Chain 1: 2700 -8061.829 0.014 0.012
Chain 1: 2800 -8026.656 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003775 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 37.75 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8402794.916 1.000 1.000
Chain 1: 200 -1584129.087 2.652 4.304
Chain 1: 300 -890576.516 2.028 1.000
Chain 1: 400 -457837.285 1.757 1.000
Chain 1: 500 -358090.616 1.461 0.945
Chain 1: 600 -233071.589 1.307 0.945
Chain 1: 700 -119169.203 1.257 0.945
Chain 1: 800 -86314.802 1.147 0.945
Chain 1: 900 -66633.653 1.053 0.779
Chain 1: 1000 -51404.801 0.977 0.779
Chain 1: 1100 -38862.387 0.909 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38035.487 0.481 0.381
Chain 1: 1300 -25979.537 0.450 0.381
Chain 1: 1400 -25696.518 0.356 0.323
Chain 1: 1500 -22280.333 0.344 0.323
Chain 1: 1600 -21495.535 0.294 0.296
Chain 1: 1700 -20368.153 0.204 0.295
Chain 1: 1800 -20311.847 0.166 0.153
Chain 1: 1900 -20637.895 0.138 0.055
Chain 1: 2000 -19148.680 0.116 0.055
Chain 1: 2100 -19387.083 0.085 0.037
Chain 1: 2200 -19613.505 0.084 0.037
Chain 1: 2300 -19230.762 0.040 0.020
Chain 1: 2400 -19002.893 0.040 0.020
Chain 1: 2500 -18804.910 0.025 0.016
Chain 1: 2600 -18435.242 0.024 0.016
Chain 1: 2700 -18392.253 0.018 0.012
Chain 1: 2800 -18109.152 0.020 0.016
Chain 1: 2900 -18390.392 0.020 0.015
Chain 1: 3000 -18376.593 0.012 0.012
Chain 1: 3100 -18461.534 0.011 0.012
Chain 1: 3200 -18152.323 0.012 0.015
Chain 1: 3300 -18356.973 0.011 0.012
Chain 1: 3400 -17832.050 0.013 0.015
Chain 1: 3500 -18443.671 0.015 0.016
Chain 1: 3600 -17750.730 0.017 0.016
Chain 1: 3700 -18137.265 0.019 0.017
Chain 1: 3800 -17097.488 0.023 0.021
Chain 1: 3900 -17093.649 0.022 0.021
Chain 1: 4000 -17210.959 0.022 0.021
Chain 1: 4100 -17124.726 0.022 0.021
Chain 1: 4200 -16941.088 0.022 0.021
Chain 1: 4300 -17079.404 0.021 0.021
Chain 1: 4400 -17036.351 0.019 0.011
Chain 1: 4500 -16938.891 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002315 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 23.15 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13069.512 1.000 1.000
Chain 1: 200 -9791.011 0.667 1.000
Chain 1: 300 -8297.521 0.505 0.335
Chain 1: 400 -8366.793 0.381 0.335
Chain 1: 500 -8317.474 0.306 0.180
Chain 1: 600 -8177.212 0.258 0.180
Chain 1: 700 -8030.747 0.223 0.018
Chain 1: 800 -8049.710 0.196 0.018
Chain 1: 900 -8142.982 0.175 0.017
Chain 1: 1000 -8184.264 0.158 0.017
Chain 1: 1100 -8184.590 0.058 0.011
Chain 1: 1200 -8096.394 0.026 0.011
Chain 1: 1300 -8023.328 0.009 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001616 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 16.16 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -47914.764 1.000 1.000
Chain 1: 200 -16261.526 1.473 1.947
Chain 1: 300 -8778.205 1.266 1.000
Chain 1: 400 -8064.401 0.972 1.000
Chain 1: 500 -9238.238 0.803 0.852
Chain 1: 600 -8331.304 0.687 0.852
Chain 1: 700 -8572.809 0.593 0.127
Chain 1: 800 -8361.854 0.522 0.127
Chain 1: 900 -8057.816 0.468 0.109
Chain 1: 1000 -7770.090 0.425 0.109
Chain 1: 1100 -7843.764 0.326 0.089
Chain 1: 1200 -7613.352 0.134 0.038
Chain 1: 1300 -7906.303 0.053 0.037
Chain 1: 1400 -7639.960 0.048 0.037
Chain 1: 1500 -7536.367 0.036 0.035
Chain 1: 1600 -7819.149 0.029 0.035
Chain 1: 1700 -7618.150 0.029 0.035
Chain 1: 1800 -7537.658 0.027 0.035
Chain 1: 1900 -7579.534 0.024 0.030
Chain 1: 2000 -7631.801 0.021 0.026
Chain 1: 2100 -7572.663 0.021 0.026
Chain 1: 2200 -7785.094 0.021 0.026
Chain 1: 2300 -7606.427 0.019 0.023
Chain 1: 2400 -7620.074 0.016 0.014
Chain 1: 2500 -7654.103 0.015 0.011
Chain 1: 2600 -7514.677 0.013 0.011
Chain 1: 2700 -7417.912 0.012 0.011
Chain 1: 2800 -7654.030 0.014 0.013
Chain 1: 2900 -7363.981 0.017 0.019
Chain 1: 3000 -7514.891 0.019 0.020
Chain 1: 3100 -7518.357 0.018 0.020
Chain 1: 3200 -7723.884 0.018 0.020
Chain 1: 3300 -7425.587 0.020 0.020
Chain 1: 3400 -7672.931 0.023 0.027
Chain 1: 3500 -7419.741 0.026 0.031
Chain 1: 3600 -7485.804 0.025 0.031
Chain 1: 3700 -7436.721 0.024 0.031
Chain 1: 3800 -7408.571 0.021 0.027
Chain 1: 3900 -7388.229 0.018 0.020
Chain 1: 4000 -7384.182 0.016 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00395 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 39.5 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86565.381 1.000 1.000
Chain 1: 200 -14031.056 3.085 5.170
Chain 1: 300 -10239.349 2.180 1.000
Chain 1: 400 -12022.674 1.672 1.000
Chain 1: 500 -8694.706 1.414 0.383
Chain 1: 600 -8686.338 1.179 0.383
Chain 1: 700 -9389.680 1.021 0.370
Chain 1: 800 -9180.529 0.896 0.370
Chain 1: 900 -9057.093 0.798 0.148
Chain 1: 1000 -8631.515 0.723 0.148
Chain 1: 1100 -8882.261 0.626 0.075 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8520.576 0.113 0.049
Chain 1: 1300 -8852.583 0.080 0.042
Chain 1: 1400 -8789.749 0.066 0.038
Chain 1: 1500 -8726.097 0.028 0.028
Chain 1: 1600 -8775.135 0.029 0.028
Chain 1: 1700 -8844.881 0.022 0.023
Chain 1: 1800 -8382.784 0.025 0.028
Chain 1: 1900 -8504.475 0.025 0.028
Chain 1: 2000 -8520.596 0.021 0.014
Chain 1: 2100 -8611.320 0.019 0.011
Chain 1: 2200 -8392.474 0.017 0.011
Chain 1: 2300 -8562.513 0.016 0.011
Chain 1: 2400 -8399.359 0.017 0.014
Chain 1: 2500 -8473.877 0.017 0.014
Chain 1: 2600 -8384.532 0.017 0.014
Chain 1: 2700 -8418.494 0.017 0.014
Chain 1: 2800 -8369.571 0.012 0.011
Chain 1: 2900 -8484.257 0.012 0.011
Chain 1: 3000 -8398.028 0.013 0.011
Chain 1: 3100 -8361.802 0.012 0.011
Chain 1: 3200 -8333.765 0.010 0.010
Chain 1: 3300 -8593.562 0.011 0.010
Chain 1: 3400 -8634.679 0.010 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003594 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.94 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8383239.090 1.000 1.000
Chain 1: 200 -1582100.774 2.649 4.299
Chain 1: 300 -891682.745 2.024 1.000
Chain 1: 400 -458816.068 1.754 1.000
Chain 1: 500 -359615.459 1.458 0.943
Chain 1: 600 -234410.151 1.304 0.943
Chain 1: 700 -120229.208 1.254 0.943
Chain 1: 800 -87333.258 1.144 0.943
Chain 1: 900 -67596.590 1.049 0.774
Chain 1: 1000 -52342.007 0.974 0.774
Chain 1: 1100 -39761.333 0.905 0.534 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38938.013 0.478 0.377
Chain 1: 1300 -26822.811 0.445 0.377
Chain 1: 1400 -26538.905 0.352 0.316
Chain 1: 1500 -23106.993 0.339 0.316
Chain 1: 1600 -22318.806 0.289 0.292
Chain 1: 1700 -21183.261 0.200 0.291
Chain 1: 1800 -21125.773 0.162 0.149
Chain 1: 1900 -21452.742 0.135 0.054
Chain 1: 2000 -19957.391 0.113 0.054
Chain 1: 2100 -20196.201 0.083 0.035
Chain 1: 2200 -20424.001 0.082 0.035
Chain 1: 2300 -20039.784 0.038 0.019
Chain 1: 2400 -19811.475 0.038 0.019
Chain 1: 2500 -19613.743 0.025 0.015
Chain 1: 2600 -19242.865 0.023 0.015
Chain 1: 2700 -19199.484 0.018 0.012
Chain 1: 2800 -18916.091 0.019 0.015
Chain 1: 2900 -19197.837 0.019 0.015
Chain 1: 3000 -19183.880 0.012 0.012
Chain 1: 3100 -19269.013 0.011 0.012
Chain 1: 3200 -18959.093 0.011 0.015
Chain 1: 3300 -19164.275 0.010 0.012
Chain 1: 3400 -18638.211 0.012 0.015
Chain 1: 3500 -19251.701 0.014 0.015
Chain 1: 3600 -18556.296 0.016 0.015
Chain 1: 3700 -18944.721 0.018 0.016
Chain 1: 3800 -17901.231 0.022 0.021
Chain 1: 3900 -17897.321 0.021 0.021
Chain 1: 4000 -18014.596 0.021 0.021
Chain 1: 4100 -17928.233 0.021 0.021
Chain 1: 4200 -17743.743 0.021 0.021
Chain 1: 4300 -17882.624 0.021 0.021
Chain 1: 4400 -17838.883 0.018 0.010
Chain 1: 4500 -17741.334 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001406 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.06 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48751.381 1.000 1.000
Chain 1: 200 -43165.946 0.565 1.000
Chain 1: 300 -13548.290 1.105 1.000
Chain 1: 400 -13657.551 0.831 1.000
Chain 1: 500 -14722.578 0.679 0.129
Chain 1: 600 -22468.542 0.623 0.345
Chain 1: 700 -18800.053 0.562 0.195
Chain 1: 800 -13605.782 0.540 0.345
Chain 1: 900 -12166.938 0.493 0.195
Chain 1: 1000 -18803.815 0.479 0.345
Chain 1: 1100 -24454.752 0.402 0.231
Chain 1: 1200 -13404.967 0.471 0.345
Chain 1: 1300 -9644.174 0.292 0.345
Chain 1: 1400 -10503.300 0.299 0.345
Chain 1: 1500 -14593.205 0.320 0.345
Chain 1: 1600 -24077.603 0.325 0.353
Chain 1: 1700 -13484.548 0.384 0.382
Chain 1: 1800 -9446.179 0.389 0.390
Chain 1: 1900 -26966.534 0.442 0.394
Chain 1: 2000 -17717.382 0.459 0.428
Chain 1: 2100 -9393.959 0.524 0.522 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2200 -9159.762 0.444 0.428
Chain 1: 2300 -9113.265 0.406 0.428
Chain 1: 2400 -9680.672 0.403 0.428
Chain 1: 2500 -15921.730 0.415 0.428
Chain 1: 2600 -9019.725 0.452 0.522 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2700 -9333.905 0.377 0.428
Chain 1: 2800 -10121.104 0.342 0.392
Chain 1: 2900 -8599.098 0.294 0.177
Chain 1: 3000 -11323.429 0.266 0.177
Chain 1: 3100 -9180.982 0.201 0.177
Chain 1: 3200 -15329.010 0.238 0.233
Chain 1: 3300 -8836.479 0.311 0.241
Chain 1: 3400 -12938.560 0.337 0.317
Chain 1: 3500 -8882.015 0.344 0.317
Chain 1: 3600 -12813.704 0.298 0.307
Chain 1: 3700 -8773.094 0.341 0.317
Chain 1: 3800 -9025.472 0.336 0.317
Chain 1: 3900 -8721.414 0.321 0.317
Chain 1: 4000 -8527.949 0.300 0.317
Chain 1: 4100 -8861.984 0.280 0.317
Chain 1: 4200 -11816.617 0.265 0.307
Chain 1: 4300 -8555.423 0.230 0.307
Chain 1: 4400 -8368.769 0.200 0.250
Chain 1: 4500 -8808.925 0.159 0.050
Chain 1: 4600 -10668.577 0.146 0.050
Chain 1: 4700 -11121.058 0.104 0.041
Chain 1: 4800 -8666.751 0.130 0.050
Chain 1: 4900 -10409.771 0.143 0.167
Chain 1: 5000 -13146.341 0.161 0.174
Chain 1: 5100 -8505.804 0.212 0.208
Chain 1: 5200 -8674.315 0.189 0.174
Chain 1: 5300 -9247.009 0.157 0.167
Chain 1: 5400 -8390.509 0.165 0.167
Chain 1: 5500 -8419.910 0.161 0.167
Chain 1: 5600 -13933.899 0.183 0.167
Chain 1: 5700 -15938.682 0.191 0.167
Chain 1: 5800 -8469.774 0.251 0.167
Chain 1: 5900 -8326.615 0.236 0.126
Chain 1: 6000 -8774.847 0.220 0.102
Chain 1: 6100 -12601.391 0.196 0.102
Chain 1: 6200 -8266.494 0.247 0.126
Chain 1: 6300 -8128.969 0.242 0.126
Chain 1: 6400 -9397.858 0.246 0.135
Chain 1: 6500 -9951.148 0.251 0.135
Chain 1: 6600 -11639.508 0.226 0.135
Chain 1: 6700 -12099.680 0.217 0.135
Chain 1: 6800 -12503.605 0.132 0.056
Chain 1: 6900 -8286.605 0.181 0.135
Chain 1: 7000 -8208.931 0.177 0.135
Chain 1: 7100 -8080.253 0.148 0.056
Chain 1: 7200 -8966.704 0.106 0.056
Chain 1: 7300 -10394.674 0.118 0.099
Chain 1: 7400 -8043.474 0.133 0.099
Chain 1: 7500 -11262.311 0.156 0.137
Chain 1: 7600 -8544.018 0.174 0.137
Chain 1: 7700 -8888.607 0.174 0.137
Chain 1: 7800 -8890.223 0.171 0.137
Chain 1: 7900 -8156.859 0.129 0.099
Chain 1: 8000 -9000.205 0.137 0.099
Chain 1: 8100 -11797.734 0.159 0.137
Chain 1: 8200 -9800.094 0.170 0.204
Chain 1: 8300 -8488.399 0.171 0.204
Chain 1: 8400 -11313.826 0.167 0.204
Chain 1: 8500 -9001.092 0.164 0.204
Chain 1: 8600 -8509.412 0.138 0.155
Chain 1: 8700 -8603.908 0.135 0.155
Chain 1: 8800 -8019.142 0.143 0.155
Chain 1: 8900 -8750.545 0.142 0.155
Chain 1: 9000 -9444.885 0.140 0.155
Chain 1: 9100 -10396.932 0.126 0.092
Chain 1: 9200 -9150.901 0.119 0.092
Chain 1: 9300 -8165.028 0.115 0.092
Chain 1: 9400 -11632.022 0.120 0.092
Chain 1: 9500 -8056.126 0.139 0.092
Chain 1: 9600 -10153.521 0.154 0.121
Chain 1: 9700 -8294.541 0.175 0.136
Chain 1: 9800 -9175.377 0.177 0.136
Chain 1: 9900 -10844.744 0.184 0.154
Chain 1: 10000 -7977.418 0.213 0.207
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001432 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.32 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61366.636 1.000 1.000
Chain 1: 200 -17508.881 1.752 2.505
Chain 1: 300 -8682.101 1.507 1.017
Chain 1: 400 -8234.855 1.144 1.017
Chain 1: 500 -8285.868 0.916 1.000
Chain 1: 600 -8302.789 0.764 1.000
Chain 1: 700 -7715.315 0.666 0.076
Chain 1: 800 -7692.461 0.583 0.076
Chain 1: 900 -7587.045 0.520 0.054
Chain 1: 1000 -7727.959 0.470 0.054
Chain 1: 1100 -7636.083 0.371 0.018
Chain 1: 1200 -7514.908 0.122 0.016
Chain 1: 1300 -7619.194 0.022 0.014
Chain 1: 1400 -7565.448 0.017 0.014
Chain 1: 1500 -7555.319 0.016 0.014
Chain 1: 1600 -7457.783 0.017 0.014
Chain 1: 1700 -7451.660 0.010 0.013 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003173 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.73 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85889.717 1.000 1.000
Chain 1: 200 -13197.850 3.254 5.508
Chain 1: 300 -9665.034 2.291 1.000
Chain 1: 400 -10582.661 1.740 1.000
Chain 1: 500 -8605.206 1.438 0.366
Chain 1: 600 -8218.856 1.206 0.366
Chain 1: 700 -8493.197 1.038 0.230
Chain 1: 800 -8792.936 0.913 0.230
Chain 1: 900 -8543.559 0.815 0.087
Chain 1: 1000 -8226.733 0.737 0.087
Chain 1: 1100 -8571.430 0.641 0.047 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8244.676 0.094 0.040
Chain 1: 1300 -8241.383 0.058 0.040
Chain 1: 1400 -8236.586 0.049 0.039
Chain 1: 1500 -8268.233 0.027 0.034
Chain 1: 1600 -8277.136 0.022 0.032
Chain 1: 1700 -8204.953 0.020 0.029
Chain 1: 1800 -8091.785 0.018 0.014
Chain 1: 1900 -8208.456 0.016 0.014
Chain 1: 2000 -8168.777 0.013 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003326 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.26 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8420048.722 1.000 1.000
Chain 1: 200 -1586416.214 2.654 4.308
Chain 1: 300 -890579.532 2.030 1.000
Chain 1: 400 -457413.130 1.759 1.000
Chain 1: 500 -357410.656 1.463 0.947
Chain 1: 600 -232361.603 1.309 0.947
Chain 1: 700 -118707.468 1.259 0.947
Chain 1: 800 -85976.760 1.149 0.947
Chain 1: 900 -66343.350 1.054 0.781
Chain 1: 1000 -51165.234 0.978 0.781
Chain 1: 1100 -38672.783 0.911 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37847.512 0.482 0.381
Chain 1: 1300 -25840.448 0.451 0.381
Chain 1: 1400 -25561.024 0.357 0.323
Chain 1: 1500 -22158.760 0.344 0.323
Chain 1: 1600 -21377.944 0.294 0.297
Chain 1: 1700 -20256.313 0.204 0.296
Chain 1: 1800 -20201.423 0.166 0.154
Chain 1: 1900 -20526.961 0.138 0.055
Chain 1: 2000 -19041.982 0.116 0.055
Chain 1: 2100 -19279.955 0.085 0.037
Chain 1: 2200 -19505.694 0.084 0.037
Chain 1: 2300 -19123.730 0.040 0.020
Chain 1: 2400 -18896.082 0.040 0.020
Chain 1: 2500 -18698.089 0.025 0.016
Chain 1: 2600 -18328.844 0.024 0.016
Chain 1: 2700 -18286.071 0.019 0.012
Chain 1: 2800 -18003.151 0.020 0.016
Chain 1: 2900 -18284.111 0.020 0.015
Chain 1: 3000 -18270.316 0.012 0.012
Chain 1: 3100 -18355.197 0.011 0.012
Chain 1: 3200 -18046.294 0.012 0.015
Chain 1: 3300 -18250.733 0.011 0.012
Chain 1: 3400 -17726.362 0.013 0.015
Chain 1: 3500 -18337.124 0.015 0.016
Chain 1: 3600 -17645.285 0.017 0.016
Chain 1: 3700 -18030.922 0.019 0.017
Chain 1: 3800 -16992.918 0.023 0.021
Chain 1: 3900 -16989.135 0.022 0.021
Chain 1: 4000 -17106.424 0.022 0.021
Chain 1: 4100 -17020.263 0.023 0.021
Chain 1: 4200 -16837.061 0.022 0.021
Chain 1: 4300 -16975.064 0.022 0.021
Chain 1: 4400 -16932.274 0.019 0.011
Chain 1: 4500 -16834.906 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001573 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.73 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48741.815 1.000 1.000
Chain 1: 200 -20197.516 1.207 1.413
Chain 1: 300 -16136.888 0.888 1.000
Chain 1: 400 -11728.498 0.760 1.000
Chain 1: 500 -12049.025 0.613 0.376
Chain 1: 600 -15348.843 0.547 0.376
Chain 1: 700 -15183.772 0.470 0.252
Chain 1: 800 -23257.509 0.455 0.347
Chain 1: 900 -10857.536 0.531 0.347
Chain 1: 1000 -10434.785 0.482 0.347
Chain 1: 1100 -10188.789 0.385 0.252
Chain 1: 1200 -19699.477 0.292 0.252
Chain 1: 1300 -10251.410 0.359 0.347
Chain 1: 1400 -10424.226 0.323 0.215
Chain 1: 1500 -10255.518 0.322 0.215
Chain 1: 1600 -13442.469 0.324 0.237
Chain 1: 1700 -10281.042 0.354 0.308
Chain 1: 1800 -20627.163 0.369 0.308
Chain 1: 1900 -9550.276 0.371 0.308
Chain 1: 2000 -18805.778 0.416 0.483
Chain 1: 2100 -17194.591 0.423 0.483
Chain 1: 2200 -13373.804 0.403 0.308
Chain 1: 2300 -11071.532 0.332 0.286
Chain 1: 2400 -9223.285 0.350 0.286
Chain 1: 2500 -10224.680 0.358 0.286
Chain 1: 2600 -10239.116 0.335 0.286
Chain 1: 2700 -9712.727 0.309 0.208
Chain 1: 2800 -10866.265 0.270 0.200
Chain 1: 2900 -9744.878 0.165 0.115
Chain 1: 3000 -15087.671 0.152 0.115
Chain 1: 3100 -9034.820 0.209 0.200
Chain 1: 3200 -8767.359 0.184 0.115
Chain 1: 3300 -9429.884 0.170 0.106
Chain 1: 3400 -13344.471 0.179 0.106
Chain 1: 3500 -8882.317 0.220 0.115
Chain 1: 3600 -8936.273 0.220 0.115
Chain 1: 3700 -10004.372 0.225 0.115
Chain 1: 3800 -8671.892 0.230 0.154
Chain 1: 3900 -13254.835 0.253 0.293
Chain 1: 4000 -9932.630 0.251 0.293
Chain 1: 4100 -8873.896 0.196 0.154
Chain 1: 4200 -9696.013 0.202 0.154
Chain 1: 4300 -10462.528 0.202 0.154
Chain 1: 4400 -13791.659 0.197 0.154
Chain 1: 4500 -8959.345 0.200 0.154
Chain 1: 4600 -8694.758 0.203 0.154
Chain 1: 4700 -14263.800 0.231 0.241
Chain 1: 4800 -9590.516 0.265 0.334
Chain 1: 4900 -9136.288 0.235 0.241
Chain 1: 5000 -11084.906 0.219 0.176
Chain 1: 5100 -8602.999 0.236 0.241
Chain 1: 5200 -14892.114 0.270 0.288
Chain 1: 5300 -11623.822 0.291 0.288
Chain 1: 5400 -13461.634 0.280 0.288
Chain 1: 5500 -9189.938 0.273 0.288
Chain 1: 5600 -8723.780 0.275 0.288
Chain 1: 5700 -13821.727 0.273 0.288
Chain 1: 5800 -9347.870 0.272 0.288
Chain 1: 5900 -13128.662 0.296 0.288
Chain 1: 6000 -11437.114 0.293 0.288
Chain 1: 6100 -8529.102 0.298 0.341
Chain 1: 6200 -8282.611 0.259 0.288
Chain 1: 6300 -11622.788 0.260 0.288
Chain 1: 6400 -8759.605 0.279 0.327
Chain 1: 6500 -11316.878 0.255 0.288
Chain 1: 6600 -8394.103 0.284 0.327
Chain 1: 6700 -8371.241 0.248 0.288
Chain 1: 6800 -12017.333 0.230 0.288
Chain 1: 6900 -8697.554 0.239 0.303
Chain 1: 7000 -9110.021 0.229 0.303
Chain 1: 7100 -9840.227 0.203 0.287
Chain 1: 7200 -11038.565 0.210 0.287
Chain 1: 7300 -11952.214 0.189 0.226
Chain 1: 7400 -12332.279 0.160 0.109
Chain 1: 7500 -9987.631 0.161 0.109
Chain 1: 7600 -8817.746 0.139 0.109
Chain 1: 7700 -8373.286 0.144 0.109
Chain 1: 7800 -10439.068 0.134 0.109
Chain 1: 7900 -8252.781 0.122 0.109
Chain 1: 8000 -11281.918 0.144 0.133
Chain 1: 8100 -8533.489 0.169 0.198
Chain 1: 8200 -9442.107 0.168 0.198
Chain 1: 8300 -8820.542 0.167 0.198
Chain 1: 8400 -13414.511 0.198 0.235
Chain 1: 8500 -8843.015 0.227 0.265
Chain 1: 8600 -9121.009 0.216 0.265
Chain 1: 8700 -8563.763 0.218 0.265
Chain 1: 8800 -8175.561 0.202 0.265
Chain 1: 8900 -8507.152 0.180 0.096
Chain 1: 9000 -11927.590 0.182 0.096
Chain 1: 9100 -11212.318 0.156 0.070
Chain 1: 9200 -8550.681 0.177 0.070
Chain 1: 9300 -11385.241 0.195 0.249
Chain 1: 9400 -9719.144 0.178 0.171
Chain 1: 9500 -8095.719 0.146 0.171
Chain 1: 9600 -10375.522 0.165 0.201
Chain 1: 9700 -13170.487 0.180 0.212
Chain 1: 9800 -9367.193 0.216 0.220
Chain 1: 9900 -8304.004 0.225 0.220
Chain 1: 10000 -9731.335 0.211 0.212
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001437 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.37 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57081.275 1.000 1.000
Chain 1: 200 -17422.410 1.638 2.276
Chain 1: 300 -8681.513 1.428 1.007
Chain 1: 400 -8256.286 1.084 1.007
Chain 1: 500 -8500.620 0.873 1.000
Chain 1: 600 -9232.735 0.740 1.000
Chain 1: 700 -8638.497 0.644 0.079
Chain 1: 800 -8086.590 0.572 0.079
Chain 1: 900 -7894.148 0.512 0.069
Chain 1: 1000 -7752.762 0.462 0.069
Chain 1: 1100 -7565.274 0.365 0.068
Chain 1: 1200 -7595.454 0.137 0.052
Chain 1: 1300 -7566.300 0.037 0.029
Chain 1: 1400 -7780.551 0.035 0.028
Chain 1: 1500 -7558.961 0.035 0.028
Chain 1: 1600 -7707.098 0.029 0.025
Chain 1: 1700 -7489.643 0.025 0.025
Chain 1: 1800 -7520.758 0.018 0.024
Chain 1: 1900 -7553.355 0.016 0.019
Chain 1: 2000 -7592.863 0.015 0.019
Chain 1: 2100 -7585.148 0.013 0.005 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003721 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 37.21 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85851.419 1.000 1.000
Chain 1: 200 -13446.286 3.192 5.385
Chain 1: 300 -9852.643 2.250 1.000
Chain 1: 400 -10702.580 1.707 1.000
Chain 1: 500 -8763.172 1.410 0.365
Chain 1: 600 -8518.645 1.180 0.365
Chain 1: 700 -8707.124 1.014 0.221
Chain 1: 800 -9161.020 0.894 0.221
Chain 1: 900 -8651.723 0.801 0.079
Chain 1: 1000 -8490.296 0.723 0.079
Chain 1: 1100 -8606.212 0.624 0.059 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8383.652 0.088 0.050
Chain 1: 1300 -8583.057 0.054 0.029
Chain 1: 1400 -8566.132 0.046 0.027
Chain 1: 1500 -8430.242 0.026 0.023
Chain 1: 1600 -8538.727 0.024 0.022
Chain 1: 1700 -8626.268 0.023 0.019
Chain 1: 1800 -8218.634 0.023 0.019
Chain 1: 1900 -8316.505 0.018 0.016
Chain 1: 2000 -8288.381 0.017 0.013
Chain 1: 2100 -8408.762 0.017 0.014
Chain 1: 2200 -8213.124 0.017 0.014
Chain 1: 2300 -8355.044 0.016 0.014
Chain 1: 2400 -8361.894 0.016 0.014
Chain 1: 2500 -8329.836 0.015 0.013
Chain 1: 2600 -8328.165 0.013 0.012
Chain 1: 2700 -8241.474 0.014 0.012
Chain 1: 2800 -8207.035 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003411 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.11 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8392413.165 1.000 1.000
Chain 1: 200 -1582784.048 2.651 4.302
Chain 1: 300 -890037.512 2.027 1.000
Chain 1: 400 -457182.002 1.757 1.000
Chain 1: 500 -357859.079 1.461 0.947
Chain 1: 600 -232826.315 1.307 0.947
Chain 1: 700 -119126.959 1.257 0.947
Chain 1: 800 -86381.176 1.147 0.947
Chain 1: 900 -66728.692 1.052 0.778
Chain 1: 1000 -51529.219 0.977 0.778
Chain 1: 1100 -39008.081 0.909 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38182.302 0.481 0.379
Chain 1: 1300 -26139.050 0.449 0.379
Chain 1: 1400 -25856.992 0.355 0.321
Chain 1: 1500 -22445.715 0.343 0.321
Chain 1: 1600 -21662.512 0.293 0.295
Chain 1: 1700 -20536.333 0.203 0.295
Chain 1: 1800 -20480.472 0.165 0.152
Chain 1: 1900 -20806.491 0.137 0.055
Chain 1: 2000 -19318.086 0.115 0.055
Chain 1: 2100 -19556.280 0.084 0.036
Chain 1: 2200 -19782.841 0.083 0.036
Chain 1: 2300 -19400.027 0.039 0.020
Chain 1: 2400 -19172.179 0.039 0.020
Chain 1: 2500 -18974.324 0.025 0.016
Chain 1: 2600 -18604.616 0.024 0.016
Chain 1: 2700 -18561.526 0.018 0.012
Chain 1: 2800 -18278.591 0.020 0.015
Chain 1: 2900 -18559.735 0.020 0.015
Chain 1: 3000 -18545.837 0.012 0.012
Chain 1: 3100 -18630.874 0.011 0.012
Chain 1: 3200 -18321.622 0.012 0.015
Chain 1: 3300 -18526.267 0.011 0.012
Chain 1: 3400 -18001.425 0.013 0.015
Chain 1: 3500 -18613.020 0.015 0.015
Chain 1: 3600 -17920.002 0.017 0.015
Chain 1: 3700 -18306.650 0.019 0.017
Chain 1: 3800 -17266.910 0.023 0.021
Chain 1: 3900 -17263.084 0.022 0.021
Chain 1: 4000 -17380.362 0.022 0.021
Chain 1: 4100 -17294.228 0.022 0.021
Chain 1: 4200 -17110.535 0.022 0.021
Chain 1: 4300 -17248.863 0.021 0.021
Chain 1: 4400 -17205.767 0.019 0.011
Chain 1: 4500 -17108.340 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001319 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.19 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13115.407 1.000 1.000
Chain 1: 200 -9920.156 0.661 1.000
Chain 1: 300 -8600.320 0.492 0.322
Chain 1: 400 -8783.991 0.374 0.322
Chain 1: 500 -8628.405 0.303 0.153
Chain 1: 600 -8445.463 0.256 0.153
Chain 1: 700 -8404.495 0.220 0.022
Chain 1: 800 -8379.941 0.193 0.022
Chain 1: 900 -8404.632 0.172 0.021
Chain 1: 1000 -8473.958 0.156 0.021
Chain 1: 1100 -8516.286 0.056 0.018
Chain 1: 1200 -8457.319 0.024 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001487 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.87 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -47311.921 1.000 1.000
Chain 1: 200 -16315.500 1.450 1.900
Chain 1: 300 -9094.382 1.231 1.000
Chain 1: 400 -8255.851 0.949 1.000
Chain 1: 500 -8087.966 0.763 0.794
Chain 1: 600 -8739.163 0.648 0.794
Chain 1: 700 -7805.571 0.573 0.120
Chain 1: 800 -8444.996 0.511 0.120
Chain 1: 900 -7915.086 0.461 0.102
Chain 1: 1000 -7927.761 0.415 0.102
Chain 1: 1100 -7589.184 0.320 0.076
Chain 1: 1200 -8002.361 0.135 0.075
Chain 1: 1300 -7889.688 0.057 0.067
Chain 1: 1400 -8100.895 0.050 0.052
Chain 1: 1500 -7621.449 0.054 0.063
Chain 1: 1600 -7696.591 0.047 0.052
Chain 1: 1700 -7830.731 0.037 0.045
Chain 1: 1800 -7716.522 0.031 0.026
Chain 1: 1900 -7694.710 0.025 0.017
Chain 1: 2000 -7733.924 0.025 0.017
Chain 1: 2100 -7618.362 0.022 0.015
Chain 1: 2200 -7819.693 0.019 0.015
Chain 1: 2300 -7682.468 0.020 0.017
Chain 1: 2400 -7627.868 0.018 0.015
Chain 1: 2500 -7735.630 0.013 0.015
Chain 1: 2600 -7586.568 0.014 0.015
Chain 1: 2700 -7611.258 0.013 0.015
Chain 1: 2800 -7552.735 0.012 0.014
Chain 1: 2900 -7496.862 0.012 0.014
Chain 1: 3000 -7577.714 0.013 0.014
Chain 1: 3100 -7592.019 0.012 0.011
Chain 1: 3200 -7818.493 0.012 0.011
Chain 1: 3300 -7501.656 0.014 0.011
Chain 1: 3400 -7732.139 0.017 0.014
Chain 1: 3500 -7481.860 0.019 0.020
Chain 1: 3600 -7559.217 0.018 0.011
Chain 1: 3700 -7514.117 0.018 0.011
Chain 1: 3800 -7519.688 0.017 0.011
Chain 1: 3900 -7462.690 0.017 0.011
Chain 1: 4000 -7459.682 0.016 0.010
Chain 1: 4100 -7465.664 0.016 0.010
Chain 1: 4200 -7548.675 0.014 0.010
Chain 1: 4300 -7447.483 0.011 0.010
Chain 1: 4400 -7489.412 0.009 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003496 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.96 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87145.144 1.000 1.000
Chain 1: 200 -14213.351 3.066 5.131
Chain 1: 300 -10504.195 2.161 1.000
Chain 1: 400 -11696.102 1.647 1.000
Chain 1: 500 -9506.137 1.363 0.353
Chain 1: 600 -8921.011 1.147 0.353
Chain 1: 700 -8921.698 0.983 0.230
Chain 1: 800 -9393.440 0.867 0.230
Chain 1: 900 -9252.202 0.772 0.102
Chain 1: 1000 -9261.319 0.695 0.102
Chain 1: 1100 -9276.918 0.595 0.066 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8872.741 0.086 0.050
Chain 1: 1300 -9165.099 0.054 0.046
Chain 1: 1400 -9110.452 0.045 0.032
Chain 1: 1500 -9004.651 0.023 0.015
Chain 1: 1600 -9119.316 0.018 0.013
Chain 1: 1700 -9183.320 0.018 0.013
Chain 1: 1800 -8746.444 0.018 0.013
Chain 1: 1900 -8850.754 0.018 0.012
Chain 1: 2000 -8827.736 0.018 0.012
Chain 1: 2100 -8968.950 0.019 0.013
Chain 1: 2200 -8756.719 0.017 0.013
Chain 1: 2300 -8915.943 0.016 0.013
Chain 1: 2400 -8754.113 0.017 0.016
Chain 1: 2500 -8825.097 0.017 0.016
Chain 1: 2600 -8737.124 0.017 0.016
Chain 1: 2700 -8770.592 0.016 0.016
Chain 1: 2800 -8730.249 0.012 0.012
Chain 1: 2900 -8824.262 0.012 0.011
Chain 1: 3000 -8659.507 0.013 0.016
Chain 1: 3100 -8813.004 0.013 0.017
Chain 1: 3200 -8684.737 0.012 0.015
Chain 1: 3300 -8693.977 0.011 0.011
Chain 1: 3400 -8857.137 0.011 0.011
Chain 1: 3500 -8869.120 0.010 0.011
Chain 1: 3600 -8641.085 0.012 0.015
Chain 1: 3700 -8788.016 0.013 0.017
Chain 1: 3800 -8647.319 0.014 0.017
Chain 1: 3900 -8581.538 0.014 0.017
Chain 1: 4000 -8659.323 0.013 0.016
Chain 1: 4100 -8652.906 0.011 0.015
Chain 1: 4200 -8637.405 0.010 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003387 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.87 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8415535.998 1.000 1.000
Chain 1: 200 -1590363.670 2.646 4.292
Chain 1: 300 -892522.327 2.024 1.000
Chain 1: 400 -458603.822 1.755 1.000
Chain 1: 500 -358436.633 1.460 0.946
Chain 1: 600 -233279.810 1.306 0.946
Chain 1: 700 -119686.436 1.255 0.946
Chain 1: 800 -86958.506 1.145 0.946
Chain 1: 900 -67353.698 1.050 0.782
Chain 1: 1000 -52205.326 0.974 0.782
Chain 1: 1100 -39727.158 0.906 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38912.344 0.479 0.376
Chain 1: 1300 -26907.850 0.445 0.376
Chain 1: 1400 -26632.922 0.351 0.314
Chain 1: 1500 -23229.585 0.338 0.314
Chain 1: 1600 -22449.360 0.288 0.291
Chain 1: 1700 -21327.356 0.198 0.290
Chain 1: 1800 -21272.794 0.161 0.147
Chain 1: 1900 -21599.344 0.133 0.053
Chain 1: 2000 -20111.641 0.112 0.053
Chain 1: 2100 -20350.203 0.081 0.035
Chain 1: 2200 -20576.535 0.080 0.035
Chain 1: 2300 -20193.691 0.038 0.019
Chain 1: 2400 -19965.626 0.038 0.019
Chain 1: 2500 -19767.426 0.024 0.015
Chain 1: 2600 -19397.519 0.023 0.015
Chain 1: 2700 -19354.359 0.018 0.012
Chain 1: 2800 -19070.952 0.019 0.015
Chain 1: 2900 -19352.292 0.019 0.015
Chain 1: 3000 -19338.576 0.011 0.012
Chain 1: 3100 -19423.644 0.011 0.011
Chain 1: 3200 -19114.087 0.011 0.015
Chain 1: 3300 -19318.948 0.010 0.011
Chain 1: 3400 -18793.400 0.012 0.015
Chain 1: 3500 -19405.949 0.014 0.015
Chain 1: 3600 -18711.655 0.016 0.015
Chain 1: 3700 -19099.149 0.018 0.016
Chain 1: 3800 -18057.382 0.022 0.020
Chain 1: 3900 -18053.413 0.021 0.020
Chain 1: 4000 -18170.783 0.021 0.020
Chain 1: 4100 -18084.483 0.021 0.020
Chain 1: 4200 -17900.347 0.021 0.020
Chain 1: 4300 -18039.051 0.020 0.020
Chain 1: 4400 -17995.597 0.018 0.010
Chain 1: 4500 -17898.016 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001248 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.48 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13421.074 1.000 1.000
Chain 1: 200 -9885.323 0.679 1.000
Chain 1: 300 -8562.258 0.504 0.358
Chain 1: 400 -8335.934 0.385 0.358
Chain 1: 500 -8117.582 0.313 0.155
Chain 1: 600 -8131.932 0.261 0.155
Chain 1: 700 -8001.667 0.226 0.027
Chain 1: 800 -8004.174 0.198 0.027
Chain 1: 900 -8086.923 0.177 0.027
Chain 1: 1000 -8058.123 0.160 0.027
Chain 1: 1100 -8037.292 0.060 0.016
Chain 1: 1200 -8041.173 0.024 0.010
Chain 1: 1300 -7965.084 0.010 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00139 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.9 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -53696.223 1.000 1.000
Chain 1: 200 -17259.358 1.556 2.111
Chain 1: 300 -8802.266 1.357 1.000
Chain 1: 400 -9142.380 1.027 1.000
Chain 1: 500 -8211.533 0.844 0.961
Chain 1: 600 -8536.635 0.710 0.961
Chain 1: 700 -7800.393 0.622 0.113
Chain 1: 800 -8333.095 0.552 0.113
Chain 1: 900 -7682.056 0.500 0.094
Chain 1: 1000 -7960.732 0.454 0.094
Chain 1: 1100 -7605.958 0.359 0.085
Chain 1: 1200 -7826.467 0.150 0.064
Chain 1: 1300 -7606.151 0.057 0.047
Chain 1: 1400 -7816.645 0.056 0.047
Chain 1: 1500 -7483.371 0.049 0.045
Chain 1: 1600 -7622.107 0.047 0.045
Chain 1: 1700 -7478.918 0.040 0.035
Chain 1: 1800 -7552.812 0.034 0.029
Chain 1: 1900 -7587.856 0.026 0.028
Chain 1: 2000 -7598.606 0.023 0.027
Chain 1: 2100 -7541.064 0.019 0.019
Chain 1: 2200 -7654.601 0.018 0.018
Chain 1: 2300 -7471.969 0.017 0.018
Chain 1: 2400 -7455.683 0.015 0.015
Chain 1: 2500 -7562.799 0.012 0.014
Chain 1: 2600 -7447.937 0.011 0.014
Chain 1: 2700 -7359.363 0.011 0.012
Chain 1: 2800 -7543.020 0.012 0.014
Chain 1: 2900 -7289.560 0.015 0.015
Chain 1: 3000 -7454.879 0.017 0.015
Chain 1: 3100 -7434.079 0.017 0.015
Chain 1: 3200 -7658.236 0.018 0.022
Chain 1: 3300 -7359.107 0.020 0.022
Chain 1: 3400 -7612.625 0.023 0.024
Chain 1: 3500 -7355.199 0.025 0.029
Chain 1: 3600 -7413.295 0.024 0.029
Chain 1: 3700 -7370.091 0.024 0.029
Chain 1: 3800 -7379.460 0.021 0.029
Chain 1: 3900 -7329.869 0.018 0.022
Chain 1: 4000 -7314.468 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003223 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.23 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86818.358 1.000 1.000
Chain 1: 200 -13907.675 3.121 5.242
Chain 1: 300 -10146.999 2.204 1.000
Chain 1: 400 -11754.237 1.687 1.000
Chain 1: 500 -8777.951 1.418 0.371
Chain 1: 600 -8606.496 1.185 0.371
Chain 1: 700 -8867.605 1.020 0.339
Chain 1: 800 -9188.379 0.897 0.339
Chain 1: 900 -8866.495 0.801 0.137
Chain 1: 1000 -9067.016 0.723 0.137
Chain 1: 1100 -8767.541 0.627 0.036 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8446.481 0.106 0.036
Chain 1: 1300 -8816.009 0.073 0.036
Chain 1: 1400 -8717.821 0.061 0.035
Chain 1: 1500 -8623.720 0.028 0.034
Chain 1: 1600 -8736.604 0.027 0.034
Chain 1: 1700 -8797.258 0.025 0.034
Chain 1: 1800 -8345.640 0.027 0.034
Chain 1: 1900 -8456.039 0.025 0.022
Chain 1: 2000 -8443.155 0.022 0.013
Chain 1: 2100 -8573.208 0.021 0.013
Chain 1: 2200 -8351.341 0.019 0.013
Chain 1: 2300 -8513.450 0.017 0.013
Chain 1: 2400 -8351.594 0.018 0.015
Chain 1: 2500 -8427.971 0.018 0.015
Chain 1: 2600 -8344.755 0.017 0.015
Chain 1: 2700 -8373.320 0.017 0.015
Chain 1: 2800 -8327.219 0.012 0.013
Chain 1: 2900 -8435.769 0.012 0.013
Chain 1: 3000 -8386.145 0.013 0.013
Chain 1: 3100 -8318.384 0.012 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003114 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.14 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8428341.447 1.000 1.000
Chain 1: 200 -1589467.669 2.651 4.303
Chain 1: 300 -892215.325 2.028 1.000
Chain 1: 400 -458375.829 1.758 1.000
Chain 1: 500 -358171.803 1.462 0.946
Chain 1: 600 -233156.773 1.308 0.946
Chain 1: 700 -119514.514 1.257 0.946
Chain 1: 800 -86723.207 1.147 0.946
Chain 1: 900 -67105.977 1.052 0.781
Chain 1: 1000 -51944.150 0.976 0.781
Chain 1: 1100 -39449.512 0.908 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38636.949 0.479 0.378
Chain 1: 1300 -26619.438 0.446 0.378
Chain 1: 1400 -26343.222 0.353 0.317
Chain 1: 1500 -22936.309 0.340 0.317
Chain 1: 1600 -22155.195 0.290 0.292
Chain 1: 1700 -21031.935 0.200 0.292
Chain 1: 1800 -20977.158 0.162 0.149
Chain 1: 1900 -21303.886 0.135 0.053
Chain 1: 2000 -19815.409 0.113 0.053
Chain 1: 2100 -20053.868 0.083 0.035
Chain 1: 2200 -20280.314 0.082 0.035
Chain 1: 2300 -19897.404 0.038 0.019
Chain 1: 2400 -19669.330 0.038 0.019
Chain 1: 2500 -19470.997 0.025 0.015
Chain 1: 2600 -19100.772 0.023 0.015
Chain 1: 2700 -19057.721 0.018 0.012
Chain 1: 2800 -18774.069 0.019 0.015
Chain 1: 2900 -19055.669 0.019 0.015
Chain 1: 3000 -19041.877 0.012 0.012
Chain 1: 3100 -19126.876 0.011 0.012
Chain 1: 3200 -18817.204 0.011 0.015
Chain 1: 3300 -19022.262 0.011 0.012
Chain 1: 3400 -18496.343 0.012 0.015
Chain 1: 3500 -19109.288 0.014 0.015
Chain 1: 3600 -18414.694 0.016 0.015
Chain 1: 3700 -18802.348 0.018 0.016
Chain 1: 3800 -17759.859 0.022 0.021
Chain 1: 3900 -17755.923 0.021 0.021
Chain 1: 4000 -17873.295 0.022 0.021
Chain 1: 4100 -17786.829 0.022 0.021
Chain 1: 4200 -17602.679 0.021 0.021
Chain 1: 4300 -17741.401 0.021 0.021
Chain 1: 4400 -17697.854 0.018 0.010
Chain 1: 4500 -17600.295 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001394 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.94 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48822.077 1.000 1.000
Chain 1: 200 -18473.809 1.321 1.643
Chain 1: 300 -38904.983 1.056 1.000
Chain 1: 400 -34402.197 0.825 1.000
Chain 1: 500 -17402.535 0.855 0.977
Chain 1: 600 -12524.138 0.778 0.977
Chain 1: 700 -16372.009 0.700 0.525
Chain 1: 800 -14393.326 0.630 0.525
Chain 1: 900 -11236.593 0.591 0.390
Chain 1: 1000 -14921.313 0.557 0.390
Chain 1: 1100 -29131.180 0.505 0.390 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -13090.838 0.464 0.390
Chain 1: 1300 -12026.348 0.420 0.281
Chain 1: 1400 -23575.629 0.456 0.390
Chain 1: 1500 -10846.901 0.475 0.390
Chain 1: 1600 -12626.223 0.451 0.281
Chain 1: 1700 -12508.114 0.428 0.281
Chain 1: 1800 -10905.015 0.429 0.281
Chain 1: 1900 -10544.894 0.404 0.247
Chain 1: 2000 -15531.458 0.412 0.321
Chain 1: 2100 -9774.439 0.422 0.321
Chain 1: 2200 -9768.160 0.299 0.147
Chain 1: 2300 -10121.575 0.294 0.147
Chain 1: 2400 -9307.043 0.254 0.141
Chain 1: 2500 -9485.735 0.138 0.088
Chain 1: 2600 -10052.265 0.130 0.056
Chain 1: 2700 -9139.321 0.139 0.088
Chain 1: 2800 -10241.945 0.135 0.088
Chain 1: 2900 -9999.455 0.134 0.088
Chain 1: 3000 -10037.235 0.102 0.056
Chain 1: 3100 -9818.520 0.046 0.035
Chain 1: 3200 -9075.012 0.054 0.056
Chain 1: 3300 -9555.549 0.055 0.056
Chain 1: 3400 -13147.377 0.074 0.056
Chain 1: 3500 -9071.882 0.117 0.082
Chain 1: 3600 -9893.822 0.120 0.083
Chain 1: 3700 -9307.856 0.116 0.082
Chain 1: 3800 -9363.830 0.106 0.063
Chain 1: 3900 -9201.873 0.105 0.063
Chain 1: 4000 -9495.481 0.108 0.063
Chain 1: 4100 -10429.950 0.114 0.082
Chain 1: 4200 -10073.031 0.110 0.063
Chain 1: 4300 -9333.987 0.113 0.079
Chain 1: 4400 -8598.797 0.094 0.079
Chain 1: 4500 -8940.508 0.053 0.063
Chain 1: 4600 -10239.926 0.057 0.063
Chain 1: 4700 -8839.842 0.067 0.079
Chain 1: 4800 -9671.114 0.075 0.085
Chain 1: 4900 -9109.231 0.079 0.085
Chain 1: 5000 -15993.489 0.119 0.086
Chain 1: 5100 -9039.339 0.187 0.086
Chain 1: 5200 -8821.298 0.186 0.086
Chain 1: 5300 -12426.063 0.207 0.127
Chain 1: 5400 -12427.535 0.199 0.127
Chain 1: 5500 -10514.179 0.213 0.158
Chain 1: 5600 -9417.949 0.212 0.158
Chain 1: 5700 -11991.076 0.218 0.182
Chain 1: 5800 -9019.099 0.242 0.215
Chain 1: 5900 -9100.232 0.237 0.215
Chain 1: 6000 -8571.538 0.200 0.182
Chain 1: 6100 -8770.956 0.125 0.116
Chain 1: 6200 -12396.094 0.152 0.182
Chain 1: 6300 -11624.164 0.129 0.116
Chain 1: 6400 -8487.469 0.166 0.182
Chain 1: 6500 -9163.445 0.156 0.116
Chain 1: 6600 -9234.666 0.145 0.074
Chain 1: 6700 -9886.707 0.130 0.066
Chain 1: 6800 -12966.069 0.121 0.066
Chain 1: 6900 -14103.791 0.128 0.074
Chain 1: 7000 -8533.603 0.187 0.081
Chain 1: 7100 -8452.202 0.186 0.081
Chain 1: 7200 -8796.030 0.160 0.074
Chain 1: 7300 -8207.825 0.161 0.074
Chain 1: 7400 -10296.370 0.144 0.074
Chain 1: 7500 -8371.902 0.160 0.081
Chain 1: 7600 -8484.707 0.160 0.081
Chain 1: 7700 -10065.334 0.169 0.157
Chain 1: 7800 -9962.388 0.147 0.081
Chain 1: 7900 -9102.474 0.148 0.094
Chain 1: 8000 -10463.013 0.096 0.094
Chain 1: 8100 -9445.636 0.106 0.108
Chain 1: 8200 -8981.763 0.107 0.108
Chain 1: 8300 -10865.856 0.117 0.130
Chain 1: 8400 -8722.045 0.121 0.130
Chain 1: 8500 -8798.986 0.099 0.108
Chain 1: 8600 -8701.127 0.099 0.108
Chain 1: 8700 -8640.625 0.084 0.094
Chain 1: 8800 -8561.900 0.084 0.094
Chain 1: 8900 -8695.866 0.076 0.052
Chain 1: 9000 -9143.971 0.068 0.049
Chain 1: 9100 -10289.457 0.068 0.049
Chain 1: 9200 -8737.194 0.081 0.049
Chain 1: 9300 -9661.827 0.073 0.049
Chain 1: 9400 -8299.776 0.065 0.049
Chain 1: 9500 -8691.557 0.069 0.049
Chain 1: 9600 -8434.044 0.071 0.049
Chain 1: 9700 -8455.693 0.070 0.049
Chain 1: 9800 -11953.951 0.098 0.096
Chain 1: 9900 -8504.426 0.137 0.111
Chain 1: 10000 -10096.756 0.148 0.158
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001373 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.73 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57018.840 1.000 1.000
Chain 1: 200 -17554.151 1.624 2.248
Chain 1: 300 -8748.123 1.418 1.007
Chain 1: 400 -8230.075 1.079 1.007
Chain 1: 500 -8585.990 0.872 1.000
Chain 1: 600 -8161.852 0.735 1.000
Chain 1: 700 -8214.081 0.631 0.063
Chain 1: 800 -8008.598 0.555 0.063
Chain 1: 900 -7826.327 0.496 0.052
Chain 1: 1000 -7647.488 0.449 0.052
Chain 1: 1100 -7672.865 0.349 0.041
Chain 1: 1200 -7686.119 0.125 0.026
Chain 1: 1300 -7605.913 0.025 0.023
Chain 1: 1400 -7801.455 0.021 0.023
Chain 1: 1500 -7599.184 0.020 0.023
Chain 1: 1600 -7657.450 0.015 0.023
Chain 1: 1700 -7522.320 0.017 0.023
Chain 1: 1800 -7574.239 0.015 0.018
Chain 1: 1900 -7586.779 0.012 0.011
Chain 1: 2000 -7623.737 0.011 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003334 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.34 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85942.358 1.000 1.000
Chain 1: 200 -13559.868 3.169 5.338
Chain 1: 300 -9953.279 2.233 1.000
Chain 1: 400 -10885.515 1.696 1.000
Chain 1: 500 -8835.263 1.404 0.362
Chain 1: 600 -8431.500 1.178 0.362
Chain 1: 700 -8594.979 1.012 0.232
Chain 1: 800 -8894.565 0.890 0.232
Chain 1: 900 -8751.593 0.793 0.086
Chain 1: 1000 -8553.225 0.716 0.086
Chain 1: 1100 -8782.845 0.618 0.048 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8416.093 0.089 0.044
Chain 1: 1300 -8636.761 0.055 0.034
Chain 1: 1400 -8642.011 0.047 0.026
Chain 1: 1500 -8534.410 0.025 0.026
Chain 1: 1600 -8637.497 0.021 0.023
Chain 1: 1700 -8725.463 0.020 0.023
Chain 1: 1800 -8317.310 0.022 0.023
Chain 1: 1900 -8413.915 0.021 0.023
Chain 1: 2000 -8386.181 0.019 0.013
Chain 1: 2100 -8506.940 0.018 0.013
Chain 1: 2200 -8325.752 0.016 0.013
Chain 1: 2300 -8453.333 0.015 0.013
Chain 1: 2400 -8463.791 0.015 0.013
Chain 1: 2500 -8425.781 0.014 0.012
Chain 1: 2600 -8424.642 0.013 0.011
Chain 1: 2700 -8339.487 0.013 0.011
Chain 1: 2800 -8304.369 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00332 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.2 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8401382.913 1.000 1.000
Chain 1: 200 -1586146.944 2.648 4.297
Chain 1: 300 -891748.396 2.025 1.000
Chain 1: 400 -457827.775 1.756 1.000
Chain 1: 500 -358082.814 1.460 0.948
Chain 1: 600 -233026.254 1.306 0.948
Chain 1: 700 -119265.045 1.256 0.948
Chain 1: 800 -86478.622 1.146 0.948
Chain 1: 900 -66828.520 1.052 0.779
Chain 1: 1000 -51631.952 0.976 0.779
Chain 1: 1100 -39111.235 0.908 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38286.145 0.480 0.379
Chain 1: 1300 -26246.660 0.448 0.379
Chain 1: 1400 -25965.350 0.355 0.320
Chain 1: 1500 -22553.671 0.342 0.320
Chain 1: 1600 -21770.087 0.292 0.294
Chain 1: 1700 -20644.621 0.202 0.294
Chain 1: 1800 -20588.770 0.164 0.151
Chain 1: 1900 -20914.780 0.137 0.055
Chain 1: 2000 -19426.396 0.115 0.055
Chain 1: 2100 -19664.798 0.084 0.036
Chain 1: 2200 -19891.131 0.083 0.036
Chain 1: 2300 -19508.477 0.039 0.020
Chain 1: 2400 -19280.612 0.039 0.020
Chain 1: 2500 -19082.585 0.025 0.016
Chain 1: 2600 -18713.026 0.023 0.016
Chain 1: 2700 -18670.014 0.018 0.012
Chain 1: 2800 -18386.945 0.019 0.015
Chain 1: 2900 -18668.071 0.019 0.015
Chain 1: 3000 -18654.344 0.012 0.012
Chain 1: 3100 -18739.312 0.011 0.012
Chain 1: 3200 -18430.092 0.012 0.015
Chain 1: 3300 -18634.706 0.011 0.012
Chain 1: 3400 -18109.813 0.012 0.015
Chain 1: 3500 -18721.430 0.015 0.015
Chain 1: 3600 -18028.441 0.017 0.015
Chain 1: 3700 -18414.995 0.018 0.017
Chain 1: 3800 -17375.232 0.023 0.021
Chain 1: 3900 -17371.378 0.021 0.021
Chain 1: 4000 -17488.691 0.022 0.021
Chain 1: 4100 -17402.496 0.022 0.021
Chain 1: 4200 -17218.838 0.021 0.021
Chain 1: 4300 -17357.175 0.021 0.021
Chain 1: 4400 -17314.086 0.018 0.011
Chain 1: 4500 -17216.628 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001512 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.12 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48814.176 1.000 1.000
Chain 1: 200 -47311.413 0.516 1.000
Chain 1: 300 -39049.208 0.414 0.212
Chain 1: 400 -21766.369 0.509 0.794
Chain 1: 500 -13037.321 0.541 0.670
Chain 1: 600 -14969.517 0.473 0.670
Chain 1: 700 -21311.301 0.448 0.298
Chain 1: 800 -29252.576 0.426 0.298
Chain 1: 900 -12385.522 0.530 0.298
Chain 1: 1000 -24722.911 0.527 0.499
Chain 1: 1100 -19596.427 0.453 0.298
Chain 1: 1200 -17236.406 0.463 0.298
Chain 1: 1300 -18799.186 0.450 0.298
Chain 1: 1400 -10690.267 0.447 0.298
Chain 1: 1500 -9246.227 0.396 0.271
Chain 1: 1600 -13489.627 0.414 0.298
Chain 1: 1700 -11349.939 0.403 0.271
Chain 1: 1800 -22601.933 0.426 0.315
Chain 1: 1900 -10446.469 0.406 0.315
Chain 1: 2000 -10362.500 0.357 0.262
Chain 1: 2100 -9730.608 0.337 0.189
Chain 1: 2200 -12401.796 0.345 0.215
Chain 1: 2300 -11261.125 0.347 0.215
Chain 1: 2400 -8677.941 0.301 0.215
Chain 1: 2500 -11160.647 0.307 0.222
Chain 1: 2600 -10433.402 0.283 0.215
Chain 1: 2700 -10416.480 0.264 0.215
Chain 1: 2800 -9141.466 0.228 0.139
Chain 1: 2900 -9001.179 0.114 0.101
Chain 1: 3000 -10303.941 0.125 0.126
Chain 1: 3100 -14769.731 0.149 0.139
Chain 1: 3200 -9276.715 0.187 0.139
Chain 1: 3300 -9028.001 0.179 0.139
Chain 1: 3400 -10202.290 0.161 0.126
Chain 1: 3500 -8791.240 0.155 0.126
Chain 1: 3600 -8642.438 0.150 0.126
Chain 1: 3700 -9519.590 0.159 0.126
Chain 1: 3800 -12204.414 0.167 0.126
Chain 1: 3900 -8553.866 0.208 0.161
Chain 1: 4000 -8687.113 0.197 0.161
Chain 1: 4100 -9716.427 0.177 0.115
Chain 1: 4200 -13052.680 0.144 0.115
Chain 1: 4300 -12564.107 0.145 0.115
Chain 1: 4400 -12364.070 0.135 0.106
Chain 1: 4500 -8693.136 0.161 0.106
Chain 1: 4600 -8534.199 0.161 0.106
Chain 1: 4700 -10429.845 0.170 0.182
Chain 1: 4800 -8613.214 0.169 0.182
Chain 1: 4900 -8318.162 0.130 0.106
Chain 1: 5000 -12741.521 0.163 0.182
Chain 1: 5100 -9317.519 0.189 0.211
Chain 1: 5200 -9068.973 0.167 0.182
Chain 1: 5300 -10133.148 0.173 0.182
Chain 1: 5400 -9272.886 0.181 0.182
Chain 1: 5500 -8572.374 0.147 0.105
Chain 1: 5600 -11160.946 0.168 0.182
Chain 1: 5700 -8629.396 0.179 0.211
Chain 1: 5800 -8957.038 0.162 0.105
Chain 1: 5900 -9065.563 0.160 0.105
Chain 1: 6000 -9081.751 0.125 0.093
Chain 1: 6100 -8558.654 0.094 0.082
Chain 1: 6200 -8887.736 0.095 0.082
Chain 1: 6300 -13650.648 0.120 0.082
Chain 1: 6400 -11944.379 0.125 0.082
Chain 1: 6500 -9539.307 0.142 0.143
Chain 1: 6600 -12691.400 0.143 0.143
Chain 1: 6700 -8330.321 0.166 0.143
Chain 1: 6800 -12851.164 0.198 0.248
Chain 1: 6900 -8470.447 0.248 0.252
Chain 1: 7000 -10936.471 0.271 0.252
Chain 1: 7100 -8121.062 0.299 0.347
Chain 1: 7200 -8371.304 0.299 0.347
Chain 1: 7300 -8250.602 0.265 0.252
Chain 1: 7400 -8807.645 0.257 0.252
Chain 1: 7500 -9626.447 0.241 0.248
Chain 1: 7600 -8724.957 0.226 0.225
Chain 1: 7700 -10571.788 0.191 0.175
Chain 1: 7800 -9264.517 0.170 0.141
Chain 1: 7900 -9695.939 0.123 0.103
Chain 1: 8000 -12573.455 0.123 0.103
Chain 1: 8100 -8464.109 0.137 0.103
Chain 1: 8200 -9458.044 0.145 0.105
Chain 1: 8300 -8160.915 0.159 0.141
Chain 1: 8400 -8063.915 0.154 0.141
Chain 1: 8500 -7989.423 0.146 0.141
Chain 1: 8600 -8196.197 0.139 0.141
Chain 1: 8700 -8338.757 0.123 0.105
Chain 1: 8800 -8196.020 0.110 0.044
Chain 1: 8900 -8883.608 0.114 0.077
Chain 1: 9000 -10029.722 0.102 0.077
Chain 1: 9100 -8707.385 0.069 0.077
Chain 1: 9200 -8434.714 0.062 0.032
Chain 1: 9300 -8315.910 0.047 0.025
Chain 1: 9400 -11328.752 0.073 0.032
Chain 1: 9500 -8209.349 0.110 0.077
Chain 1: 9600 -8080.654 0.109 0.077
Chain 1: 9700 -8683.426 0.114 0.077
Chain 1: 9800 -8181.566 0.118 0.077
Chain 1: 9900 -9112.484 0.121 0.102
Chain 1: 10000 -8151.104 0.121 0.102
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001581 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.81 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57935.175 1.000 1.000
Chain 1: 200 -17492.703 1.656 2.312
Chain 1: 300 -8591.727 1.449 1.036
Chain 1: 400 -8152.643 1.100 1.036
Chain 1: 500 -8037.467 0.883 1.000
Chain 1: 600 -8587.572 0.747 1.000
Chain 1: 700 -8199.583 0.647 0.064
Chain 1: 800 -8039.522 0.568 0.064
Chain 1: 900 -7789.250 0.509 0.054
Chain 1: 1000 -7740.297 0.459 0.054
Chain 1: 1100 -7658.481 0.360 0.047
Chain 1: 1200 -7656.547 0.128 0.032
Chain 1: 1300 -7642.650 0.025 0.020
Chain 1: 1400 -7805.572 0.022 0.020
Chain 1: 1500 -7594.913 0.023 0.021
Chain 1: 1600 -7750.549 0.019 0.020
Chain 1: 1700 -7471.708 0.018 0.020
Chain 1: 1800 -7565.679 0.017 0.020
Chain 1: 1900 -7536.659 0.014 0.012
Chain 1: 2000 -7559.589 0.014 0.012
Chain 1: 2100 -7559.529 0.013 0.012
Chain 1: 2200 -7661.765 0.014 0.013
Chain 1: 2300 -7796.900 0.016 0.017
Chain 1: 2400 -7610.837 0.016 0.017
Chain 1: 2500 -7618.653 0.013 0.013
Chain 1: 2600 -7473.088 0.013 0.013
Chain 1: 2700 -7557.105 0.011 0.012
Chain 1: 2800 -7530.818 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002746 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 27.46 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86753.506 1.000 1.000
Chain 1: 200 -13320.780 3.256 5.513
Chain 1: 300 -9691.663 2.296 1.000
Chain 1: 400 -10687.862 1.745 1.000
Chain 1: 500 -8509.963 1.447 0.374
Chain 1: 600 -8496.161 1.206 0.374
Chain 1: 700 -8130.466 1.040 0.256
Chain 1: 800 -8427.168 0.915 0.256
Chain 1: 900 -8479.100 0.814 0.093
Chain 1: 1000 -8387.786 0.734 0.093
Chain 1: 1100 -8539.324 0.635 0.045 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8171.300 0.089 0.045
Chain 1: 1300 -8385.287 0.054 0.035
Chain 1: 1400 -8397.752 0.044 0.026
Chain 1: 1500 -8253.181 0.021 0.018
Chain 1: 1600 -8365.941 0.022 0.018
Chain 1: 1700 -8448.654 0.018 0.018
Chain 1: 1800 -8036.432 0.020 0.018
Chain 1: 1900 -8132.531 0.020 0.018
Chain 1: 2000 -8105.664 0.020 0.018
Chain 1: 2100 -8228.157 0.019 0.015
Chain 1: 2200 -8048.184 0.017 0.015
Chain 1: 2300 -8127.543 0.016 0.013
Chain 1: 2400 -8197.166 0.016 0.013
Chain 1: 2500 -8142.483 0.015 0.012
Chain 1: 2600 -8141.831 0.014 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003272 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.72 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8418601.167 1.000 1.000
Chain 1: 200 -1590568.563 2.646 4.293
Chain 1: 300 -891294.973 2.026 1.000
Chain 1: 400 -456941.860 1.757 1.000
Chain 1: 500 -356993.584 1.462 0.951
Chain 1: 600 -231810.391 1.308 0.951
Chain 1: 700 -118499.526 1.258 0.951
Chain 1: 800 -85823.790 1.148 0.951
Chain 1: 900 -66273.498 1.053 0.785
Chain 1: 1000 -51167.024 0.978 0.785
Chain 1: 1100 -38727.304 0.910 0.540 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37914.189 0.482 0.381
Chain 1: 1300 -25956.509 0.450 0.381
Chain 1: 1400 -25683.075 0.356 0.321
Chain 1: 1500 -22292.315 0.343 0.321
Chain 1: 1600 -21515.149 0.293 0.295
Chain 1: 1700 -20399.310 0.203 0.295
Chain 1: 1800 -20345.858 0.165 0.152
Chain 1: 1900 -20672.017 0.137 0.055
Chain 1: 2000 -19188.291 0.115 0.055
Chain 1: 2100 -19426.585 0.084 0.036
Chain 1: 2200 -19652.123 0.083 0.036
Chain 1: 2300 -19270.078 0.039 0.020
Chain 1: 2400 -19042.270 0.039 0.020
Chain 1: 2500 -18843.841 0.025 0.016
Chain 1: 2600 -18474.635 0.024 0.016
Chain 1: 2700 -18431.689 0.018 0.012
Chain 1: 2800 -18148.406 0.020 0.016
Chain 1: 2900 -18429.482 0.020 0.015
Chain 1: 3000 -18415.789 0.012 0.012
Chain 1: 3100 -18500.792 0.011 0.012
Chain 1: 3200 -18191.609 0.012 0.015
Chain 1: 3300 -18396.186 0.011 0.012
Chain 1: 3400 -17871.211 0.013 0.015
Chain 1: 3500 -18482.845 0.015 0.016
Chain 1: 3600 -17789.733 0.017 0.016
Chain 1: 3700 -18176.337 0.019 0.017
Chain 1: 3800 -17136.361 0.023 0.021
Chain 1: 3900 -17132.419 0.022 0.021
Chain 1: 4000 -17249.799 0.022 0.021
Chain 1: 4100 -17163.597 0.022 0.021
Chain 1: 4200 -16979.853 0.022 0.021
Chain 1: 4300 -17118.300 0.021 0.021
Chain 1: 4400 -17075.184 0.019 0.011
Chain 1: 4500 -16977.640 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001431 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.31 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12532.994 1.000 1.000
Chain 1: 200 -9446.777 0.663 1.000
Chain 1: 300 -8225.690 0.492 0.327
Chain 1: 400 -8422.391 0.375 0.327
Chain 1: 500 -8263.666 0.304 0.148
Chain 1: 600 -8142.914 0.255 0.148
Chain 1: 700 -8075.754 0.220 0.023
Chain 1: 800 -8049.986 0.193 0.023
Chain 1: 900 -7960.665 0.173 0.019
Chain 1: 1000 -8080.983 0.157 0.019
Chain 1: 1100 -8146.601 0.058 0.015
Chain 1: 1200 -8059.423 0.026 0.015
Chain 1: 1300 -8052.860 0.011 0.011
Chain 1: 1400 -8043.170 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001522 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.22 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57371.041 1.000 1.000
Chain 1: 200 -17603.026 1.630 2.259
Chain 1: 300 -8799.569 1.420 1.000
Chain 1: 400 -8136.589 1.085 1.000
Chain 1: 500 -8575.845 0.878 1.000
Chain 1: 600 -8254.233 0.739 1.000
Chain 1: 700 -8556.733 0.638 0.081
Chain 1: 800 -8386.952 0.561 0.081
Chain 1: 900 -8099.269 0.502 0.051
Chain 1: 1000 -7824.217 0.456 0.051
Chain 1: 1100 -7797.713 0.356 0.039
Chain 1: 1200 -7964.169 0.132 0.036
Chain 1: 1300 -7584.021 0.037 0.036
Chain 1: 1400 -8046.154 0.035 0.036
Chain 1: 1500 -7599.308 0.036 0.036
Chain 1: 1600 -7785.328 0.034 0.035
Chain 1: 1700 -7514.400 0.034 0.036
Chain 1: 1800 -7554.903 0.033 0.036
Chain 1: 1900 -7590.772 0.030 0.035
Chain 1: 2000 -7649.115 0.027 0.024
Chain 1: 2100 -7564.388 0.028 0.024
Chain 1: 2200 -7720.376 0.028 0.024
Chain 1: 2300 -7538.320 0.025 0.024
Chain 1: 2400 -7555.286 0.019 0.020
Chain 1: 2500 -7458.881 0.015 0.013
Chain 1: 2600 -7527.429 0.013 0.011
Chain 1: 2700 -7456.062 0.011 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003491 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.91 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86279.223 1.000 1.000
Chain 1: 200 -13677.598 3.154 5.308
Chain 1: 300 -10041.516 2.223 1.000
Chain 1: 400 -11094.049 1.691 1.000
Chain 1: 500 -8875.438 1.403 0.362
Chain 1: 600 -8914.186 1.170 0.362
Chain 1: 700 -8526.843 1.009 0.250
Chain 1: 800 -8809.342 0.887 0.250
Chain 1: 900 -8887.883 0.790 0.095
Chain 1: 1000 -8699.141 0.713 0.095
Chain 1: 1100 -8836.630 0.614 0.045 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8499.576 0.087 0.040
Chain 1: 1300 -8725.667 0.054 0.032
Chain 1: 1400 -8740.213 0.045 0.026
Chain 1: 1500 -8589.899 0.021 0.022
Chain 1: 1600 -8703.630 0.022 0.022
Chain 1: 1700 -8783.313 0.019 0.017
Chain 1: 1800 -8365.026 0.020 0.017
Chain 1: 1900 -8463.467 0.021 0.017
Chain 1: 2000 -8437.432 0.019 0.016
Chain 1: 2100 -8561.392 0.019 0.014
Chain 1: 2200 -8375.012 0.017 0.014
Chain 1: 2300 -8458.076 0.015 0.013
Chain 1: 2400 -8527.609 0.016 0.013
Chain 1: 2500 -8473.551 0.015 0.012
Chain 1: 2600 -8473.862 0.013 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002969 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 29.69 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8410824.455 1.000 1.000
Chain 1: 200 -1583814.094 2.655 4.310
Chain 1: 300 -890633.967 2.030 1.000
Chain 1: 400 -458281.048 1.758 1.000
Chain 1: 500 -358735.694 1.462 0.943
Chain 1: 600 -233567.250 1.308 0.943
Chain 1: 700 -119541.025 1.257 0.943
Chain 1: 800 -86782.895 1.147 0.943
Chain 1: 900 -67071.018 1.052 0.778
Chain 1: 1000 -51841.158 0.976 0.778
Chain 1: 1100 -39300.214 0.908 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38471.547 0.479 0.377
Chain 1: 1300 -26398.060 0.447 0.377
Chain 1: 1400 -26115.176 0.354 0.319
Chain 1: 1500 -22696.414 0.341 0.319
Chain 1: 1600 -21912.051 0.291 0.294
Chain 1: 1700 -20781.299 0.201 0.294
Chain 1: 1800 -20724.634 0.164 0.151
Chain 1: 1900 -21050.853 0.136 0.054
Chain 1: 2000 -19560.163 0.114 0.054
Chain 1: 2100 -19798.320 0.084 0.036
Chain 1: 2200 -20025.490 0.083 0.036
Chain 1: 2300 -19642.063 0.039 0.020
Chain 1: 2400 -19414.092 0.039 0.020
Chain 1: 2500 -19216.579 0.025 0.015
Chain 1: 2600 -18846.299 0.023 0.015
Chain 1: 2700 -18803.082 0.018 0.012
Chain 1: 2800 -18520.193 0.019 0.015
Chain 1: 2900 -18801.458 0.019 0.015
Chain 1: 3000 -18787.457 0.012 0.012
Chain 1: 3100 -18872.563 0.011 0.012
Chain 1: 3200 -18563.089 0.012 0.015
Chain 1: 3300 -18767.916 0.011 0.012
Chain 1: 3400 -18242.847 0.012 0.015
Chain 1: 3500 -18854.890 0.015 0.015
Chain 1: 3600 -18161.251 0.016 0.015
Chain 1: 3700 -18548.375 0.018 0.017
Chain 1: 3800 -17507.797 0.023 0.021
Chain 1: 3900 -17503.984 0.021 0.021
Chain 1: 4000 -17621.203 0.022 0.021
Chain 1: 4100 -17535.071 0.022 0.021
Chain 1: 4200 -17351.180 0.021 0.021
Chain 1: 4300 -17489.591 0.021 0.021
Chain 1: 4400 -17446.337 0.018 0.011
Chain 1: 4500 -17348.916 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001316 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.16 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12247.299 1.000 1.000
Chain 1: 200 -9209.896 0.665 1.000
Chain 1: 300 -7976.628 0.495 0.330
Chain 1: 400 -8182.420 0.377 0.330
Chain 1: 500 -8038.768 0.305 0.155
Chain 1: 600 -7896.224 0.258 0.155
Chain 1: 700 -7818.700 0.222 0.025
Chain 1: 800 -7826.051 0.195 0.025
Chain 1: 900 -7722.706 0.174 0.018
Chain 1: 1000 -7826.859 0.158 0.018
Chain 1: 1100 -7891.902 0.059 0.018
Chain 1: 1200 -7845.739 0.027 0.013
Chain 1: 1300 -7786.149 0.012 0.013
Chain 1: 1400 -7808.401 0.010 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001441 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.41 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56590.786 1.000 1.000
Chain 1: 200 -17236.769 1.642 2.283
Chain 1: 300 -8668.444 1.424 1.000
Chain 1: 400 -8165.811 1.083 1.000
Chain 1: 500 -8068.608 0.869 0.988
Chain 1: 600 -8645.804 0.735 0.988
Chain 1: 700 -7792.190 0.646 0.110
Chain 1: 800 -8114.059 0.570 0.110
Chain 1: 900 -8031.875 0.508 0.067
Chain 1: 1000 -7953.156 0.458 0.067
Chain 1: 1100 -7814.911 0.360 0.062
Chain 1: 1200 -7749.272 0.132 0.040
Chain 1: 1300 -7670.133 0.035 0.018
Chain 1: 1400 -7707.577 0.029 0.012
Chain 1: 1500 -7609.307 0.029 0.013
Chain 1: 1600 -7657.783 0.023 0.010
Chain 1: 1700 -7564.645 0.013 0.010
Chain 1: 1800 -7614.240 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003213 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.13 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86571.657 1.000 1.000
Chain 1: 200 -13374.372 3.236 5.473
Chain 1: 300 -9770.299 2.281 1.000
Chain 1: 400 -10554.979 1.729 1.000
Chain 1: 500 -8716.745 1.425 0.369
Chain 1: 600 -8288.133 1.196 0.369
Chain 1: 700 -8182.510 1.027 0.211
Chain 1: 800 -8675.565 0.906 0.211
Chain 1: 900 -8608.394 0.806 0.074
Chain 1: 1000 -8320.566 0.729 0.074
Chain 1: 1100 -8569.489 0.632 0.057 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8244.181 0.089 0.052
Chain 1: 1300 -8465.439 0.054 0.039
Chain 1: 1400 -8460.814 0.047 0.035
Chain 1: 1500 -8352.488 0.027 0.029
Chain 1: 1600 -8455.892 0.023 0.026
Chain 1: 1700 -8545.388 0.023 0.026
Chain 1: 1800 -8140.088 0.022 0.026
Chain 1: 1900 -8238.425 0.023 0.026
Chain 1: 2000 -8210.026 0.020 0.013
Chain 1: 2100 -8329.923 0.018 0.013
Chain 1: 2200 -8138.323 0.017 0.013
Chain 1: 2300 -8275.488 0.016 0.013
Chain 1: 2400 -8277.925 0.016 0.013
Chain 1: 2500 -8252.582 0.015 0.012
Chain 1: 2600 -8251.931 0.013 0.012
Chain 1: 2700 -8162.499 0.013 0.012
Chain 1: 2800 -8129.255 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003604 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 36.04 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8403762.973 1.000 1.000
Chain 1: 200 -1584822.164 2.651 4.303
Chain 1: 300 -891108.827 2.027 1.000
Chain 1: 400 -457851.118 1.757 1.000
Chain 1: 500 -358240.655 1.461 0.946
Chain 1: 600 -232911.672 1.307 0.946
Chain 1: 700 -119085.832 1.257 0.946
Chain 1: 800 -86284.827 1.147 0.946
Chain 1: 900 -66613.560 1.053 0.778
Chain 1: 1000 -51413.202 0.977 0.778
Chain 1: 1100 -38894.730 0.909 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38070.688 0.481 0.380
Chain 1: 1300 -26037.929 0.449 0.380
Chain 1: 1400 -25756.238 0.356 0.322
Chain 1: 1500 -22346.988 0.343 0.322
Chain 1: 1600 -21564.142 0.293 0.296
Chain 1: 1700 -20439.676 0.203 0.295
Chain 1: 1800 -20384.162 0.165 0.153
Chain 1: 1900 -20710.102 0.137 0.055
Chain 1: 2000 -19222.663 0.116 0.055
Chain 1: 2100 -19460.868 0.085 0.036
Chain 1: 2200 -19687.039 0.084 0.036
Chain 1: 2300 -19304.576 0.039 0.020
Chain 1: 2400 -19076.810 0.040 0.020
Chain 1: 2500 -18878.741 0.025 0.016
Chain 1: 2600 -18509.161 0.024 0.016
Chain 1: 2700 -18466.260 0.018 0.012
Chain 1: 2800 -18183.169 0.020 0.016
Chain 1: 2900 -18464.343 0.020 0.015
Chain 1: 3000 -18450.491 0.012 0.012
Chain 1: 3100 -18535.454 0.011 0.012
Chain 1: 3200 -18226.291 0.012 0.015
Chain 1: 3300 -18430.923 0.011 0.012
Chain 1: 3400 -17906.089 0.013 0.015
Chain 1: 3500 -18517.535 0.015 0.016
Chain 1: 3600 -17824.820 0.017 0.016
Chain 1: 3700 -18211.170 0.019 0.017
Chain 1: 3800 -17171.735 0.023 0.021
Chain 1: 3900 -17167.918 0.022 0.021
Chain 1: 4000 -17285.216 0.022 0.021
Chain 1: 4100 -17199.001 0.022 0.021
Chain 1: 4200 -17015.478 0.022 0.021
Chain 1: 4300 -17153.707 0.021 0.021
Chain 1: 4400 -17110.694 0.019 0.011
Chain 1: 4500 -17013.279 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001355 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.55 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13323.448 1.000 1.000
Chain 1: 200 -10093.291 0.660 1.000
Chain 1: 300 -8700.547 0.493 0.320
Chain 1: 400 -8886.819 0.375 0.320
Chain 1: 500 -8503.334 0.309 0.160
Chain 1: 600 -8598.479 0.260 0.160
Chain 1: 700 -8895.678 0.227 0.045
Chain 1: 800 -8509.305 0.205 0.045
Chain 1: 900 -8589.651 0.183 0.045
Chain 1: 1000 -8545.477 0.165 0.045
Chain 1: 1100 -8603.749 0.066 0.033
Chain 1: 1200 -8523.656 0.035 0.021
Chain 1: 1300 -8490.504 0.019 0.011
Chain 1: 1400 -8489.561 0.017 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00145 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.5 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -63267.017 1.000 1.000
Chain 1: 200 -18971.827 1.667 2.335
Chain 1: 300 -9498.907 1.444 1.000
Chain 1: 400 -8488.216 1.113 1.000
Chain 1: 500 -8531.926 0.891 0.997
Chain 1: 600 -9361.869 0.757 0.997
Chain 1: 700 -9741.323 0.655 0.119
Chain 1: 800 -8731.193 0.587 0.119
Chain 1: 900 -7910.422 0.534 0.116
Chain 1: 1000 -8014.085 0.482 0.116
Chain 1: 1100 -7925.428 0.383 0.104
Chain 1: 1200 -7755.984 0.151 0.089
Chain 1: 1300 -8010.794 0.055 0.039
Chain 1: 1400 -7880.493 0.045 0.032
Chain 1: 1500 -7669.844 0.047 0.032
Chain 1: 1600 -7985.958 0.042 0.032
Chain 1: 1700 -7697.552 0.042 0.032
Chain 1: 1800 -7652.349 0.031 0.027
Chain 1: 1900 -7660.913 0.021 0.022
Chain 1: 2000 -7905.798 0.022 0.027
Chain 1: 2100 -7721.969 0.024 0.027
Chain 1: 2200 -7935.163 0.024 0.027
Chain 1: 2300 -7715.044 0.024 0.027
Chain 1: 2400 -7729.428 0.022 0.027
Chain 1: 2500 -7744.328 0.020 0.027
Chain 1: 2600 -7648.979 0.017 0.024
Chain 1: 2700 -7620.794 0.014 0.012
Chain 1: 2800 -7688.588 0.014 0.012
Chain 1: 2900 -7501.682 0.016 0.024
Chain 1: 3000 -7657.970 0.015 0.020
Chain 1: 3100 -7636.755 0.013 0.012
Chain 1: 3200 -7864.718 0.013 0.012
Chain 1: 3300 -7565.339 0.015 0.012
Chain 1: 3400 -7846.830 0.018 0.020
Chain 1: 3500 -7564.051 0.021 0.025
Chain 1: 3600 -7606.657 0.021 0.025
Chain 1: 3700 -7576.016 0.021 0.025
Chain 1: 3800 -7567.160 0.020 0.025
Chain 1: 3900 -7514.973 0.018 0.020
Chain 1: 4000 -7502.996 0.016 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002601 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 26.01 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87496.689 1.000 1.000
Chain 1: 200 -14527.356 3.011 5.023
Chain 1: 300 -10718.415 2.126 1.000
Chain 1: 400 -12723.511 1.634 1.000
Chain 1: 500 -9105.635 1.387 0.397
Chain 1: 600 -9198.170 1.157 0.397
Chain 1: 700 -9530.892 0.997 0.355
Chain 1: 800 -9821.449 0.876 0.355
Chain 1: 900 -9594.492 0.781 0.158
Chain 1: 1000 -9732.730 0.705 0.158
Chain 1: 1100 -9402.978 0.608 0.035 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8982.930 0.110 0.035
Chain 1: 1300 -9298.937 0.078 0.035
Chain 1: 1400 -9212.263 0.063 0.034
Chain 1: 1500 -9196.342 0.024 0.030
Chain 1: 1600 -9238.558 0.023 0.030
Chain 1: 1700 -9305.255 0.021 0.024
Chain 1: 1800 -8860.958 0.023 0.024
Chain 1: 1900 -8959.372 0.021 0.014
Chain 1: 2000 -8980.691 0.020 0.011
Chain 1: 2100 -9066.968 0.018 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003397 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.97 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8436577.541 1.000 1.000
Chain 1: 200 -1586849.014 2.658 4.317
Chain 1: 300 -891080.629 2.032 1.000
Chain 1: 400 -458405.202 1.760 1.000
Chain 1: 500 -358473.423 1.464 0.944
Chain 1: 600 -233485.291 1.309 0.944
Chain 1: 700 -119976.621 1.257 0.944
Chain 1: 800 -87280.751 1.147 0.944
Chain 1: 900 -67685.786 1.052 0.781
Chain 1: 1000 -52545.556 0.975 0.781
Chain 1: 1100 -40071.304 0.906 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39261.843 0.477 0.375
Chain 1: 1300 -27249.837 0.443 0.375
Chain 1: 1400 -26976.198 0.350 0.311
Chain 1: 1500 -23571.884 0.336 0.311
Chain 1: 1600 -22792.662 0.286 0.289
Chain 1: 1700 -21669.030 0.197 0.288
Chain 1: 1800 -21614.577 0.159 0.144
Chain 1: 1900 -21941.620 0.132 0.052
Chain 1: 2000 -20452.810 0.110 0.052
Chain 1: 2100 -20691.079 0.080 0.034
Chain 1: 2200 -20917.995 0.079 0.034
Chain 1: 2300 -20534.633 0.037 0.019
Chain 1: 2400 -20306.478 0.037 0.019
Chain 1: 2500 -20108.368 0.024 0.015
Chain 1: 2600 -19737.659 0.022 0.015
Chain 1: 2700 -19694.421 0.017 0.012
Chain 1: 2800 -19410.828 0.019 0.015
Chain 1: 2900 -19692.482 0.018 0.014
Chain 1: 3000 -19678.571 0.011 0.012
Chain 1: 3100 -19763.699 0.011 0.011
Chain 1: 3200 -19453.757 0.011 0.014
Chain 1: 3300 -19659.009 0.010 0.011
Chain 1: 3400 -19132.798 0.012 0.014
Chain 1: 3500 -19746.255 0.014 0.015
Chain 1: 3600 -19050.889 0.016 0.015
Chain 1: 3700 -19439.182 0.018 0.016
Chain 1: 3800 -18395.628 0.022 0.020
Chain 1: 3900 -18391.686 0.020 0.020
Chain 1: 4000 -18509.014 0.021 0.020
Chain 1: 4100 -18422.595 0.021 0.020
Chain 1: 4200 -18238.152 0.020 0.020
Chain 1: 4300 -18377.034 0.020 0.020
Chain 1: 4400 -18333.270 0.018 0.010
Chain 1: 4500 -18235.709 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.0015 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49298.331 1.000 1.000
Chain 1: 200 -113273.251 0.782 1.000
Chain 1: 300 -15819.394 2.575 1.000
Chain 1: 400 -17580.471 1.956 1.000
Chain 1: 500 -14035.422 1.616 0.565
Chain 1: 600 -16102.525 1.368 0.565
Chain 1: 700 -14312.534 1.190 0.253
Chain 1: 800 -11855.164 1.067 0.253
Chain 1: 900 -14716.289 0.970 0.207
Chain 1: 1000 -17859.761 0.891 0.207
Chain 1: 1100 -15406.983 0.807 0.194 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -12273.260 0.776 0.194 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1300 -12824.047 0.164 0.176
Chain 1: 1400 -16546.209 0.177 0.194
Chain 1: 1500 -12771.783 0.181 0.194
Chain 1: 1600 -11382.978 0.180 0.194
Chain 1: 1700 -11854.863 0.172 0.194
Chain 1: 1800 -11898.210 0.151 0.176
Chain 1: 1900 -10278.540 0.148 0.159
Chain 1: 2000 -21552.819 0.182 0.159
Chain 1: 2100 -12394.480 0.240 0.225
Chain 1: 2200 -10194.048 0.236 0.216
Chain 1: 2300 -12633.514 0.251 0.216
Chain 1: 2400 -9890.763 0.257 0.216
Chain 1: 2500 -22128.327 0.282 0.216
Chain 1: 2600 -9628.216 0.400 0.277
Chain 1: 2700 -9434.595 0.398 0.277
Chain 1: 2800 -10523.927 0.408 0.277
Chain 1: 2900 -9476.829 0.403 0.277
Chain 1: 3000 -13217.674 0.379 0.277
Chain 1: 3100 -9678.359 0.342 0.277
Chain 1: 3200 -9155.657 0.326 0.277
Chain 1: 3300 -9573.568 0.311 0.277
Chain 1: 3400 -10247.957 0.290 0.110
Chain 1: 3500 -9419.742 0.244 0.104
Chain 1: 3600 -9269.914 0.115 0.088
Chain 1: 3700 -8824.814 0.118 0.088
Chain 1: 3800 -10519.483 0.124 0.088
Chain 1: 3900 -10596.912 0.114 0.066
Chain 1: 4000 -10282.592 0.089 0.057
Chain 1: 4100 -9592.981 0.059 0.057
Chain 1: 4200 -9171.514 0.058 0.050
Chain 1: 4300 -9694.774 0.059 0.054
Chain 1: 4400 -9132.581 0.059 0.054
Chain 1: 4500 -9181.239 0.050 0.050
Chain 1: 4600 -11348.781 0.068 0.054
Chain 1: 4700 -12415.126 0.071 0.062
Chain 1: 4800 -9322.853 0.089 0.062
Chain 1: 4900 -8921.477 0.092 0.062
Chain 1: 5000 -12096.806 0.115 0.072
Chain 1: 5100 -8600.042 0.149 0.086
Chain 1: 5200 -13569.525 0.181 0.191
Chain 1: 5300 -9967.775 0.212 0.262
Chain 1: 5400 -9341.180 0.212 0.262
Chain 1: 5500 -8784.572 0.218 0.262
Chain 1: 5600 -9055.650 0.202 0.262
Chain 1: 5700 -9924.840 0.202 0.262
Chain 1: 5800 -8999.890 0.179 0.103
Chain 1: 5900 -17629.507 0.224 0.262
Chain 1: 6000 -9063.195 0.292 0.361
Chain 1: 6100 -12596.197 0.279 0.280
Chain 1: 6200 -9088.960 0.281 0.280
Chain 1: 6300 -8729.899 0.249 0.103
Chain 1: 6400 -10312.275 0.258 0.153
Chain 1: 6500 -9501.772 0.260 0.153
Chain 1: 6600 -8805.893 0.265 0.153
Chain 1: 6700 -14107.636 0.294 0.280
Chain 1: 6800 -11213.264 0.309 0.280
Chain 1: 6900 -13038.831 0.274 0.258
Chain 1: 7000 -8725.922 0.229 0.258
Chain 1: 7100 -8691.695 0.202 0.153
Chain 1: 7200 -8644.667 0.164 0.140
Chain 1: 7300 -10905.930 0.180 0.153
Chain 1: 7400 -8550.955 0.192 0.207
Chain 1: 7500 -12593.921 0.216 0.258
Chain 1: 7600 -9039.057 0.247 0.275
Chain 1: 7700 -9352.340 0.213 0.258
Chain 1: 7800 -9637.913 0.190 0.207
Chain 1: 7900 -9543.350 0.177 0.207
Chain 1: 8000 -8559.679 0.139 0.115
Chain 1: 8100 -8494.243 0.140 0.115
Chain 1: 8200 -8677.885 0.141 0.115
Chain 1: 8300 -13249.160 0.155 0.115
Chain 1: 8400 -9502.532 0.167 0.115
Chain 1: 8500 -9384.925 0.136 0.033
Chain 1: 8600 -8307.773 0.110 0.033
Chain 1: 8700 -8838.186 0.112 0.060
Chain 1: 8800 -10947.891 0.129 0.115
Chain 1: 8900 -10873.543 0.128 0.115
Chain 1: 9000 -8693.036 0.142 0.130
Chain 1: 9100 -9014.134 0.145 0.130
Chain 1: 9200 -10790.849 0.159 0.165
Chain 1: 9300 -8612.892 0.150 0.165
Chain 1: 9400 -11076.061 0.133 0.165
Chain 1: 9500 -8847.847 0.157 0.193
Chain 1: 9600 -8753.746 0.145 0.193
Chain 1: 9700 -8392.282 0.143 0.193
Chain 1: 9800 -9370.550 0.134 0.165
Chain 1: 9900 -11334.145 0.151 0.173
Chain 1: 10000 -8340.873 0.162 0.173
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00154 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.4 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57410.653 1.000 1.000
Chain 1: 200 -17875.295 1.606 2.212
Chain 1: 300 -8936.498 1.404 1.000
Chain 1: 400 -8174.915 1.076 1.000
Chain 1: 500 -8961.173 0.879 1.000
Chain 1: 600 -9305.726 0.738 1.000
Chain 1: 700 -8122.535 0.654 0.146
Chain 1: 800 -8456.329 0.577 0.146
Chain 1: 900 -7916.437 0.520 0.093
Chain 1: 1000 -7852.115 0.469 0.093
Chain 1: 1100 -8012.950 0.371 0.088
Chain 1: 1200 -7935.875 0.151 0.068
Chain 1: 1300 -7695.496 0.054 0.039
Chain 1: 1400 -7912.732 0.047 0.037
Chain 1: 1500 -7593.427 0.043 0.037
Chain 1: 1600 -7526.512 0.040 0.031
Chain 1: 1700 -7730.907 0.028 0.027
Chain 1: 1800 -7646.880 0.025 0.026
Chain 1: 1900 -7746.173 0.020 0.020
Chain 1: 2000 -7592.246 0.021 0.020
Chain 1: 2100 -7580.639 0.019 0.020
Chain 1: 2200 -7753.153 0.020 0.022
Chain 1: 2300 -7610.559 0.019 0.020
Chain 1: 2400 -7558.189 0.017 0.019
Chain 1: 2500 -7654.138 0.014 0.013
Chain 1: 2600 -7543.798 0.015 0.015
Chain 1: 2700 -7674.977 0.014 0.015
Chain 1: 2800 -7521.321 0.015 0.017
Chain 1: 2900 -7380.394 0.015 0.019
Chain 1: 3000 -7520.810 0.015 0.019
Chain 1: 3100 -7531.303 0.015 0.019
Chain 1: 3200 -7750.899 0.016 0.019
Chain 1: 3300 -7453.373 0.018 0.019
Chain 1: 3400 -7700.201 0.020 0.019
Chain 1: 3500 -7446.551 0.023 0.020
Chain 1: 3600 -7507.948 0.022 0.020
Chain 1: 3700 -7462.232 0.021 0.020
Chain 1: 3800 -7452.129 0.019 0.019
Chain 1: 3900 -7419.471 0.017 0.019
Chain 1: 4000 -7412.094 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003265 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.65 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86838.305 1.000 1.000
Chain 1: 200 -13971.125 3.108 5.216
Chain 1: 300 -10228.341 2.194 1.000
Chain 1: 400 -11807.239 1.679 1.000
Chain 1: 500 -8989.406 1.406 0.366
Chain 1: 600 -9460.013 1.180 0.366
Chain 1: 700 -8952.936 1.019 0.313
Chain 1: 800 -9706.358 0.902 0.313
Chain 1: 900 -9072.902 0.809 0.134
Chain 1: 1000 -9054.499 0.728 0.134
Chain 1: 1100 -8927.561 0.630 0.078 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8551.970 0.113 0.070
Chain 1: 1300 -8876.289 0.080 0.057
Chain 1: 1400 -8714.359 0.068 0.050
Chain 1: 1500 -8726.243 0.037 0.044
Chain 1: 1600 -8833.893 0.033 0.037
Chain 1: 1700 -8891.018 0.028 0.019
Chain 1: 1800 -8447.111 0.026 0.019
Chain 1: 1900 -8553.560 0.020 0.014
Chain 1: 2000 -8539.830 0.020 0.014
Chain 1: 2100 -8657.520 0.020 0.014
Chain 1: 2200 -8449.934 0.018 0.014
Chain 1: 2300 -8546.883 0.015 0.012
Chain 1: 2400 -8612.677 0.014 0.012
Chain 1: 2500 -8561.873 0.015 0.012
Chain 1: 2600 -8575.952 0.014 0.011
Chain 1: 2700 -8483.044 0.014 0.011
Chain 1: 2800 -8429.843 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003396 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.96 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8389208.174 1.000 1.000
Chain 1: 200 -1580478.611 2.654 4.308
Chain 1: 300 -890761.908 2.027 1.000
Chain 1: 400 -457891.613 1.757 1.000
Chain 1: 500 -358923.565 1.461 0.945
Chain 1: 600 -233892.765 1.306 0.945
Chain 1: 700 -119972.963 1.255 0.945
Chain 1: 800 -87158.563 1.146 0.945
Chain 1: 900 -67455.540 1.051 0.774
Chain 1: 1000 -52216.007 0.975 0.774
Chain 1: 1100 -39655.416 0.906 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38830.641 0.478 0.376
Chain 1: 1300 -26732.008 0.446 0.376
Chain 1: 1400 -26447.744 0.352 0.317
Chain 1: 1500 -23021.087 0.339 0.317
Chain 1: 1600 -22234.706 0.290 0.292
Chain 1: 1700 -21100.953 0.200 0.292
Chain 1: 1800 -21043.771 0.163 0.149
Chain 1: 1900 -21370.496 0.135 0.054
Chain 1: 2000 -19876.916 0.113 0.054
Chain 1: 2100 -20115.431 0.083 0.035
Chain 1: 2200 -20343.059 0.082 0.035
Chain 1: 2300 -19959.061 0.038 0.019
Chain 1: 2400 -19730.913 0.038 0.019
Chain 1: 2500 -19533.241 0.025 0.015
Chain 1: 2600 -19162.579 0.023 0.015
Chain 1: 2700 -19119.245 0.018 0.012
Chain 1: 2800 -18836.099 0.019 0.015
Chain 1: 2900 -19117.580 0.019 0.015
Chain 1: 3000 -19103.599 0.012 0.012
Chain 1: 3100 -19188.780 0.011 0.012
Chain 1: 3200 -18878.973 0.011 0.015
Chain 1: 3300 -19084.060 0.011 0.012
Chain 1: 3400 -18558.348 0.012 0.015
Chain 1: 3500 -19171.339 0.014 0.015
Chain 1: 3600 -18476.489 0.016 0.015
Chain 1: 3700 -18864.536 0.018 0.016
Chain 1: 3800 -17822.034 0.022 0.021
Chain 1: 3900 -17818.175 0.021 0.021
Chain 1: 4000 -17935.394 0.022 0.021
Chain 1: 4100 -17849.145 0.022 0.021
Chain 1: 4200 -17664.878 0.021 0.021
Chain 1: 4300 -17803.602 0.021 0.021
Chain 1: 4400 -17760.035 0.018 0.010
Chain 1: 4500 -17662.524 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001391 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.91 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48189.727 1.000 1.000
Chain 1: 200 -14672.454 1.642 2.284
Chain 1: 300 -16683.301 1.135 1.000
Chain 1: 400 -19121.600 0.883 1.000
Chain 1: 500 -14363.206 0.773 0.331
Chain 1: 600 -10772.616 0.700 0.333
Chain 1: 700 -12263.486 0.617 0.331
Chain 1: 800 -18023.801 0.580 0.331
Chain 1: 900 -10235.512 0.600 0.331
Chain 1: 1000 -12452.753 0.558 0.331
Chain 1: 1100 -10173.372 0.480 0.320
Chain 1: 1200 -9744.557 0.256 0.224
Chain 1: 1300 -13351.366 0.271 0.270
Chain 1: 1400 -10248.213 0.289 0.303
Chain 1: 1500 -9376.465 0.265 0.270
Chain 1: 1600 -13532.763 0.262 0.270
Chain 1: 1700 -9406.366 0.294 0.303
Chain 1: 1800 -12389.571 0.286 0.270
Chain 1: 1900 -10428.608 0.229 0.241
Chain 1: 2000 -9908.994 0.216 0.241
Chain 1: 2100 -11462.228 0.207 0.241
Chain 1: 2200 -10017.913 0.217 0.241
Chain 1: 2300 -9792.373 0.193 0.188
Chain 1: 2400 -8993.103 0.171 0.144
Chain 1: 2500 -11003.697 0.180 0.183
Chain 1: 2600 -9207.277 0.169 0.183
Chain 1: 2700 -9157.845 0.126 0.144
Chain 1: 2800 -8816.845 0.105 0.136
Chain 1: 2900 -9413.592 0.093 0.089
Chain 1: 3000 -9758.021 0.091 0.089
Chain 1: 3100 -8872.654 0.088 0.089
Chain 1: 3200 -17627.442 0.123 0.089
Chain 1: 3300 -15239.735 0.136 0.100
Chain 1: 3400 -15609.164 0.130 0.100
Chain 1: 3500 -9516.930 0.175 0.100
Chain 1: 3600 -9762.166 0.158 0.063
Chain 1: 3700 -8820.372 0.169 0.100
Chain 1: 3800 -15387.514 0.207 0.107
Chain 1: 3900 -9592.514 0.262 0.157
Chain 1: 4000 -8742.423 0.268 0.157
Chain 1: 4100 -9835.200 0.269 0.157
Chain 1: 4200 -13060.934 0.244 0.157
Chain 1: 4300 -9619.419 0.264 0.247
Chain 1: 4400 -10200.470 0.267 0.247
Chain 1: 4500 -8532.856 0.223 0.195
Chain 1: 4600 -9184.866 0.227 0.195
Chain 1: 4700 -8700.486 0.222 0.195
Chain 1: 4800 -8605.941 0.181 0.111
Chain 1: 4900 -15392.025 0.164 0.111
Chain 1: 5000 -9280.868 0.221 0.195
Chain 1: 5100 -8367.131 0.220 0.195
Chain 1: 5200 -11032.464 0.220 0.195
Chain 1: 5300 -8157.020 0.219 0.195
Chain 1: 5400 -8428.900 0.217 0.195
Chain 1: 5500 -9323.283 0.207 0.109
Chain 1: 5600 -9209.638 0.201 0.109
Chain 1: 5700 -9362.849 0.197 0.109
Chain 1: 5800 -8522.426 0.206 0.109
Chain 1: 5900 -9891.949 0.176 0.109
Chain 1: 6000 -8239.531 0.130 0.109
Chain 1: 6100 -11486.794 0.147 0.138
Chain 1: 6200 -8143.271 0.164 0.138
Chain 1: 6300 -8518.850 0.133 0.099
Chain 1: 6400 -11469.074 0.156 0.138
Chain 1: 6500 -8067.799 0.188 0.201
Chain 1: 6600 -9850.210 0.205 0.201
Chain 1: 6700 -10719.984 0.212 0.201
Chain 1: 6800 -9307.765 0.217 0.201
Chain 1: 6900 -8458.053 0.213 0.201
Chain 1: 7000 -11257.492 0.218 0.249
Chain 1: 7100 -8065.962 0.229 0.249
Chain 1: 7200 -13345.196 0.228 0.249
Chain 1: 7300 -8897.638 0.273 0.257
Chain 1: 7400 -7917.669 0.260 0.249
Chain 1: 7500 -7943.151 0.218 0.181
Chain 1: 7600 -11656.082 0.232 0.249
Chain 1: 7700 -7974.032 0.270 0.319
Chain 1: 7800 -8283.549 0.258 0.319
Chain 1: 7900 -8919.728 0.256 0.319
Chain 1: 8000 -9609.787 0.238 0.319
Chain 1: 8100 -8144.214 0.216 0.180
Chain 1: 8200 -7802.528 0.181 0.124
Chain 1: 8300 -11344.064 0.162 0.124
Chain 1: 8400 -10211.081 0.161 0.111
Chain 1: 8500 -8207.602 0.185 0.180
Chain 1: 8600 -7939.332 0.157 0.111
Chain 1: 8700 -10253.856 0.133 0.111
Chain 1: 8800 -7886.497 0.159 0.180
Chain 1: 8900 -11229.030 0.182 0.226
Chain 1: 9000 -7947.873 0.216 0.244
Chain 1: 9100 -10425.152 0.222 0.244
Chain 1: 9200 -8385.830 0.242 0.244
Chain 1: 9300 -8750.724 0.215 0.243
Chain 1: 9400 -10016.355 0.216 0.243
Chain 1: 9500 -7987.153 0.217 0.243
Chain 1: 9600 -8137.487 0.216 0.243
Chain 1: 9700 -10637.216 0.217 0.243
Chain 1: 9800 -8387.884 0.214 0.243
Chain 1: 9900 -9848.108 0.199 0.238
Chain 1: 10000 -8579.093 0.172 0.235
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001431 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.31 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56384.737 1.000 1.000
Chain 1: 200 -16896.536 1.669 2.337
Chain 1: 300 -8506.440 1.441 1.000
Chain 1: 400 -8624.376 1.084 1.000
Chain 1: 500 -8002.270 0.883 0.986
Chain 1: 600 -8678.454 0.749 0.986
Chain 1: 700 -7912.448 0.656 0.097
Chain 1: 800 -8135.983 0.577 0.097
Chain 1: 900 -7836.146 0.517 0.078
Chain 1: 1000 -7693.076 0.467 0.078
Chain 1: 1100 -7660.539 0.368 0.078
Chain 1: 1200 -7615.633 0.135 0.038
Chain 1: 1300 -7697.159 0.037 0.027
Chain 1: 1400 -7850.337 0.038 0.027
Chain 1: 1500 -7614.266 0.033 0.027
Chain 1: 1600 -7513.986 0.027 0.020
Chain 1: 1700 -7506.250 0.017 0.019
Chain 1: 1800 -7525.740 0.015 0.013
Chain 1: 1900 -7595.824 0.012 0.011
Chain 1: 2000 -7586.002 0.010 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002745 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 27.45 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86311.603 1.000 1.000
Chain 1: 200 -12997.604 3.320 5.641
Chain 1: 300 -9495.651 2.336 1.000
Chain 1: 400 -10308.609 1.772 1.000
Chain 1: 500 -8342.875 1.465 0.369
Chain 1: 600 -8115.573 1.225 0.369
Chain 1: 700 -8213.876 1.052 0.236
Chain 1: 800 -8377.314 0.923 0.236
Chain 1: 900 -8383.057 0.820 0.079
Chain 1: 1000 -8118.475 0.742 0.079
Chain 1: 1100 -8429.893 0.645 0.037 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8216.314 0.084 0.033
Chain 1: 1300 -8277.533 0.048 0.028
Chain 1: 1400 -8257.640 0.040 0.026
Chain 1: 1500 -8171.232 0.018 0.020
Chain 1: 1600 -8253.656 0.016 0.012
Chain 1: 1700 -8357.755 0.016 0.012
Chain 1: 1800 -7980.525 0.019 0.012
Chain 1: 1900 -8077.933 0.020 0.012
Chain 1: 2000 -8048.744 0.017 0.012
Chain 1: 2100 -8195.111 0.015 0.012
Chain 1: 2200 -7972.030 0.015 0.012
Chain 1: 2300 -8115.184 0.016 0.012
Chain 1: 2400 -8114.905 0.016 0.012
Chain 1: 2500 -8085.805 0.015 0.012
Chain 1: 2600 -8078.720 0.014 0.012
Chain 1: 2700 -7989.914 0.014 0.012
Chain 1: 2800 -7976.036 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002613 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 26.13 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8393463.066 1.000 1.000
Chain 1: 200 -1583137.560 2.651 4.302
Chain 1: 300 -890756.902 2.026 1.000
Chain 1: 400 -457156.149 1.757 1.000
Chain 1: 500 -357578.951 1.461 0.948
Chain 1: 600 -232544.851 1.307 0.948
Chain 1: 700 -118740.515 1.257 0.948
Chain 1: 800 -85905.557 1.148 0.948
Chain 1: 900 -66243.581 1.053 0.777
Chain 1: 1000 -51022.081 0.978 0.777
Chain 1: 1100 -38491.052 0.911 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37658.576 0.483 0.382
Chain 1: 1300 -25627.008 0.452 0.382
Chain 1: 1400 -25342.490 0.358 0.326
Chain 1: 1500 -21933.037 0.346 0.326
Chain 1: 1600 -21148.906 0.296 0.298
Chain 1: 1700 -20025.366 0.205 0.297
Chain 1: 1800 -19969.402 0.167 0.155
Chain 1: 1900 -20294.703 0.139 0.056
Chain 1: 2000 -18808.626 0.117 0.056
Chain 1: 2100 -19046.889 0.086 0.037
Chain 1: 2200 -19272.501 0.085 0.037
Chain 1: 2300 -18890.646 0.040 0.020
Chain 1: 2400 -18663.083 0.040 0.020
Chain 1: 2500 -18464.964 0.026 0.016
Chain 1: 2600 -18096.236 0.024 0.016
Chain 1: 2700 -18053.475 0.019 0.013
Chain 1: 2800 -17770.698 0.020 0.016
Chain 1: 2900 -18051.500 0.020 0.016
Chain 1: 3000 -18037.804 0.012 0.013
Chain 1: 3100 -18122.643 0.011 0.012
Chain 1: 3200 -17813.959 0.012 0.016
Chain 1: 3300 -18018.160 0.011 0.012
Chain 1: 3400 -17494.170 0.013 0.016
Chain 1: 3500 -18104.414 0.015 0.016
Chain 1: 3600 -17413.225 0.017 0.016
Chain 1: 3700 -17798.453 0.019 0.017
Chain 1: 3800 -16761.438 0.024 0.022
Chain 1: 3900 -16757.640 0.022 0.022
Chain 1: 4000 -16874.952 0.023 0.022
Chain 1: 4100 -16788.893 0.023 0.022
Chain 1: 4200 -16605.820 0.022 0.022
Chain 1: 4300 -16743.748 0.022 0.022
Chain 1: 4400 -16701.179 0.019 0.011
Chain 1: 4500 -16603.789 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001354 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.54 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12057.109 1.000 1.000
Chain 1: 200 -8998.119 0.670 1.000
Chain 1: 300 -8009.409 0.488 0.340
Chain 1: 400 -8108.971 0.369 0.340
Chain 1: 500 -7939.926 0.299 0.123
Chain 1: 600 -7824.512 0.252 0.123
Chain 1: 700 -7754.034 0.217 0.021
Chain 1: 800 -7721.998 0.191 0.021
Chain 1: 900 -7759.948 0.170 0.015
Chain 1: 1000 -7815.690 0.154 0.015
Chain 1: 1100 -7891.608 0.055 0.012
Chain 1: 1200 -7784.793 0.022 0.012
Chain 1: 1300 -7748.117 0.010 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001463 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.63 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -51037.481 1.000 1.000
Chain 1: 200 -15925.199 1.602 2.205
Chain 1: 300 -8572.917 1.354 1.000
Chain 1: 400 -8263.243 1.025 1.000
Chain 1: 500 -8776.694 0.832 0.858
Chain 1: 600 -8732.350 0.694 0.858
Chain 1: 700 -7786.804 0.612 0.121
Chain 1: 800 -8132.560 0.541 0.121
Chain 1: 900 -7636.720 0.488 0.065
Chain 1: 1000 -7687.287 0.440 0.065
Chain 1: 1100 -7659.269 0.340 0.059
Chain 1: 1200 -7588.945 0.121 0.043
Chain 1: 1300 -7720.710 0.037 0.037
Chain 1: 1400 -7837.391 0.034 0.017
Chain 1: 1500 -7614.210 0.031 0.017
Chain 1: 1600 -7537.269 0.032 0.017
Chain 1: 1700 -7520.744 0.020 0.015
Chain 1: 1800 -7561.310 0.016 0.010
Chain 1: 1900 -7546.510 0.010 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002883 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 28.83 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85966.147 1.000 1.000
Chain 1: 200 -13150.652 3.269 5.537
Chain 1: 300 -9626.928 2.301 1.000
Chain 1: 400 -10513.991 1.747 1.000
Chain 1: 500 -8535.221 1.444 0.366
Chain 1: 600 -8185.230 1.210 0.366
Chain 1: 700 -8560.722 1.044 0.232
Chain 1: 800 -8974.543 0.919 0.232
Chain 1: 900 -8479.603 0.823 0.084
Chain 1: 1000 -8224.680 0.744 0.084
Chain 1: 1100 -8469.770 0.647 0.058 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8183.365 0.097 0.046
Chain 1: 1300 -8380.072 0.063 0.044
Chain 1: 1400 -8371.674 0.054 0.043
Chain 1: 1500 -8271.969 0.032 0.035
Chain 1: 1600 -8361.449 0.029 0.031
Chain 1: 1700 -8457.562 0.026 0.029
Chain 1: 1800 -8071.255 0.026 0.029
Chain 1: 1900 -8172.728 0.021 0.023
Chain 1: 2000 -8142.442 0.019 0.012
Chain 1: 2100 -8281.127 0.017 0.012
Chain 1: 2200 -8062.691 0.017 0.012
Chain 1: 2300 -8204.858 0.016 0.012
Chain 1: 2400 -8214.653 0.016 0.012
Chain 1: 2500 -8179.806 0.015 0.012
Chain 1: 2600 -8177.378 0.014 0.012
Chain 1: 2700 -8087.485 0.014 0.012
Chain 1: 2800 -8067.783 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002595 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.95 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8400927.430 1.000 1.000
Chain 1: 200 -1584685.464 2.651 4.301
Chain 1: 300 -890396.015 2.027 1.000
Chain 1: 400 -456900.729 1.757 1.000
Chain 1: 500 -357152.499 1.462 0.949
Chain 1: 600 -232261.425 1.308 0.949
Chain 1: 700 -118673.618 1.258 0.949
Chain 1: 800 -85923.688 1.148 0.949
Chain 1: 900 -66296.827 1.053 0.780
Chain 1: 1000 -51116.095 0.978 0.780
Chain 1: 1100 -38616.975 0.910 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37791.575 0.482 0.381
Chain 1: 1300 -25784.568 0.451 0.381
Chain 1: 1400 -25504.231 0.357 0.324
Chain 1: 1500 -22101.392 0.345 0.324
Chain 1: 1600 -21319.744 0.294 0.297
Chain 1: 1700 -20198.590 0.204 0.296
Chain 1: 1800 -20143.603 0.166 0.154
Chain 1: 1900 -20469.063 0.138 0.056
Chain 1: 2000 -18984.098 0.117 0.056
Chain 1: 2100 -19222.248 0.085 0.037
Chain 1: 2200 -19447.792 0.084 0.037
Chain 1: 2300 -19065.984 0.040 0.020
Chain 1: 2400 -18838.355 0.040 0.020
Chain 1: 2500 -18640.197 0.026 0.016
Chain 1: 2600 -18271.259 0.024 0.016
Chain 1: 2700 -18228.508 0.019 0.012
Chain 1: 2800 -17945.567 0.020 0.016
Chain 1: 2900 -18226.484 0.020 0.015
Chain 1: 3000 -18212.761 0.012 0.012
Chain 1: 3100 -18297.618 0.011 0.012
Chain 1: 3200 -17988.804 0.012 0.015
Chain 1: 3300 -18193.142 0.011 0.012
Chain 1: 3400 -17668.889 0.013 0.015
Chain 1: 3500 -18279.465 0.015 0.016
Chain 1: 3600 -17587.852 0.017 0.016
Chain 1: 3700 -17973.347 0.019 0.017
Chain 1: 3800 -16935.653 0.023 0.021
Chain 1: 3900 -16931.841 0.022 0.021
Chain 1: 4000 -17049.166 0.023 0.021
Chain 1: 4100 -16963.025 0.023 0.021
Chain 1: 4200 -16779.860 0.022 0.021
Chain 1: 4300 -16917.861 0.022 0.021
Chain 1: 4400 -16875.147 0.019 0.011
Chain 1: 4500 -16777.750 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00149 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.9 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12082.766 1.000 1.000
Chain 1: 200 -9041.371 0.668 1.000
Chain 1: 300 -7969.991 0.490 0.336
Chain 1: 400 -8090.925 0.371 0.336
Chain 1: 500 -8084.578 0.297 0.134
Chain 1: 600 -8013.054 0.249 0.134
Chain 1: 700 -7811.434 0.217 0.026
Chain 1: 800 -7797.264 0.190 0.026
Chain 1: 900 -7788.946 0.169 0.015
Chain 1: 1000 -7819.623 0.153 0.015
Chain 1: 1100 -7867.908 0.053 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001695 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 16.95 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -46062.459 1.000 1.000
Chain 1: 200 -15240.953 1.511 2.022
Chain 1: 300 -8537.239 1.269 1.000
Chain 1: 400 -8361.725 0.957 1.000
Chain 1: 500 -8172.016 0.770 0.785
Chain 1: 600 -7828.134 0.649 0.785
Chain 1: 700 -7715.804 0.559 0.044
Chain 1: 800 -7996.641 0.493 0.044
Chain 1: 900 -7916.659 0.439 0.035
Chain 1: 1000 -7709.941 0.398 0.035
Chain 1: 1100 -7717.591 0.298 0.027
Chain 1: 1200 -7633.719 0.097 0.023
Chain 1: 1300 -7578.028 0.019 0.021
Chain 1: 1400 -7891.340 0.021 0.023
Chain 1: 1500 -7591.776 0.023 0.027
Chain 1: 1600 -7510.192 0.020 0.015
Chain 1: 1700 -7492.152 0.018 0.011
Chain 1: 1800 -7534.187 0.015 0.011
Chain 1: 1900 -7563.364 0.015 0.011
Chain 1: 2000 -7574.161 0.012 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002607 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 26.07 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86165.213 1.000 1.000
Chain 1: 200 -13163.473 3.273 5.546
Chain 1: 300 -9618.596 2.305 1.000
Chain 1: 400 -10451.286 1.748 1.000
Chain 1: 500 -8550.629 1.443 0.369
Chain 1: 600 -8202.280 1.210 0.369
Chain 1: 700 -8411.615 1.041 0.222
Chain 1: 800 -8971.172 0.918 0.222
Chain 1: 900 -8457.265 0.823 0.080
Chain 1: 1000 -8213.587 0.744 0.080
Chain 1: 1100 -8454.122 0.646 0.062 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8158.763 0.096 0.061
Chain 1: 1300 -8374.196 0.061 0.042
Chain 1: 1400 -8366.919 0.053 0.036
Chain 1: 1500 -8269.836 0.032 0.030
Chain 1: 1600 -8362.788 0.029 0.028
Chain 1: 1700 -8463.816 0.028 0.028
Chain 1: 1800 -8073.954 0.026 0.028
Chain 1: 1900 -8175.568 0.022 0.026
Chain 1: 2000 -8145.518 0.019 0.012
Chain 1: 2100 -8283.405 0.018 0.012
Chain 1: 2200 -8065.411 0.017 0.012
Chain 1: 2300 -8207.345 0.016 0.012
Chain 1: 2400 -8214.846 0.016 0.012
Chain 1: 2500 -8184.315 0.015 0.012
Chain 1: 2600 -8180.603 0.014 0.012
Chain 1: 2700 -8091.272 0.014 0.012
Chain 1: 2800 -8070.891 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003338 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.38 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8400113.823 1.000 1.000
Chain 1: 200 -1585369.166 2.649 4.299
Chain 1: 300 -891808.822 2.025 1.000
Chain 1: 400 -457658.558 1.756 1.000
Chain 1: 500 -357962.677 1.461 0.949
Chain 1: 600 -232828.096 1.307 0.949
Chain 1: 700 -118984.221 1.257 0.949
Chain 1: 800 -86127.434 1.147 0.949
Chain 1: 900 -66458.952 1.053 0.778
Chain 1: 1000 -51233.627 0.977 0.778
Chain 1: 1100 -38692.559 0.910 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37862.979 0.482 0.381
Chain 1: 1300 -25818.779 0.451 0.381
Chain 1: 1400 -25534.571 0.357 0.324
Chain 1: 1500 -22121.384 0.345 0.324
Chain 1: 1600 -21336.313 0.295 0.297
Chain 1: 1700 -20211.234 0.205 0.296
Chain 1: 1800 -20155.107 0.167 0.154
Chain 1: 1900 -20480.624 0.139 0.056
Chain 1: 2000 -18993.221 0.117 0.056
Chain 1: 2100 -19231.631 0.086 0.037
Chain 1: 2200 -19457.451 0.085 0.037
Chain 1: 2300 -19075.372 0.040 0.020
Chain 1: 2400 -18847.695 0.040 0.020
Chain 1: 2500 -18649.567 0.026 0.016
Chain 1: 2600 -18280.587 0.024 0.016
Chain 1: 2700 -18237.777 0.019 0.012
Chain 1: 2800 -17954.836 0.020 0.016
Chain 1: 2900 -18235.838 0.020 0.015
Chain 1: 3000 -18222.113 0.012 0.012
Chain 1: 3100 -18306.963 0.011 0.012
Chain 1: 3200 -17998.121 0.012 0.015
Chain 1: 3300 -18202.471 0.011 0.012
Chain 1: 3400 -17678.150 0.013 0.015
Chain 1: 3500 -18288.847 0.015 0.016
Chain 1: 3600 -17597.135 0.017 0.016
Chain 1: 3700 -17982.741 0.019 0.017
Chain 1: 3800 -16944.845 0.023 0.021
Chain 1: 3900 -16941.036 0.022 0.021
Chain 1: 4000 -17058.361 0.023 0.021
Chain 1: 4100 -16972.204 0.023 0.021
Chain 1: 4200 -16788.984 0.022 0.021
Chain 1: 4300 -16927.025 0.022 0.021
Chain 1: 4400 -16884.302 0.019 0.011
Chain 1: 4500 -16786.889 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00134 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.4 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13771.249 1.000 1.000
Chain 1: 200 -10364.695 0.664 1.000
Chain 1: 300 -8453.798 0.518 0.329
Chain 1: 400 -8171.127 0.397 0.329
Chain 1: 500 -8333.093 0.322 0.226
Chain 1: 600 -8168.957 0.271 0.226
Chain 1: 700 -8114.731 0.234 0.035
Chain 1: 800 -8034.060 0.206 0.035
Chain 1: 900 -8363.876 0.187 0.035
Chain 1: 1000 -8104.059 0.172 0.035
Chain 1: 1100 -8189.562 0.073 0.032
Chain 1: 1200 -8097.546 0.041 0.020
Chain 1: 1300 -8007.305 0.020 0.019
Chain 1: 1400 -8038.377 0.016 0.011
Chain 1: 1500 -8127.239 0.016 0.011
Chain 1: 1600 -8036.263 0.015 0.011
Chain 1: 1700 -8011.302 0.014 0.011
Chain 1: 1800 -7978.390 0.014 0.011
Chain 1: 1900 -8006.992 0.010 0.011
Chain 1: 2000 -7942.579 0.008 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00137 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.7 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -51600.287 1.000 1.000
Chain 1: 200 -17243.195 1.496 1.993
Chain 1: 300 -9024.509 1.301 1.000
Chain 1: 400 -9936.188 0.999 1.000
Chain 1: 500 -8407.029 0.835 0.911
Chain 1: 600 -8739.131 0.702 0.911
Chain 1: 700 -9275.958 0.610 0.182
Chain 1: 800 -8229.047 0.550 0.182
Chain 1: 900 -8483.612 0.492 0.127
Chain 1: 1000 -8441.159 0.443 0.127
Chain 1: 1100 -7827.729 0.351 0.092
Chain 1: 1200 -7844.795 0.152 0.078
Chain 1: 1300 -7927.348 0.062 0.058
Chain 1: 1400 -7750.931 0.055 0.038
Chain 1: 1500 -7591.835 0.039 0.030
Chain 1: 1600 -7779.232 0.038 0.024
Chain 1: 1700 -7723.452 0.033 0.023
Chain 1: 1800 -7634.119 0.021 0.021
Chain 1: 1900 -7672.240 0.019 0.012
Chain 1: 2000 -7788.267 0.020 0.015
Chain 1: 2100 -7622.027 0.014 0.015
Chain 1: 2200 -7922.659 0.018 0.021
Chain 1: 2300 -7738.259 0.019 0.022
Chain 1: 2400 -7658.060 0.018 0.021
Chain 1: 2500 -7696.005 0.016 0.015
Chain 1: 2600 -7644.106 0.014 0.012
Chain 1: 2700 -7484.963 0.016 0.015
Chain 1: 2800 -7576.894 0.016 0.015
Chain 1: 2900 -7421.889 0.017 0.021
Chain 1: 3000 -7593.694 0.018 0.021
Chain 1: 3100 -7573.566 0.016 0.021
Chain 1: 3200 -7807.557 0.016 0.021
Chain 1: 3300 -7453.050 0.018 0.021
Chain 1: 3400 -7718.984 0.020 0.021
Chain 1: 3500 -7493.322 0.023 0.023
Chain 1: 3600 -7563.289 0.023 0.023
Chain 1: 3700 -7505.999 0.022 0.023
Chain 1: 3800 -7490.380 0.021 0.023
Chain 1: 3900 -7458.657 0.019 0.023
Chain 1: 4000 -7446.072 0.017 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003061 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 30.61 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86541.447 1.000 1.000
Chain 1: 200 -14278.863 3.030 5.061
Chain 1: 300 -10368.385 2.146 1.000
Chain 1: 400 -12669.155 1.655 1.000
Chain 1: 500 -8836.895 1.411 0.434
Chain 1: 600 -8933.920 1.177 0.434
Chain 1: 700 -8589.136 1.015 0.377
Chain 1: 800 -9352.298 0.898 0.377
Chain 1: 900 -8822.294 0.805 0.182
Chain 1: 1000 -9123.191 0.728 0.182
Chain 1: 1100 -9079.667 0.628 0.082 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8452.989 0.130 0.074
Chain 1: 1300 -8915.114 0.097 0.060
Chain 1: 1400 -8830.917 0.080 0.052
Chain 1: 1500 -8763.700 0.037 0.040
Chain 1: 1600 -8801.428 0.037 0.040
Chain 1: 1700 -8895.869 0.034 0.033
Chain 1: 1800 -8398.810 0.032 0.033
Chain 1: 1900 -8527.423 0.027 0.015
Chain 1: 2000 -8534.944 0.024 0.011
Chain 1: 2100 -8677.590 0.025 0.015
Chain 1: 2200 -8395.057 0.021 0.015
Chain 1: 2300 -8483.073 0.017 0.011
Chain 1: 2400 -8576.404 0.017 0.011
Chain 1: 2500 -8477.712 0.017 0.012
Chain 1: 2600 -8528.979 0.017 0.012
Chain 1: 2700 -8432.875 0.018 0.012
Chain 1: 2800 -8397.204 0.012 0.011
Chain 1: 2900 -8487.551 0.012 0.011
Chain 1: 3000 -8414.426 0.012 0.011
Chain 1: 3100 -8371.090 0.011 0.011
Chain 1: 3200 -8331.301 0.008 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003278 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.78 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8396892.040 1.000 1.000
Chain 1: 200 -1584512.618 2.650 4.299
Chain 1: 300 -892231.822 2.025 1.000
Chain 1: 400 -459284.129 1.754 1.000
Chain 1: 500 -359683.993 1.459 0.943
Chain 1: 600 -234401.412 1.305 0.943
Chain 1: 700 -120358.123 1.254 0.943
Chain 1: 800 -87468.894 1.144 0.943
Chain 1: 900 -67766.737 1.049 0.776
Chain 1: 1000 -52545.032 0.973 0.776
Chain 1: 1100 -39988.342 0.905 0.534 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -39172.776 0.477 0.376
Chain 1: 1300 -27078.223 0.444 0.376
Chain 1: 1400 -26798.244 0.351 0.314
Chain 1: 1500 -23370.621 0.338 0.314
Chain 1: 1600 -22584.290 0.288 0.291
Chain 1: 1700 -21450.947 0.198 0.290
Chain 1: 1800 -21394.213 0.161 0.147
Chain 1: 1900 -21721.615 0.133 0.053
Chain 1: 2000 -20226.451 0.112 0.053
Chain 1: 2100 -20465.340 0.082 0.035
Chain 1: 2200 -20693.176 0.081 0.035
Chain 1: 2300 -20308.819 0.038 0.019
Chain 1: 2400 -20080.379 0.038 0.019
Chain 1: 2500 -19882.423 0.024 0.015
Chain 1: 2600 -19511.258 0.023 0.015
Chain 1: 2700 -19467.816 0.018 0.012
Chain 1: 2800 -19184.081 0.019 0.015
Chain 1: 2900 -19466.012 0.019 0.014
Chain 1: 3000 -19452.149 0.011 0.012
Chain 1: 3100 -19537.311 0.011 0.011
Chain 1: 3200 -19227.123 0.011 0.014
Chain 1: 3300 -19432.512 0.010 0.011
Chain 1: 3400 -18905.859 0.012 0.014
Chain 1: 3500 -19520.137 0.014 0.015
Chain 1: 3600 -18823.730 0.016 0.015
Chain 1: 3700 -19212.850 0.018 0.016
Chain 1: 3800 -18167.719 0.022 0.020
Chain 1: 3900 -18163.736 0.021 0.020
Chain 1: 4000 -18281.065 0.021 0.020
Chain 1: 4100 -18194.578 0.021 0.020
Chain 1: 4200 -18009.750 0.021 0.020
Chain 1: 4300 -18148.906 0.020 0.020
Chain 1: 4400 -18104.876 0.018 0.010
Chain 1: 4500 -18007.254 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001215 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.15 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49343.852 1.000 1.000
Chain 1: 200 -15883.634 1.553 2.107
Chain 1: 300 -19119.689 1.092 1.000
Chain 1: 400 -16115.396 0.866 1.000
Chain 1: 500 -17940.928 0.713 0.186
Chain 1: 600 -15437.674 0.621 0.186
Chain 1: 700 -16847.620 0.544 0.169
Chain 1: 800 -14058.442 0.501 0.186
Chain 1: 900 -11976.191 0.465 0.174
Chain 1: 1000 -13739.196 0.431 0.174
Chain 1: 1100 -24076.034 0.374 0.174
Chain 1: 1200 -13940.069 0.236 0.174
Chain 1: 1300 -12918.932 0.227 0.174
Chain 1: 1400 -13065.847 0.209 0.162
Chain 1: 1500 -11947.709 0.209 0.162
Chain 1: 1600 -12687.404 0.198 0.128
Chain 1: 1700 -12925.501 0.192 0.128
Chain 1: 1800 -26931.827 0.224 0.128
Chain 1: 1900 -11288.976 0.345 0.128
Chain 1: 2000 -10197.428 0.343 0.107
Chain 1: 2100 -10354.559 0.302 0.094
Chain 1: 2200 -9865.067 0.234 0.079
Chain 1: 2300 -9656.232 0.228 0.058
Chain 1: 2400 -10131.636 0.232 0.058
Chain 1: 2500 -15432.557 0.257 0.058
Chain 1: 2600 -10046.123 0.304 0.107
Chain 1: 2700 -10116.562 0.303 0.107
Chain 1: 2800 -11851.407 0.266 0.107
Chain 1: 2900 -16632.971 0.156 0.107
Chain 1: 3000 -10062.158 0.211 0.146
Chain 1: 3100 -10095.140 0.209 0.146
Chain 1: 3200 -9994.638 0.206 0.146
Chain 1: 3300 -13775.606 0.231 0.274
Chain 1: 3400 -9355.143 0.273 0.287
Chain 1: 3500 -14478.663 0.274 0.287
Chain 1: 3600 -10458.926 0.259 0.287
Chain 1: 3700 -12017.265 0.272 0.287
Chain 1: 3800 -8939.842 0.291 0.344
Chain 1: 3900 -9570.309 0.269 0.344
Chain 1: 4000 -11843.971 0.223 0.274
Chain 1: 4100 -9376.490 0.249 0.274
Chain 1: 4200 -14211.598 0.282 0.340
Chain 1: 4300 -9138.387 0.310 0.344
Chain 1: 4400 -8769.158 0.267 0.340
Chain 1: 4500 -9144.818 0.236 0.263
Chain 1: 4600 -14362.491 0.234 0.263
Chain 1: 4700 -9362.424 0.274 0.340
Chain 1: 4800 -9601.823 0.242 0.263
Chain 1: 4900 -10058.859 0.240 0.263
Chain 1: 5000 -9964.250 0.222 0.263
Chain 1: 5100 -8829.394 0.208 0.129
Chain 1: 5200 -9667.903 0.183 0.087
Chain 1: 5300 -12148.545 0.148 0.087
Chain 1: 5400 -8777.873 0.182 0.129
Chain 1: 5500 -9325.404 0.184 0.129
Chain 1: 5600 -10349.004 0.157 0.099
Chain 1: 5700 -9618.437 0.112 0.087
Chain 1: 5800 -8971.563 0.116 0.087
Chain 1: 5900 -14382.632 0.149 0.099
Chain 1: 6000 -11724.003 0.171 0.129
Chain 1: 6100 -8933.668 0.190 0.204
Chain 1: 6200 -9335.923 0.185 0.204
Chain 1: 6300 -13141.841 0.194 0.227
Chain 1: 6400 -9533.185 0.193 0.227
Chain 1: 6500 -9984.148 0.192 0.227
Chain 1: 6600 -12326.999 0.201 0.227
Chain 1: 6700 -8578.432 0.237 0.290
Chain 1: 6800 -9285.290 0.237 0.290
Chain 1: 6900 -13803.459 0.233 0.290
Chain 1: 7000 -13560.797 0.212 0.290
Chain 1: 7100 -8915.140 0.233 0.290
Chain 1: 7200 -9716.580 0.237 0.290
Chain 1: 7300 -10941.112 0.219 0.190
Chain 1: 7400 -8486.319 0.210 0.190
Chain 1: 7500 -8794.484 0.209 0.190
Chain 1: 7600 -12048.348 0.217 0.270
Chain 1: 7700 -8898.915 0.209 0.270
Chain 1: 7800 -12406.961 0.229 0.283
Chain 1: 7900 -8539.513 0.242 0.283
Chain 1: 8000 -8641.982 0.241 0.283
Chain 1: 8100 -8622.233 0.189 0.270
Chain 1: 8200 -8681.963 0.182 0.270
Chain 1: 8300 -8833.029 0.172 0.270
Chain 1: 8400 -10494.930 0.159 0.158
Chain 1: 8500 -8568.017 0.178 0.225
Chain 1: 8600 -9103.547 0.157 0.158
Chain 1: 8700 -10014.094 0.131 0.091
Chain 1: 8800 -9468.261 0.108 0.059
Chain 1: 8900 -9666.301 0.065 0.058
Chain 1: 9000 -11577.661 0.080 0.059
Chain 1: 9100 -8463.517 0.117 0.091
Chain 1: 9200 -9392.369 0.126 0.099
Chain 1: 9300 -8448.085 0.135 0.112
Chain 1: 9400 -8603.430 0.121 0.099
Chain 1: 9500 -13162.375 0.134 0.099
Chain 1: 9600 -8969.740 0.174 0.112
Chain 1: 9700 -8609.310 0.170 0.112
Chain 1: 9800 -8837.691 0.166 0.112
Chain 1: 9900 -10695.216 0.182 0.165
Chain 1: 10000 -8615.614 0.189 0.174
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001504 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.04 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57675.395 1.000 1.000
Chain 1: 200 -18073.104 1.596 2.191
Chain 1: 300 -8968.101 1.402 1.015
Chain 1: 400 -8171.898 1.076 1.015
Chain 1: 500 -8848.659 0.876 1.000
Chain 1: 600 -8268.051 0.742 1.000
Chain 1: 700 -8687.822 0.643 0.097
Chain 1: 800 -8158.093 0.570 0.097
Chain 1: 900 -7918.217 0.510 0.076
Chain 1: 1000 -8008.376 0.461 0.076
Chain 1: 1100 -7880.106 0.362 0.070
Chain 1: 1200 -7749.875 0.145 0.065
Chain 1: 1300 -7996.544 0.046 0.048
Chain 1: 1400 -7864.739 0.038 0.031
Chain 1: 1500 -7525.386 0.035 0.031
Chain 1: 1600 -7726.495 0.031 0.030
Chain 1: 1700 -7575.220 0.028 0.026
Chain 1: 1800 -7549.913 0.022 0.020
Chain 1: 1900 -7551.956 0.019 0.017
Chain 1: 2000 -7691.014 0.019 0.018
Chain 1: 2100 -7558.379 0.019 0.018
Chain 1: 2200 -7781.248 0.021 0.020
Chain 1: 2300 -7599.175 0.020 0.020
Chain 1: 2400 -7672.703 0.019 0.020
Chain 1: 2500 -7593.183 0.016 0.018
Chain 1: 2600 -7531.055 0.014 0.018
Chain 1: 2700 -7553.139 0.012 0.010
Chain 1: 2800 -7635.468 0.013 0.011
Chain 1: 2900 -7368.968 0.017 0.018
Chain 1: 3000 -7536.795 0.017 0.018
Chain 1: 3100 -7515.946 0.016 0.011
Chain 1: 3200 -7740.638 0.016 0.011
Chain 1: 3300 -7434.171 0.017 0.011
Chain 1: 3400 -7692.708 0.020 0.022
Chain 1: 3500 -7436.163 0.022 0.029
Chain 1: 3600 -7490.840 0.022 0.029
Chain 1: 3700 -7449.393 0.022 0.029
Chain 1: 3800 -7446.819 0.021 0.029
Chain 1: 3900 -7399.517 0.018 0.022
Chain 1: 4000 -7392.674 0.016 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003112 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.12 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87279.244 1.000 1.000
Chain 1: 200 -14030.381 3.110 5.221
Chain 1: 300 -10276.303 2.195 1.000
Chain 1: 400 -11959.135 1.682 1.000
Chain 1: 500 -8997.118 1.411 0.365
Chain 1: 600 -9042.090 1.177 0.365
Chain 1: 700 -9426.508 1.015 0.329
Chain 1: 800 -8531.057 0.901 0.329
Chain 1: 900 -8554.212 0.801 0.141
Chain 1: 1000 -9238.199 0.728 0.141
Chain 1: 1100 -8790.823 0.633 0.105 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -9177.863 0.116 0.074
Chain 1: 1300 -8577.895 0.086 0.070
Chain 1: 1400 -8735.248 0.074 0.051
Chain 1: 1500 -8655.973 0.042 0.042
Chain 1: 1600 -8666.032 0.041 0.042
Chain 1: 1700 -8545.527 0.039 0.042
Chain 1: 1800 -8601.908 0.029 0.018
Chain 1: 1900 -8449.074 0.030 0.018
Chain 1: 2000 -8551.178 0.024 0.018
Chain 1: 2100 -8535.598 0.019 0.014
Chain 1: 2200 -8493.446 0.016 0.012
Chain 1: 2300 -8653.039 0.010 0.012
Chain 1: 2400 -8470.550 0.011 0.012
Chain 1: 2500 -8544.017 0.011 0.012
Chain 1: 2600 -8449.654 0.012 0.012
Chain 1: 2700 -8493.888 0.011 0.011
Chain 1: 2800 -8445.787 0.011 0.011
Chain 1: 2900 -8553.458 0.010 0.011
Chain 1: 3000 -8504.744 0.010 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003195 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.95 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8417180.685 1.000 1.000
Chain 1: 200 -1587649.873 2.651 4.302
Chain 1: 300 -891653.790 2.027 1.000
Chain 1: 400 -458173.848 1.757 1.000
Chain 1: 500 -358326.034 1.461 0.946
Chain 1: 600 -233130.757 1.307 0.946
Chain 1: 700 -119540.728 1.256 0.946
Chain 1: 800 -86814.290 1.146 0.946
Chain 1: 900 -67208.264 1.051 0.781
Chain 1: 1000 -52060.095 0.975 0.781
Chain 1: 1100 -39574.922 0.907 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38762.840 0.479 0.377
Chain 1: 1300 -26743.480 0.446 0.377
Chain 1: 1400 -26467.758 0.352 0.315
Chain 1: 1500 -23061.450 0.339 0.315
Chain 1: 1600 -22280.938 0.289 0.292
Chain 1: 1700 -21156.753 0.199 0.291
Chain 1: 1800 -21101.856 0.162 0.148
Chain 1: 1900 -21428.695 0.134 0.053
Chain 1: 2000 -19939.784 0.112 0.053
Chain 1: 2100 -20178.145 0.082 0.035
Chain 1: 2200 -20404.984 0.081 0.035
Chain 1: 2300 -20021.708 0.038 0.019
Chain 1: 2400 -19793.605 0.038 0.019
Chain 1: 2500 -19595.537 0.024 0.015
Chain 1: 2600 -19225.145 0.023 0.015
Chain 1: 2700 -19181.930 0.018 0.012
Chain 1: 2800 -18898.484 0.019 0.015
Chain 1: 2900 -19179.999 0.019 0.015
Chain 1: 3000 -19166.134 0.012 0.012
Chain 1: 3100 -19251.242 0.011 0.012
Chain 1: 3200 -18941.485 0.011 0.015
Chain 1: 3300 -19146.567 0.011 0.012
Chain 1: 3400 -18620.685 0.012 0.015
Chain 1: 3500 -19233.712 0.014 0.015
Chain 1: 3600 -18538.901 0.016 0.015
Chain 1: 3700 -18926.799 0.018 0.016
Chain 1: 3800 -17884.156 0.022 0.020
Chain 1: 3900 -17880.236 0.021 0.020
Chain 1: 4000 -17997.555 0.021 0.020
Chain 1: 4100 -17911.207 0.022 0.020
Chain 1: 4200 -17726.934 0.021 0.020
Chain 1: 4300 -17865.703 0.021 0.020
Chain 1: 4400 -17822.101 0.018 0.010
Chain 1: 4500 -17724.551 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001368 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.68 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49004.212 1.000 1.000
Chain 1: 200 -19242.838 1.273 1.547
Chain 1: 300 -20360.643 0.867 1.000
Chain 1: 400 -15330.297 0.732 1.000
Chain 1: 500 -17912.100 0.615 0.328
Chain 1: 600 -20243.131 0.531 0.328
Chain 1: 700 -22566.289 0.470 0.144
Chain 1: 800 -11567.873 0.530 0.328
Chain 1: 900 -17447.855 0.509 0.328
Chain 1: 1000 -10978.926 0.517 0.337
Chain 1: 1100 -10058.914 0.426 0.328
Chain 1: 1200 -12679.785 0.292 0.207
Chain 1: 1300 -16877.846 0.311 0.249
Chain 1: 1400 -11851.478 0.321 0.249
Chain 1: 1500 -10898.880 0.315 0.249
Chain 1: 1600 -10324.295 0.309 0.249
Chain 1: 1700 -11100.865 0.306 0.249
Chain 1: 1800 -18111.209 0.250 0.249
Chain 1: 1900 -9589.629 0.305 0.249
Chain 1: 2000 -14058.604 0.278 0.249
Chain 1: 2100 -9643.227 0.314 0.318
Chain 1: 2200 -16062.755 0.334 0.387
Chain 1: 2300 -9070.154 0.386 0.400
Chain 1: 2400 -11660.898 0.366 0.387
Chain 1: 2500 -10701.096 0.366 0.387
Chain 1: 2600 -9233.425 0.376 0.387
Chain 1: 2700 -10696.270 0.383 0.387
Chain 1: 2800 -9325.928 0.359 0.318
Chain 1: 2900 -9301.848 0.270 0.222
Chain 1: 3000 -8802.987 0.244 0.159
Chain 1: 3100 -9077.999 0.201 0.147
Chain 1: 3200 -13801.104 0.196 0.147
Chain 1: 3300 -12016.979 0.133 0.147
Chain 1: 3400 -8810.983 0.148 0.147
Chain 1: 3500 -16115.483 0.184 0.148
Chain 1: 3600 -10344.358 0.224 0.148
Chain 1: 3700 -10403.966 0.211 0.148
Chain 1: 3800 -9037.764 0.211 0.151
Chain 1: 3900 -9426.349 0.215 0.151
Chain 1: 4000 -10180.622 0.217 0.151
Chain 1: 4100 -9518.454 0.221 0.151
Chain 1: 4200 -11375.155 0.203 0.151
Chain 1: 4300 -9366.957 0.209 0.163
Chain 1: 4400 -9119.947 0.176 0.151
Chain 1: 4500 -16367.848 0.175 0.151
Chain 1: 4600 -9738.404 0.187 0.151
Chain 1: 4700 -13214.756 0.213 0.163
Chain 1: 4800 -10289.058 0.226 0.214
Chain 1: 4900 -9113.494 0.235 0.214
Chain 1: 5000 -14815.818 0.266 0.263
Chain 1: 5100 -12253.117 0.280 0.263
Chain 1: 5200 -8697.401 0.304 0.284
Chain 1: 5300 -10544.283 0.301 0.284
Chain 1: 5400 -10065.772 0.303 0.284
Chain 1: 5500 -10000.929 0.259 0.263
Chain 1: 5600 -9471.608 0.196 0.209
Chain 1: 5700 -8542.001 0.181 0.175
Chain 1: 5800 -8827.034 0.156 0.129
Chain 1: 5900 -14417.278 0.182 0.175
Chain 1: 6000 -8901.230 0.205 0.175
Chain 1: 6100 -10612.932 0.200 0.161
Chain 1: 6200 -10383.439 0.162 0.109
Chain 1: 6300 -12539.697 0.161 0.109
Chain 1: 6400 -11507.160 0.166 0.109
Chain 1: 6500 -9940.082 0.181 0.158
Chain 1: 6600 -8528.281 0.192 0.161
Chain 1: 6700 -8137.426 0.186 0.161
Chain 1: 6800 -8724.436 0.189 0.161
Chain 1: 6900 -10139.140 0.164 0.158
Chain 1: 7000 -12538.340 0.121 0.158
Chain 1: 7100 -9321.515 0.140 0.158
Chain 1: 7200 -8927.140 0.142 0.158
Chain 1: 7300 -11502.800 0.147 0.158
Chain 1: 7400 -10884.800 0.144 0.158
Chain 1: 7500 -8108.314 0.162 0.166
Chain 1: 7600 -8433.045 0.150 0.140
Chain 1: 7700 -8606.837 0.147 0.140
Chain 1: 7800 -11880.291 0.168 0.191
Chain 1: 7900 -8041.321 0.202 0.224
Chain 1: 8000 -8617.784 0.189 0.224
Chain 1: 8100 -10581.456 0.173 0.186
Chain 1: 8200 -8129.354 0.199 0.224
Chain 1: 8300 -11161.825 0.204 0.272
Chain 1: 8400 -7987.349 0.238 0.276
Chain 1: 8500 -9229.942 0.217 0.272
Chain 1: 8600 -10184.044 0.222 0.272
Chain 1: 8700 -11688.507 0.233 0.272
Chain 1: 8800 -8272.098 0.247 0.272
Chain 1: 8900 -8595.324 0.203 0.186
Chain 1: 9000 -10082.414 0.211 0.186
Chain 1: 9100 -8342.088 0.213 0.209
Chain 1: 9200 -8959.219 0.190 0.147
Chain 1: 9300 -9347.385 0.167 0.135
Chain 1: 9400 -8103.808 0.143 0.135
Chain 1: 9500 -8155.212 0.130 0.129
Chain 1: 9600 -10671.523 0.144 0.147
Chain 1: 9700 -8538.525 0.156 0.153
Chain 1: 9800 -8820.749 0.118 0.147
Chain 1: 9900 -11111.103 0.135 0.153
Chain 1: 10000 -8033.772 0.159 0.206
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001426 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.26 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -53059.811 1.000 1.000
Chain 1: 200 -16712.516 1.587 2.175
Chain 1: 300 -8636.816 1.370 1.000
Chain 1: 400 -8391.182 1.035 1.000
Chain 1: 500 -7993.124 0.838 0.935
Chain 1: 600 -8555.716 0.709 0.935
Chain 1: 700 -8153.260 0.615 0.066
Chain 1: 800 -8271.406 0.540 0.066
Chain 1: 900 -7985.830 0.484 0.050
Chain 1: 1000 -7872.291 0.437 0.050
Chain 1: 1100 -7717.659 0.339 0.049
Chain 1: 1200 -7614.074 0.123 0.036
Chain 1: 1300 -7631.060 0.029 0.029
Chain 1: 1400 -8010.092 0.031 0.036
Chain 1: 1500 -7654.141 0.031 0.036
Chain 1: 1600 -7792.186 0.026 0.020
Chain 1: 1700 -7574.066 0.024 0.020
Chain 1: 1800 -7629.352 0.023 0.020
Chain 1: 1900 -7636.184 0.020 0.018
Chain 1: 2000 -7642.914 0.019 0.018
Chain 1: 2100 -7632.324 0.017 0.014
Chain 1: 2200 -7734.386 0.017 0.013
Chain 1: 2300 -7849.878 0.018 0.015
Chain 1: 2400 -7674.971 0.015 0.015
Chain 1: 2500 -7649.408 0.011 0.013
Chain 1: 2600 -7562.979 0.010 0.011
Chain 1: 2700 -7622.507 0.008 0.008 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003107 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.07 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86340.907 1.000 1.000
Chain 1: 200 -13421.892 3.216 5.433
Chain 1: 300 -9756.200 2.270 1.000
Chain 1: 400 -10708.712 1.724 1.000
Chain 1: 500 -8732.464 1.425 0.376
Chain 1: 600 -8708.064 1.188 0.376
Chain 1: 700 -8475.974 1.022 0.226
Chain 1: 800 -9123.725 0.903 0.226
Chain 1: 900 -8518.268 0.811 0.089
Chain 1: 1000 -8411.794 0.731 0.089
Chain 1: 1100 -8413.134 0.631 0.071 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8183.658 0.090 0.071
Chain 1: 1300 -8426.622 0.056 0.029
Chain 1: 1400 -8443.367 0.047 0.028
Chain 1: 1500 -8293.849 0.026 0.027
Chain 1: 1600 -8407.828 0.027 0.027
Chain 1: 1700 -8483.374 0.025 0.018
Chain 1: 1800 -8059.000 0.024 0.018
Chain 1: 1900 -8160.743 0.018 0.014
Chain 1: 2000 -8135.257 0.017 0.014
Chain 1: 2100 -8261.378 0.018 0.015
Chain 1: 2200 -8062.575 0.018 0.015
Chain 1: 2300 -8155.635 0.016 0.014
Chain 1: 2400 -8224.150 0.017 0.014
Chain 1: 2500 -8170.352 0.016 0.012
Chain 1: 2600 -8172.088 0.014 0.011
Chain 1: 2700 -8088.679 0.015 0.011
Chain 1: 2800 -8048.024 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002591 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.91 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8395753.047 1.000 1.000
Chain 1: 200 -1581880.387 2.654 4.307
Chain 1: 300 -890577.148 2.028 1.000
Chain 1: 400 -457843.097 1.757 1.000
Chain 1: 500 -358207.706 1.461 0.945
Chain 1: 600 -233173.241 1.307 0.945
Chain 1: 700 -119267.213 1.257 0.945
Chain 1: 800 -86445.857 1.147 0.945
Chain 1: 900 -66763.319 1.053 0.776
Chain 1: 1000 -51547.144 0.977 0.776
Chain 1: 1100 -39010.125 0.909 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38186.222 0.480 0.380
Chain 1: 1300 -26126.351 0.449 0.380
Chain 1: 1400 -25845.111 0.355 0.321
Chain 1: 1500 -22427.934 0.343 0.321
Chain 1: 1600 -21643.419 0.293 0.295
Chain 1: 1700 -20514.934 0.203 0.295
Chain 1: 1800 -20458.762 0.165 0.152
Chain 1: 1900 -20785.007 0.137 0.055
Chain 1: 2000 -19294.866 0.115 0.055
Chain 1: 2100 -19533.342 0.085 0.036
Chain 1: 2200 -19760.010 0.084 0.036
Chain 1: 2300 -19376.997 0.039 0.020
Chain 1: 2400 -19148.993 0.039 0.020
Chain 1: 2500 -18951.079 0.025 0.016
Chain 1: 2600 -18581.059 0.024 0.016
Chain 1: 2700 -18538.000 0.018 0.012
Chain 1: 2800 -18254.796 0.020 0.016
Chain 1: 2900 -18536.154 0.020 0.015
Chain 1: 3000 -18522.332 0.012 0.012
Chain 1: 3100 -18607.320 0.011 0.012
Chain 1: 3200 -18297.900 0.012 0.015
Chain 1: 3300 -18502.713 0.011 0.012
Chain 1: 3400 -17977.488 0.013 0.015
Chain 1: 3500 -18589.599 0.015 0.016
Chain 1: 3600 -17895.981 0.017 0.016
Chain 1: 3700 -18282.995 0.019 0.017
Chain 1: 3800 -17242.240 0.023 0.021
Chain 1: 3900 -17238.370 0.022 0.021
Chain 1: 4000 -17355.674 0.022 0.021
Chain 1: 4100 -17269.397 0.022 0.021
Chain 1: 4200 -17085.547 0.022 0.021
Chain 1: 4300 -17224.007 0.021 0.021
Chain 1: 4400 -17180.747 0.019 0.011
Chain 1: 4500 -17083.265 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001298 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.98 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12436.296 1.000 1.000
Chain 1: 200 -9390.892 0.662 1.000
Chain 1: 300 -8090.005 0.495 0.324
Chain 1: 400 -8281.453 0.377 0.324
Chain 1: 500 -8145.209 0.305 0.161
Chain 1: 600 -8037.948 0.256 0.161
Chain 1: 700 -7950.035 0.221 0.023
Chain 1: 800 -7957.687 0.194 0.023
Chain 1: 900 -7864.037 0.174 0.017
Chain 1: 1000 -8053.956 0.159 0.023
Chain 1: 1100 -8089.894 0.059 0.017
Chain 1: 1200 -7984.478 0.028 0.013
Chain 1: 1300 -7921.346 0.013 0.013
Chain 1: 1400 -7943.576 0.011 0.012
Chain 1: 1500 -8030.638 0.010 0.011
Chain 1: 1600 -7993.012 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001387 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.87 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61802.741 1.000 1.000
Chain 1: 200 -17938.064 1.723 2.445
Chain 1: 300 -8949.887 1.483 1.004
Chain 1: 400 -8422.542 1.128 1.004
Chain 1: 500 -8279.192 0.906 1.000
Chain 1: 600 -8949.337 0.767 1.000
Chain 1: 700 -8388.452 0.667 0.075
Chain 1: 800 -7881.220 0.592 0.075
Chain 1: 900 -8055.646 0.529 0.067
Chain 1: 1000 -7961.210 0.477 0.067
Chain 1: 1100 -7882.697 0.378 0.064
Chain 1: 1200 -7737.978 0.135 0.063
Chain 1: 1300 -7730.402 0.035 0.022
Chain 1: 1400 -7884.952 0.031 0.020
Chain 1: 1500 -7710.716 0.031 0.022
Chain 1: 1600 -7751.822 0.024 0.020
Chain 1: 1700 -7618.511 0.019 0.019
Chain 1: 1800 -7636.827 0.013 0.017
Chain 1: 1900 -7672.147 0.011 0.012
Chain 1: 2000 -7742.515 0.011 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002677 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 26.77 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86094.596 1.000 1.000
Chain 1: 200 -13551.712 3.177 5.353
Chain 1: 300 -9919.684 2.240 1.000
Chain 1: 400 -10961.012 1.704 1.000
Chain 1: 500 -8886.622 1.410 0.366
Chain 1: 600 -8410.528 1.184 0.366
Chain 1: 700 -8334.525 1.016 0.233
Chain 1: 800 -8658.528 0.894 0.233
Chain 1: 900 -8741.248 0.796 0.095
Chain 1: 1000 -8468.654 0.719 0.095
Chain 1: 1100 -8742.683 0.622 0.057 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8555.505 0.089 0.037
Chain 1: 1300 -8566.181 0.053 0.032
Chain 1: 1400 -8622.813 0.044 0.031
Chain 1: 1500 -8490.609 0.022 0.022
Chain 1: 1600 -8603.408 0.018 0.016
Chain 1: 1700 -8684.379 0.018 0.016
Chain 1: 1800 -8271.279 0.019 0.016
Chain 1: 1900 -8367.565 0.019 0.016
Chain 1: 2000 -8340.809 0.016 0.013
Chain 1: 2100 -8463.368 0.015 0.013
Chain 1: 2200 -8283.560 0.015 0.013
Chain 1: 2300 -8362.435 0.015 0.013
Chain 1: 2400 -8432.132 0.016 0.013
Chain 1: 2500 -8377.552 0.015 0.012
Chain 1: 2600 -8376.986 0.013 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003268 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.68 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8413716.696 1.000 1.000
Chain 1: 200 -1586238.832 2.652 4.304
Chain 1: 300 -892019.007 2.027 1.000
Chain 1: 400 -458553.832 1.757 1.000
Chain 1: 500 -358486.683 1.461 0.945
Chain 1: 600 -233393.443 1.307 0.945
Chain 1: 700 -119412.917 1.257 0.945
Chain 1: 800 -86588.231 1.147 0.945
Chain 1: 900 -66899.492 1.052 0.778
Chain 1: 1000 -51678.732 0.977 0.778
Chain 1: 1100 -39141.968 0.909 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38315.359 0.480 0.379
Chain 1: 1300 -26258.456 0.448 0.379
Chain 1: 1400 -25976.514 0.355 0.320
Chain 1: 1500 -22560.439 0.342 0.320
Chain 1: 1600 -21775.989 0.292 0.295
Chain 1: 1700 -20648.060 0.202 0.294
Chain 1: 1800 -20591.805 0.165 0.151
Chain 1: 1900 -20917.852 0.137 0.055
Chain 1: 2000 -19428.314 0.115 0.055
Chain 1: 2100 -19666.716 0.084 0.036
Chain 1: 2200 -19893.294 0.083 0.036
Chain 1: 2300 -19510.405 0.039 0.020
Chain 1: 2400 -19282.489 0.039 0.020
Chain 1: 2500 -19084.607 0.025 0.016
Chain 1: 2600 -18714.786 0.023 0.016
Chain 1: 2700 -18671.718 0.018 0.012
Chain 1: 2800 -18388.651 0.020 0.015
Chain 1: 2900 -18669.875 0.019 0.015
Chain 1: 3000 -18656.091 0.012 0.012
Chain 1: 3100 -18741.078 0.011 0.012
Chain 1: 3200 -18431.760 0.012 0.015
Chain 1: 3300 -18636.464 0.011 0.012
Chain 1: 3400 -18111.469 0.012 0.015
Chain 1: 3500 -18723.252 0.015 0.015
Chain 1: 3600 -18030.030 0.017 0.015
Chain 1: 3700 -18416.771 0.018 0.017
Chain 1: 3800 -17376.680 0.023 0.021
Chain 1: 3900 -17372.826 0.021 0.021
Chain 1: 4000 -17490.127 0.022 0.021
Chain 1: 4100 -17403.913 0.022 0.021
Chain 1: 4200 -17220.181 0.021 0.021
Chain 1: 4300 -17358.551 0.021 0.021
Chain 1: 4400 -17315.401 0.018 0.011
Chain 1: 4500 -17217.948 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001789 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 17.89 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12834.925 1.000 1.000
Chain 1: 200 -9722.372 0.660 1.000
Chain 1: 300 -8170.614 0.503 0.320
Chain 1: 400 -8437.600 0.385 0.320
Chain 1: 500 -8314.877 0.311 0.190
Chain 1: 600 -8162.822 0.263 0.190
Chain 1: 700 -8279.017 0.227 0.032
Chain 1: 800 -8118.930 0.201 0.032
Chain 1: 900 -8029.722 0.180 0.020
Chain 1: 1000 -8043.786 0.162 0.020
Chain 1: 1100 -8234.269 0.064 0.020
Chain 1: 1200 -8093.914 0.034 0.019
Chain 1: 1300 -8020.538 0.016 0.017
Chain 1: 1400 -8049.105 0.013 0.015
Chain 1: 1500 -8149.601 0.013 0.014
Chain 1: 1600 -8061.740 0.012 0.012
Chain 1: 1700 -8027.515 0.011 0.011
Chain 1: 1800 -7998.559 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001379 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.79 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58345.799 1.000 1.000
Chain 1: 200 -18092.752 1.612 2.225
Chain 1: 300 -8919.107 1.418 1.029
Chain 1: 400 -8136.165 1.087 1.029
Chain 1: 500 -8495.208 0.878 1.000
Chain 1: 600 -8663.309 0.735 1.000
Chain 1: 700 -7888.208 0.644 0.098
Chain 1: 800 -7677.815 0.567 0.098
Chain 1: 900 -7996.663 0.509 0.096
Chain 1: 1000 -7836.221 0.460 0.096
Chain 1: 1100 -7914.404 0.361 0.042
Chain 1: 1200 -7783.360 0.140 0.040
Chain 1: 1300 -7765.335 0.037 0.027
Chain 1: 1400 -7992.561 0.031 0.027
Chain 1: 1500 -7670.232 0.030 0.027
Chain 1: 1600 -7855.795 0.031 0.027
Chain 1: 1700 -7647.125 0.024 0.027
Chain 1: 1800 -7754.652 0.022 0.024
Chain 1: 1900 -7806.474 0.019 0.020
Chain 1: 2000 -7695.911 0.019 0.017
Chain 1: 2100 -7568.476 0.019 0.017
Chain 1: 2200 -8002.511 0.023 0.024
Chain 1: 2300 -7659.184 0.027 0.027
Chain 1: 2400 -7712.180 0.025 0.024
Chain 1: 2500 -7713.790 0.021 0.017
Chain 1: 2600 -7605.918 0.020 0.014
Chain 1: 2700 -7615.726 0.017 0.014
Chain 1: 2800 -7713.976 0.017 0.014
Chain 1: 2900 -7464.919 0.020 0.014
Chain 1: 3000 -7614.673 0.020 0.017
Chain 1: 3100 -7605.507 0.019 0.014
Chain 1: 3200 -7815.188 0.016 0.014
Chain 1: 3300 -7528.221 0.015 0.014
Chain 1: 3400 -7763.537 0.018 0.020
Chain 1: 3500 -7515.326 0.021 0.027
Chain 1: 3600 -7581.981 0.021 0.027
Chain 1: 3700 -7531.568 0.021 0.027
Chain 1: 3800 -7528.994 0.020 0.027
Chain 1: 3900 -7494.134 0.017 0.020
Chain 1: 4000 -7485.839 0.015 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002599 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.99 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86954.134 1.000 1.000
Chain 1: 200 -13912.642 3.125 5.250
Chain 1: 300 -10165.915 2.206 1.000
Chain 1: 400 -11739.345 1.688 1.000
Chain 1: 500 -8886.400 1.415 0.369
Chain 1: 600 -9643.894 1.192 0.369
Chain 1: 700 -8592.569 1.039 0.321
Chain 1: 800 -8798.508 0.912 0.321
Chain 1: 900 -8975.376 0.813 0.134
Chain 1: 1000 -8993.503 0.732 0.134
Chain 1: 1100 -8865.750 0.633 0.122 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8511.486 0.113 0.079
Chain 1: 1300 -8815.553 0.079 0.042
Chain 1: 1400 -8749.095 0.067 0.034
Chain 1: 1500 -8664.436 0.035 0.023
Chain 1: 1600 -8776.890 0.029 0.020
Chain 1: 1700 -8834.335 0.017 0.014
Chain 1: 1800 -8391.986 0.020 0.014
Chain 1: 1900 -8497.088 0.019 0.013
Chain 1: 2000 -8482.769 0.019 0.013
Chain 1: 2100 -8599.510 0.019 0.013
Chain 1: 2200 -8393.855 0.018 0.013
Chain 1: 2300 -8489.594 0.015 0.012
Chain 1: 2400 -8556.283 0.015 0.012
Chain 1: 2500 -8504.841 0.015 0.012
Chain 1: 2600 -8519.052 0.014 0.011
Chain 1: 2700 -8426.360 0.014 0.011
Chain 1: 2800 -8373.601 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003514 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.14 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8413862.213 1.000 1.000
Chain 1: 200 -1586956.162 2.651 4.302
Chain 1: 300 -891563.446 2.027 1.000
Chain 1: 400 -457875.600 1.757 1.000
Chain 1: 500 -358007.217 1.462 0.947
Chain 1: 600 -233091.142 1.307 0.947
Chain 1: 700 -119498.625 1.256 0.947
Chain 1: 800 -86727.499 1.147 0.947
Chain 1: 900 -67115.468 1.052 0.780
Chain 1: 1000 -51950.322 0.976 0.780
Chain 1: 1100 -39452.847 0.907 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38639.824 0.479 0.378
Chain 1: 1300 -26618.538 0.446 0.378
Chain 1: 1400 -26342.450 0.353 0.317
Chain 1: 1500 -22933.983 0.340 0.317
Chain 1: 1600 -22152.528 0.290 0.292
Chain 1: 1700 -21028.662 0.200 0.292
Chain 1: 1800 -20973.703 0.162 0.149
Chain 1: 1900 -21300.361 0.135 0.053
Chain 1: 2000 -19811.529 0.113 0.053
Chain 1: 2100 -20050.202 0.083 0.035
Chain 1: 2200 -20276.602 0.082 0.035
Chain 1: 2300 -19893.675 0.038 0.019
Chain 1: 2400 -19665.598 0.038 0.019
Chain 1: 2500 -19467.318 0.025 0.015
Chain 1: 2600 -19097.296 0.023 0.015
Chain 1: 2700 -19054.186 0.018 0.012
Chain 1: 2800 -18770.647 0.019 0.015
Chain 1: 2900 -19052.143 0.019 0.015
Chain 1: 3000 -19038.393 0.012 0.012
Chain 1: 3100 -19123.423 0.011 0.012
Chain 1: 3200 -18813.802 0.011 0.015
Chain 1: 3300 -19018.768 0.011 0.012
Chain 1: 3400 -18493.052 0.012 0.015
Chain 1: 3500 -19105.776 0.014 0.015
Chain 1: 3600 -18411.354 0.016 0.015
Chain 1: 3700 -18798.941 0.018 0.016
Chain 1: 3800 -17756.824 0.022 0.021
Chain 1: 3900 -17752.857 0.021 0.021
Chain 1: 4000 -17870.235 0.022 0.021
Chain 1: 4100 -17783.854 0.022 0.021
Chain 1: 4200 -17599.705 0.021 0.021
Chain 1: 4300 -17738.428 0.021 0.021
Chain 1: 4400 -17694.956 0.018 0.010
Chain 1: 4500 -17597.361 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001343 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.43 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49573.852 1.000 1.000
Chain 1: 200 -22743.382 1.090 1.180
Chain 1: 300 -14980.279 0.899 1.000
Chain 1: 400 -21084.531 0.747 1.000
Chain 1: 500 -24747.566 0.627 0.518
Chain 1: 600 -19820.827 0.564 0.518
Chain 1: 700 -14735.897 0.533 0.345
Chain 1: 800 -17364.480 0.485 0.345
Chain 1: 900 -14206.548 0.456 0.290
Chain 1: 1000 -10930.099 0.440 0.300
Chain 1: 1100 -11445.087 0.345 0.290
Chain 1: 1200 -20098.824 0.270 0.290
Chain 1: 1300 -12602.368 0.277 0.290
Chain 1: 1400 -10838.889 0.265 0.249
Chain 1: 1500 -10526.193 0.253 0.249
Chain 1: 1600 -12302.260 0.243 0.222
Chain 1: 1700 -11336.730 0.217 0.163
Chain 1: 1800 -11959.686 0.207 0.163
Chain 1: 1900 -12646.648 0.190 0.144
Chain 1: 2000 -18273.449 0.191 0.144
Chain 1: 2100 -12033.667 0.238 0.163
Chain 1: 2200 -12537.395 0.199 0.144
Chain 1: 2300 -9745.646 0.168 0.144
Chain 1: 2400 -12021.442 0.171 0.144
Chain 1: 2500 -10115.139 0.187 0.188
Chain 1: 2600 -10390.703 0.175 0.188
Chain 1: 2700 -10185.339 0.168 0.188
Chain 1: 2800 -10332.269 0.165 0.188
Chain 1: 2900 -10098.231 0.161 0.188
Chain 1: 3000 -10789.636 0.137 0.064
Chain 1: 3100 -10407.396 0.089 0.040
Chain 1: 3200 -13672.154 0.109 0.064
Chain 1: 3300 -17208.953 0.101 0.064
Chain 1: 3400 -11257.960 0.135 0.064
Chain 1: 3500 -9724.568 0.132 0.064
Chain 1: 3600 -9697.770 0.129 0.064
Chain 1: 3700 -9414.132 0.130 0.064
Chain 1: 3800 -9848.078 0.133 0.064
Chain 1: 3900 -10192.509 0.134 0.064
Chain 1: 4000 -9253.792 0.138 0.101
Chain 1: 4100 -10789.328 0.149 0.142
Chain 1: 4200 -9927.493 0.133 0.101
Chain 1: 4300 -11418.625 0.126 0.101
Chain 1: 4400 -8975.695 0.100 0.101
Chain 1: 4500 -9232.115 0.087 0.087
Chain 1: 4600 -14214.208 0.122 0.101
Chain 1: 4700 -11623.237 0.141 0.131
Chain 1: 4800 -12831.273 0.146 0.131
Chain 1: 4900 -13653.900 0.149 0.131
Chain 1: 5000 -13108.679 0.143 0.131
Chain 1: 5100 -10541.719 0.153 0.131
Chain 1: 5200 -10885.390 0.148 0.131
Chain 1: 5300 -10940.429 0.135 0.094
Chain 1: 5400 -9317.849 0.125 0.094
Chain 1: 5500 -9833.107 0.128 0.094
Chain 1: 5600 -8991.721 0.102 0.094
Chain 1: 5700 -14171.725 0.116 0.094
Chain 1: 5800 -13613.309 0.111 0.060
Chain 1: 5900 -9868.436 0.143 0.094
Chain 1: 6000 -8988.689 0.148 0.098
Chain 1: 6100 -10905.920 0.142 0.098
Chain 1: 6200 -13023.510 0.155 0.163
Chain 1: 6300 -13544.823 0.158 0.163
Chain 1: 6400 -8964.367 0.192 0.163
Chain 1: 6500 -9195.695 0.189 0.163
Chain 1: 6600 -9136.282 0.180 0.163
Chain 1: 6700 -12829.313 0.173 0.163
Chain 1: 6800 -15212.353 0.184 0.163
Chain 1: 6900 -9219.669 0.211 0.163
Chain 1: 7000 -13049.669 0.231 0.176
Chain 1: 7100 -9035.873 0.258 0.288
Chain 1: 7200 -9678.559 0.248 0.288
Chain 1: 7300 -9176.936 0.250 0.288
Chain 1: 7400 -9398.174 0.201 0.157
Chain 1: 7500 -11279.819 0.215 0.167
Chain 1: 7600 -8858.469 0.242 0.273
Chain 1: 7700 -8896.186 0.213 0.167
Chain 1: 7800 -9658.220 0.206 0.167
Chain 1: 7900 -8781.984 0.151 0.100
Chain 1: 8000 -8682.971 0.122 0.079
Chain 1: 8100 -8748.676 0.079 0.066
Chain 1: 8200 -9817.277 0.083 0.079
Chain 1: 8300 -12204.365 0.097 0.100
Chain 1: 8400 -12396.728 0.096 0.100
Chain 1: 8500 -10295.244 0.100 0.100
Chain 1: 8600 -9348.188 0.083 0.100
Chain 1: 8700 -9509.585 0.084 0.100
Chain 1: 8800 -8732.130 0.085 0.100
Chain 1: 8900 -11321.467 0.098 0.101
Chain 1: 9000 -11655.295 0.100 0.101
Chain 1: 9100 -12754.747 0.107 0.101
Chain 1: 9200 -8744.985 0.142 0.101
Chain 1: 9300 -11345.988 0.146 0.101
Chain 1: 9400 -11536.235 0.146 0.101
Chain 1: 9500 -9866.338 0.142 0.101
Chain 1: 9600 -10245.218 0.136 0.089
Chain 1: 9700 -9725.932 0.140 0.089
Chain 1: 9800 -9312.179 0.135 0.086
Chain 1: 9900 -12362.896 0.137 0.086
Chain 1: 10000 -9522.288 0.164 0.169
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001386 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.86 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -47048.803 1.000 1.000
Chain 1: 200 -16210.471 1.451 1.902
Chain 1: 300 -9089.223 1.229 1.000
Chain 1: 400 -8335.846 0.944 1.000
Chain 1: 500 -8936.928 0.769 0.783
Chain 1: 600 -9689.139 0.654 0.783
Chain 1: 700 -7975.320 0.591 0.215
Chain 1: 800 -7815.282 0.520 0.215
Chain 1: 900 -8103.081 0.466 0.090
Chain 1: 1000 -8298.550 0.422 0.090
Chain 1: 1100 -7841.062 0.327 0.078
Chain 1: 1200 -8042.077 0.140 0.067
Chain 1: 1300 -7875.992 0.063 0.058
Chain 1: 1400 -8162.162 0.058 0.036
Chain 1: 1500 -7661.387 0.058 0.036
Chain 1: 1600 -7841.506 0.052 0.035
Chain 1: 1700 -7571.222 0.034 0.035
Chain 1: 1800 -7705.571 0.034 0.035
Chain 1: 1900 -7730.870 0.031 0.025
Chain 1: 2000 -7838.885 0.030 0.025
Chain 1: 2100 -7755.866 0.025 0.023
Chain 1: 2200 -8010.187 0.026 0.023
Chain 1: 2300 -7718.388 0.027 0.032
Chain 1: 2400 -7653.846 0.025 0.023
Chain 1: 2500 -7715.269 0.019 0.017
Chain 1: 2600 -7644.497 0.018 0.014
Chain 1: 2700 -7635.891 0.014 0.011
Chain 1: 2800 -7639.541 0.012 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003197 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 31.97 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -87665.988 1.000 1.000
Chain 1: 200 -14209.389 3.085 5.170
Chain 1: 300 -10467.937 2.176 1.000
Chain 1: 400 -11663.271 1.657 1.000
Chain 1: 500 -9429.978 1.373 0.357
Chain 1: 600 -9214.112 1.148 0.357
Chain 1: 700 -9161.202 0.985 0.237
Chain 1: 800 -8769.130 0.868 0.237
Chain 1: 900 -8780.445 0.771 0.102
Chain 1: 1000 -9090.368 0.698 0.102
Chain 1: 1100 -9232.458 0.599 0.045 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8810.207 0.087 0.045
Chain 1: 1300 -9138.502 0.055 0.036
Chain 1: 1400 -9083.927 0.045 0.034
Chain 1: 1500 -8959.321 0.023 0.023
Chain 1: 1600 -9071.522 0.022 0.015
Chain 1: 1700 -9133.392 0.022 0.015
Chain 1: 1800 -8690.889 0.022 0.015
Chain 1: 1900 -8797.153 0.024 0.015
Chain 1: 2000 -8781.633 0.020 0.014
Chain 1: 2100 -8908.597 0.020 0.014
Chain 1: 2200 -8696.399 0.018 0.014
Chain 1: 2300 -8791.156 0.015 0.012
Chain 1: 2400 -8858.312 0.015 0.012
Chain 1: 2500 -8806.202 0.015 0.012
Chain 1: 2600 -8818.378 0.014 0.011
Chain 1: 2700 -8726.824 0.014 0.011
Chain 1: 2800 -8675.210 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003545 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.45 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8405651.197 1.000 1.000
Chain 1: 200 -1586031.380 2.650 4.300
Chain 1: 300 -891494.689 2.026 1.000
Chain 1: 400 -458069.580 1.756 1.000
Chain 1: 500 -358406.102 1.461 0.946
Chain 1: 600 -233465.077 1.306 0.946
Chain 1: 700 -119811.914 1.255 0.946
Chain 1: 800 -87055.313 1.145 0.946
Chain 1: 900 -67431.724 1.050 0.779
Chain 1: 1000 -52259.508 0.974 0.779
Chain 1: 1100 -39757.525 0.906 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38942.152 0.478 0.376
Chain 1: 1300 -26911.315 0.445 0.376
Chain 1: 1400 -26634.323 0.351 0.314
Chain 1: 1500 -23223.673 0.338 0.314
Chain 1: 1600 -22441.483 0.288 0.291
Chain 1: 1700 -21316.256 0.198 0.290
Chain 1: 1800 -21261.002 0.161 0.147
Chain 1: 1900 -21587.647 0.134 0.053
Chain 1: 2000 -20098.239 0.112 0.053
Chain 1: 2100 -20336.883 0.082 0.035
Chain 1: 2200 -20563.440 0.081 0.035
Chain 1: 2300 -20180.374 0.038 0.019
Chain 1: 2400 -19952.274 0.038 0.019
Chain 1: 2500 -19754.109 0.024 0.015
Chain 1: 2600 -19383.988 0.023 0.015
Chain 1: 2700 -19340.877 0.018 0.012
Chain 1: 2800 -19057.421 0.019 0.015
Chain 1: 2900 -19338.879 0.019 0.015
Chain 1: 3000 -19325.110 0.011 0.012
Chain 1: 3100 -19410.144 0.011 0.011
Chain 1: 3200 -19100.537 0.011 0.015
Chain 1: 3300 -19305.488 0.010 0.011
Chain 1: 3400 -18779.820 0.012 0.015
Chain 1: 3500 -19392.539 0.014 0.015
Chain 1: 3600 -18698.102 0.016 0.015
Chain 1: 3700 -19085.703 0.018 0.016
Chain 1: 3800 -18043.641 0.022 0.020
Chain 1: 3900 -18039.689 0.021 0.020
Chain 1: 4000 -18157.037 0.021 0.020
Chain 1: 4100 -18070.682 0.021 0.020
Chain 1: 4200 -17886.542 0.021 0.020
Chain 1: 4300 -18025.245 0.020 0.020
Chain 1: 4400 -17981.753 0.018 0.010
Chain 1: 4500 -17884.180 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001338 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.38 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12268.421 1.000 1.000
Chain 1: 200 -9199.568 0.667 1.000
Chain 1: 300 -8162.790 0.487 0.334
Chain 1: 400 -8157.108 0.365 0.334
Chain 1: 500 -8038.547 0.295 0.127
Chain 1: 600 -7954.819 0.248 0.127
Chain 1: 700 -7861.727 0.214 0.015
Chain 1: 800 -7905.755 0.188 0.015
Chain 1: 900 -8028.689 0.169 0.015
Chain 1: 1000 -7937.206 0.153 0.015
Chain 1: 1100 -7905.527 0.053 0.012
Chain 1: 1200 -7867.701 0.021 0.012
Chain 1: 1300 -7833.773 0.008 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001542 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 15.42 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56825.841 1.000 1.000
Chain 1: 200 -17328.193 1.640 2.279
Chain 1: 300 -8680.491 1.425 1.000
Chain 1: 400 -8456.337 1.076 1.000
Chain 1: 500 -8128.229 0.868 0.996
Chain 1: 600 -8519.205 0.731 0.996
Chain 1: 700 -7830.521 0.639 0.088
Chain 1: 800 -8143.776 0.564 0.088
Chain 1: 900 -7995.440 0.504 0.046
Chain 1: 1000 -7751.910 0.456 0.046
Chain 1: 1100 -7948.790 0.359 0.040
Chain 1: 1200 -7634.277 0.135 0.040
Chain 1: 1300 -7699.379 0.036 0.038
Chain 1: 1400 -7915.964 0.036 0.038
Chain 1: 1500 -7640.311 0.036 0.036
Chain 1: 1600 -7572.080 0.032 0.031
Chain 1: 1700 -7535.815 0.024 0.027
Chain 1: 1800 -7618.318 0.021 0.025
Chain 1: 1900 -7580.343 0.020 0.025
Chain 1: 2000 -7634.727 0.017 0.011
Chain 1: 2100 -7547.811 0.016 0.011
Chain 1: 2200 -7694.060 0.014 0.011
Chain 1: 2300 -7600.402 0.014 0.012
Chain 1: 2400 -7652.916 0.012 0.011
Chain 1: 2500 -7744.564 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003309 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.09 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85848.971 1.000 1.000
Chain 1: 200 -13418.433 3.199 5.398
Chain 1: 300 -9820.681 2.255 1.000
Chain 1: 400 -10675.112 1.711 1.000
Chain 1: 500 -8654.220 1.416 0.366
Chain 1: 600 -8329.868 1.186 0.366
Chain 1: 700 -8373.190 1.017 0.234
Chain 1: 800 -8525.211 0.892 0.234
Chain 1: 900 -8656.012 0.795 0.080
Chain 1: 1000 -8378.753 0.719 0.080
Chain 1: 1100 -8627.850 0.622 0.039 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8284.776 0.086 0.039
Chain 1: 1300 -8505.535 0.052 0.033
Chain 1: 1400 -8507.859 0.044 0.029
Chain 1: 1500 -8399.592 0.022 0.026
Chain 1: 1600 -8503.611 0.019 0.018
Chain 1: 1700 -8591.963 0.020 0.018
Chain 1: 1800 -8184.509 0.023 0.026
Chain 1: 1900 -8281.446 0.023 0.026
Chain 1: 2000 -8253.510 0.020 0.013
Chain 1: 2100 -8374.034 0.018 0.013
Chain 1: 2200 -8184.784 0.016 0.013
Chain 1: 2300 -8321.170 0.015 0.013
Chain 1: 2400 -8328.373 0.016 0.013
Chain 1: 2500 -8294.674 0.015 0.012
Chain 1: 2600 -8292.649 0.013 0.012
Chain 1: 2700 -8206.726 0.013 0.012
Chain 1: 2800 -8171.913 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.004731 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 47.31 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8367944.128 1.000 1.000
Chain 1: 200 -1576627.791 2.654 4.307
Chain 1: 300 -888976.988 2.027 1.000
Chain 1: 400 -456743.348 1.757 1.000
Chain 1: 500 -357901.587 1.461 0.946
Chain 1: 600 -233009.949 1.307 0.946
Chain 1: 700 -119239.911 1.256 0.946
Chain 1: 800 -86458.245 1.147 0.946
Chain 1: 900 -66780.035 1.052 0.774
Chain 1: 1000 -51560.947 0.976 0.774
Chain 1: 1100 -39020.258 0.908 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38193.475 0.480 0.379
Chain 1: 1300 -26125.949 0.449 0.379
Chain 1: 1400 -25841.812 0.355 0.321
Chain 1: 1500 -22423.519 0.343 0.321
Chain 1: 1600 -21638.504 0.293 0.295
Chain 1: 1700 -20509.095 0.203 0.295
Chain 1: 1800 -20452.565 0.165 0.152
Chain 1: 1900 -20778.653 0.137 0.055
Chain 1: 2000 -19288.591 0.116 0.055
Chain 1: 2100 -19526.811 0.085 0.036
Chain 1: 2200 -19753.639 0.084 0.036
Chain 1: 2300 -19370.573 0.039 0.020
Chain 1: 2400 -19142.715 0.039 0.020
Chain 1: 2500 -18944.936 0.025 0.016
Chain 1: 2600 -18575.054 0.024 0.016
Chain 1: 2700 -18532.016 0.018 0.012
Chain 1: 2800 -18249.098 0.020 0.016
Chain 1: 2900 -18530.271 0.020 0.015
Chain 1: 3000 -18516.363 0.012 0.012
Chain 1: 3100 -18601.377 0.011 0.012
Chain 1: 3200 -18292.111 0.012 0.015
Chain 1: 3300 -18496.799 0.011 0.012
Chain 1: 3400 -17971.928 0.013 0.015
Chain 1: 3500 -18583.620 0.015 0.016
Chain 1: 3600 -17890.540 0.017 0.016
Chain 1: 3700 -18277.220 0.019 0.017
Chain 1: 3800 -17237.393 0.023 0.021
Chain 1: 3900 -17233.606 0.022 0.021
Chain 1: 4000 -17350.839 0.022 0.021
Chain 1: 4100 -17264.690 0.022 0.021
Chain 1: 4200 -17081.037 0.022 0.021
Chain 1: 4300 -17219.333 0.021 0.021
Chain 1: 4400 -17176.235 0.019 0.011
Chain 1: 4500 -17078.834 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001421 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.21 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12631.213 1.000 1.000
Chain 1: 200 -9502.040 0.665 1.000
Chain 1: 300 -8225.480 0.495 0.329
Chain 1: 400 -8384.796 0.376 0.329
Chain 1: 500 -8063.463 0.309 0.155
Chain 1: 600 -8159.606 0.259 0.155
Chain 1: 700 -8281.810 0.224 0.040
Chain 1: 800 -8123.844 0.199 0.040
Chain 1: 900 -8197.139 0.178 0.019
Chain 1: 1000 -8159.953 0.160 0.019
Chain 1: 1100 -8194.247 0.061 0.019
Chain 1: 1200 -8093.683 0.029 0.015
Chain 1: 1300 -8186.459 0.015 0.012
Chain 1: 1400 -8074.843 0.014 0.012
Chain 1: 1500 -8170.706 0.011 0.012
Chain 1: 1600 -8113.255 0.011 0.012
Chain 1: 1700 -8056.600 0.010 0.011
Chain 1: 1800 -8030.572 0.008 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001401 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.01 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58499.393 1.000 1.000
Chain 1: 200 -17950.110 1.629 2.259
Chain 1: 300 -9053.080 1.414 1.000
Chain 1: 400 -8479.558 1.077 1.000
Chain 1: 500 -8601.899 0.865 0.983
Chain 1: 600 -8591.642 0.721 0.983
Chain 1: 700 -7995.405 0.628 0.075
Chain 1: 800 -8584.700 0.559 0.075
Chain 1: 900 -8013.564 0.504 0.071
Chain 1: 1000 -8043.556 0.454 0.071
Chain 1: 1100 -7810.298 0.357 0.069
Chain 1: 1200 -7854.704 0.132 0.068
Chain 1: 1300 -7890.015 0.034 0.030
Chain 1: 1400 -7818.468 0.028 0.014
Chain 1: 1500 -7755.304 0.028 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003345 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.45 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86519.010 1.000 1.000
Chain 1: 200 -13793.809 3.136 5.272
Chain 1: 300 -10112.607 2.212 1.000
Chain 1: 400 -11175.397 1.683 1.000
Chain 1: 500 -9076.297 1.393 0.364
Chain 1: 600 -8516.409 1.171 0.364
Chain 1: 700 -8647.044 1.006 0.231
Chain 1: 800 -9044.230 0.886 0.231
Chain 1: 900 -8818.566 0.790 0.095
Chain 1: 1000 -8716.817 0.712 0.095
Chain 1: 1100 -8948.406 0.615 0.066 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8465.829 0.094 0.057
Chain 1: 1300 -8736.826 0.060 0.044
Chain 1: 1400 -8792.063 0.051 0.031
Chain 1: 1500 -8650.937 0.030 0.026
Chain 1: 1600 -8756.791 0.024 0.026
Chain 1: 1700 -8827.711 0.024 0.026
Chain 1: 1800 -8397.821 0.025 0.026
Chain 1: 1900 -8501.633 0.023 0.016
Chain 1: 2000 -8476.722 0.022 0.016
Chain 1: 2100 -8610.196 0.021 0.016
Chain 1: 2200 -8405.368 0.018 0.016
Chain 1: 2300 -8500.639 0.016 0.012
Chain 1: 2400 -8565.249 0.016 0.012
Chain 1: 2500 -8510.211 0.015 0.012
Chain 1: 2600 -8514.273 0.014 0.011
Chain 1: 2700 -8429.596 0.014 0.011
Chain 1: 2800 -8386.504 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003419 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.19 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8416338.256 1.000 1.000
Chain 1: 200 -1583251.342 2.658 4.316
Chain 1: 300 -890767.730 2.031 1.000
Chain 1: 400 -457817.698 1.760 1.000
Chain 1: 500 -357927.277 1.464 0.946
Chain 1: 600 -232983.128 1.309 0.946
Chain 1: 700 -119388.263 1.258 0.946
Chain 1: 800 -86643.116 1.148 0.946
Chain 1: 900 -67016.168 1.053 0.777
Chain 1: 1000 -51837.279 0.977 0.777
Chain 1: 1100 -39337.474 0.909 0.536 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38520.809 0.479 0.378
Chain 1: 1300 -26492.608 0.447 0.378
Chain 1: 1400 -26215.617 0.353 0.318
Chain 1: 1500 -22806.191 0.340 0.318
Chain 1: 1600 -22024.557 0.290 0.293
Chain 1: 1700 -20899.520 0.201 0.293
Chain 1: 1800 -20844.376 0.163 0.149
Chain 1: 1900 -21170.719 0.135 0.054
Chain 1: 2000 -19682.332 0.114 0.054
Chain 1: 2100 -19920.695 0.083 0.035
Chain 1: 2200 -20147.146 0.082 0.035
Chain 1: 2300 -19764.340 0.039 0.019
Chain 1: 2400 -19536.347 0.039 0.019
Chain 1: 2500 -19338.334 0.025 0.015
Chain 1: 2600 -18968.281 0.023 0.015
Chain 1: 2700 -18925.307 0.018 0.012
Chain 1: 2800 -18641.975 0.019 0.015
Chain 1: 2900 -18923.357 0.019 0.015
Chain 1: 3000 -18909.555 0.012 0.012
Chain 1: 3100 -18994.518 0.011 0.012
Chain 1: 3200 -18685.088 0.011 0.015
Chain 1: 3300 -18889.966 0.011 0.012
Chain 1: 3400 -18364.596 0.012 0.015
Chain 1: 3500 -18976.819 0.015 0.015
Chain 1: 3600 -18283.119 0.016 0.015
Chain 1: 3700 -18670.132 0.018 0.017
Chain 1: 3800 -17629.178 0.023 0.021
Chain 1: 3900 -17625.315 0.021 0.021
Chain 1: 4000 -17742.631 0.022 0.021
Chain 1: 4100 -17656.270 0.022 0.021
Chain 1: 4200 -17472.459 0.021 0.021
Chain 1: 4300 -17610.914 0.021 0.021
Chain 1: 4400 -17567.614 0.018 0.011
Chain 1: 4500 -17470.125 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001429 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.29 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49127.145 1.000 1.000
Chain 1: 200 -18111.224 1.356 1.713
Chain 1: 300 -16932.863 0.927 1.000
Chain 1: 400 -16788.789 0.698 1.000
Chain 1: 500 -13627.010 0.605 0.232
Chain 1: 600 -14765.275 0.517 0.232
Chain 1: 700 -11495.613 0.483 0.232
Chain 1: 800 -11822.231 0.426 0.232
Chain 1: 900 -11299.200 0.384 0.077
Chain 1: 1000 -27924.591 0.405 0.232
Chain 1: 1100 -16763.625 0.372 0.232
Chain 1: 1200 -12801.542 0.232 0.232
Chain 1: 1300 -12519.873 0.227 0.232
Chain 1: 1400 -10764.146 0.242 0.232
Chain 1: 1500 -10884.744 0.220 0.163
Chain 1: 1600 -10228.406 0.219 0.163
Chain 1: 1700 -10276.955 0.191 0.064
Chain 1: 1800 -15535.294 0.222 0.163
Chain 1: 1900 -11096.849 0.257 0.310
Chain 1: 2000 -10674.553 0.202 0.163
Chain 1: 2100 -9664.550 0.146 0.105
Chain 1: 2200 -9995.113 0.118 0.064
Chain 1: 2300 -10042.639 0.116 0.064
Chain 1: 2400 -9450.819 0.106 0.063
Chain 1: 2500 -11868.492 0.126 0.064
Chain 1: 2600 -9360.629 0.146 0.105
Chain 1: 2700 -10468.937 0.156 0.106
Chain 1: 2800 -12354.339 0.137 0.106
Chain 1: 2900 -12298.243 0.098 0.105
Chain 1: 3000 -10175.892 0.115 0.106
Chain 1: 3100 -9832.314 0.108 0.106
Chain 1: 3200 -16310.893 0.144 0.153
Chain 1: 3300 -9427.998 0.217 0.204
Chain 1: 3400 -9092.050 0.214 0.204
Chain 1: 3500 -9291.490 0.196 0.153
Chain 1: 3600 -10137.234 0.178 0.106
Chain 1: 3700 -16448.972 0.205 0.153
Chain 1: 3800 -11140.755 0.238 0.209
Chain 1: 3900 -11144.754 0.237 0.209
Chain 1: 4000 -13738.971 0.235 0.189
Chain 1: 4100 -8873.954 0.287 0.384
Chain 1: 4200 -13282.697 0.280 0.332
Chain 1: 4300 -14470.516 0.215 0.189
Chain 1: 4400 -10918.609 0.244 0.325
Chain 1: 4500 -8952.907 0.264 0.325
Chain 1: 4600 -8609.615 0.260 0.325
Chain 1: 4700 -13649.458 0.258 0.325
Chain 1: 4800 -8986.516 0.262 0.325
Chain 1: 4900 -8957.993 0.263 0.325
Chain 1: 5000 -19924.284 0.299 0.332
Chain 1: 5100 -8641.368 0.375 0.332
Chain 1: 5200 -8720.095 0.342 0.325
Chain 1: 5300 -9758.374 0.345 0.325
Chain 1: 5400 -11721.650 0.329 0.220
Chain 1: 5500 -8519.886 0.345 0.369
Chain 1: 5600 -8809.557 0.344 0.369
Chain 1: 5700 -8801.615 0.307 0.167
Chain 1: 5800 -11529.267 0.279 0.167
Chain 1: 5900 -12723.048 0.288 0.167
Chain 1: 6000 -9027.684 0.274 0.167
Chain 1: 6100 -13406.076 0.176 0.167
Chain 1: 6200 -8400.596 0.235 0.237
Chain 1: 6300 -9829.478 0.238 0.237
Chain 1: 6400 -11668.680 0.237 0.237
Chain 1: 6500 -10911.032 0.207 0.158
Chain 1: 6600 -8798.768 0.228 0.237
Chain 1: 6700 -14526.033 0.267 0.240
Chain 1: 6800 -8837.733 0.308 0.327
Chain 1: 6900 -12495.135 0.327 0.327
Chain 1: 7000 -9293.674 0.321 0.327
Chain 1: 7100 -8981.304 0.292 0.293
Chain 1: 7200 -8490.036 0.238 0.240
Chain 1: 7300 -11610.258 0.250 0.269
Chain 1: 7400 -10334.291 0.247 0.269
Chain 1: 7500 -11048.789 0.246 0.269
Chain 1: 7600 -8947.373 0.246 0.269
Chain 1: 7700 -8591.138 0.211 0.235
Chain 1: 7800 -9385.309 0.155 0.123
Chain 1: 7900 -8364.596 0.138 0.122
Chain 1: 8000 -8935.971 0.110 0.085
Chain 1: 8100 -11441.335 0.128 0.122
Chain 1: 8200 -8801.773 0.152 0.123
Chain 1: 8300 -11222.859 0.147 0.123
Chain 1: 8400 -10583.073 0.141 0.122
Chain 1: 8500 -8305.196 0.162 0.216
Chain 1: 8600 -9497.967 0.151 0.126
Chain 1: 8700 -8896.173 0.153 0.126
Chain 1: 8800 -8467.684 0.150 0.126
Chain 1: 8900 -9449.929 0.148 0.126
Chain 1: 9000 -11464.276 0.159 0.176
Chain 1: 9100 -8505.671 0.172 0.176
Chain 1: 9200 -12966.049 0.177 0.176
Chain 1: 9300 -12570.092 0.158 0.126
Chain 1: 9400 -11087.309 0.165 0.134
Chain 1: 9500 -8201.220 0.173 0.134
Chain 1: 9600 -8852.884 0.168 0.134
Chain 1: 9700 -8445.136 0.166 0.134
Chain 1: 9800 -10277.426 0.179 0.176
Chain 1: 9900 -9286.604 0.179 0.176
Chain 1: 10000 -9107.894 0.164 0.134
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001435 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.35 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -61975.494 1.000 1.000
Chain 1: 200 -17983.793 1.723 2.446
Chain 1: 300 -8960.628 1.484 1.007
Chain 1: 400 -9197.058 1.120 1.007
Chain 1: 500 -8437.491 0.914 1.000
Chain 1: 600 -8572.838 0.764 1.000
Chain 1: 700 -8071.522 0.664 0.090
Chain 1: 800 -8143.666 0.582 0.090
Chain 1: 900 -8182.674 0.518 0.062
Chain 1: 1000 -7900.941 0.470 0.062
Chain 1: 1100 -7731.589 0.372 0.036
Chain 1: 1200 -7838.495 0.129 0.026
Chain 1: 1300 -7897.524 0.029 0.022
Chain 1: 1400 -7659.998 0.029 0.022
Chain 1: 1500 -7625.890 0.021 0.016
Chain 1: 1600 -7780.024 0.021 0.020
Chain 1: 1700 -7553.959 0.018 0.020
Chain 1: 1800 -7630.571 0.018 0.020
Chain 1: 1900 -7620.044 0.018 0.020
Chain 1: 2000 -7675.443 0.015 0.014
Chain 1: 2100 -7622.364 0.013 0.010
Chain 1: 2200 -7727.553 0.013 0.010
Chain 1: 2300 -7577.008 0.014 0.014
Chain 1: 2400 -7652.777 0.012 0.010
Chain 1: 2500 -7467.998 0.014 0.014
Chain 1: 2600 -7552.786 0.013 0.011
Chain 1: 2700 -7538.933 0.011 0.010
Chain 1: 2800 -7619.473 0.011 0.011
Chain 1: 2900 -7420.862 0.013 0.011
Chain 1: 3000 -7551.556 0.014 0.014
Chain 1: 3100 -7550.756 0.014 0.014
Chain 1: 3200 -7747.870 0.015 0.017
Chain 1: 3300 -7480.000 0.016 0.017
Chain 1: 3400 -7690.554 0.018 0.025
Chain 1: 3500 -7462.162 0.019 0.025
Chain 1: 3600 -7526.779 0.018 0.025
Chain 1: 3700 -7476.092 0.019 0.025
Chain 1: 3800 -7479.293 0.018 0.025
Chain 1: 3900 -7445.024 0.016 0.017
Chain 1: 4000 -7439.903 0.014 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003619 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 36.19 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85644.961 1.000 1.000
Chain 1: 200 -13681.447 3.130 5.260
Chain 1: 300 -10024.297 2.208 1.000
Chain 1: 400 -10962.812 1.678 1.000
Chain 1: 500 -9000.438 1.386 0.365
Chain 1: 600 -8765.844 1.159 0.365
Chain 1: 700 -8421.775 0.999 0.218
Chain 1: 800 -8767.073 0.879 0.218
Chain 1: 900 -8800.193 0.782 0.086
Chain 1: 1000 -8547.586 0.707 0.086
Chain 1: 1100 -8830.088 0.610 0.041 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8426.192 0.089 0.041
Chain 1: 1300 -8657.497 0.055 0.039
Chain 1: 1400 -8706.209 0.047 0.032
Chain 1: 1500 -8552.398 0.027 0.030
Chain 1: 1600 -8662.825 0.026 0.030
Chain 1: 1700 -8741.925 0.022 0.027
Chain 1: 1800 -8314.874 0.024 0.027
Chain 1: 1900 -8417.854 0.025 0.027
Chain 1: 2000 -8392.635 0.022 0.018
Chain 1: 2100 -8519.957 0.020 0.015
Chain 1: 2200 -8318.608 0.018 0.015
Chain 1: 2300 -8413.200 0.016 0.013
Chain 1: 2400 -8480.937 0.016 0.013
Chain 1: 2500 -8427.109 0.015 0.012
Chain 1: 2600 -8429.671 0.014 0.011
Chain 1: 2700 -8345.817 0.014 0.011
Chain 1: 2800 -8304.231 0.010 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002588 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 25.88 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8384028.153 1.000 1.000
Chain 1: 200 -1578379.352 2.656 4.312
Chain 1: 300 -890397.729 2.028 1.000
Chain 1: 400 -458067.402 1.757 1.000
Chain 1: 500 -359022.617 1.461 0.944
Chain 1: 600 -234049.772 1.306 0.944
Chain 1: 700 -119877.474 1.256 0.944
Chain 1: 800 -87023.605 1.146 0.944
Chain 1: 900 -67267.490 1.051 0.773
Chain 1: 1000 -51988.394 0.976 0.773
Chain 1: 1100 -39398.430 0.908 0.534 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38566.695 0.478 0.378
Chain 1: 1300 -26442.628 0.447 0.378
Chain 1: 1400 -26155.395 0.354 0.320
Chain 1: 1500 -22722.839 0.341 0.320
Chain 1: 1600 -21934.541 0.292 0.294
Chain 1: 1700 -20797.907 0.202 0.294
Chain 1: 1800 -20739.976 0.164 0.151
Chain 1: 1900 -21066.326 0.136 0.055
Chain 1: 2000 -19572.187 0.115 0.055
Chain 1: 2100 -19810.564 0.084 0.036
Chain 1: 2200 -20038.230 0.083 0.036
Chain 1: 2300 -19654.364 0.039 0.020
Chain 1: 2400 -19426.304 0.039 0.020
Chain 1: 2500 -19228.908 0.025 0.015
Chain 1: 2600 -18858.289 0.023 0.015
Chain 1: 2700 -18815.060 0.018 0.012
Chain 1: 2800 -18532.092 0.019 0.015
Chain 1: 2900 -18813.553 0.019 0.015
Chain 1: 3000 -18799.512 0.012 0.012
Chain 1: 3100 -18884.595 0.011 0.012
Chain 1: 3200 -18575.015 0.012 0.015
Chain 1: 3300 -18779.980 0.011 0.012
Chain 1: 3400 -18254.661 0.012 0.015
Chain 1: 3500 -18867.054 0.015 0.015
Chain 1: 3600 -18173.084 0.016 0.015
Chain 1: 3700 -18560.468 0.018 0.017
Chain 1: 3800 -17519.293 0.023 0.021
Chain 1: 3900 -17515.519 0.021 0.021
Chain 1: 4000 -17632.718 0.022 0.021
Chain 1: 4100 -17546.480 0.022 0.021
Chain 1: 4200 -17362.547 0.021 0.021
Chain 1: 4300 -17500.990 0.021 0.021
Chain 1: 4400 -17457.649 0.018 0.011
Chain 1: 4500 -17360.236 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001311 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.11 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12239.831 1.000 1.000
Chain 1: 200 -8926.255 0.686 1.000
Chain 1: 300 -8155.592 0.489 0.371
Chain 1: 400 -8134.628 0.367 0.371
Chain 1: 500 -7929.471 0.299 0.094
Chain 1: 600 -7766.274 0.253 0.094
Chain 1: 700 -7731.355 0.217 0.026
Chain 1: 800 -7743.662 0.190 0.026
Chain 1: 900 -7757.353 0.169 0.021
Chain 1: 1000 -7784.597 0.153 0.021
Chain 1: 1100 -7836.699 0.053 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001729 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 17.29 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58053.078 1.000 1.000
Chain 1: 200 -17479.939 1.661 2.321
Chain 1: 300 -8564.233 1.454 1.041
Chain 1: 400 -8153.882 1.103 1.041
Chain 1: 500 -8274.111 0.885 1.000
Chain 1: 600 -8608.706 0.744 1.000
Chain 1: 700 -7771.807 0.653 0.108
Chain 1: 800 -7998.221 0.575 0.108
Chain 1: 900 -7618.212 0.517 0.050
Chain 1: 1000 -7832.588 0.468 0.050
Chain 1: 1100 -7761.454 0.369 0.050
Chain 1: 1200 -7675.279 0.138 0.039
Chain 1: 1300 -7665.034 0.034 0.028
Chain 1: 1400 -7793.229 0.030 0.027
Chain 1: 1500 -7554.493 0.032 0.028
Chain 1: 1600 -7569.732 0.029 0.027
Chain 1: 1700 -7472.546 0.019 0.016
Chain 1: 1800 -7513.876 0.017 0.013
Chain 1: 1900 -7510.343 0.012 0.011
Chain 1: 2000 -7532.611 0.009 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002633 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 26.33 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86235.879 1.000 1.000
Chain 1: 200 -13284.465 3.246 5.491
Chain 1: 300 -9683.918 2.288 1.000
Chain 1: 400 -10716.008 1.740 1.000
Chain 1: 500 -8607.083 1.441 0.372
Chain 1: 600 -8142.625 1.210 0.372
Chain 1: 700 -8443.781 1.042 0.245
Chain 1: 800 -9020.801 0.920 0.245
Chain 1: 900 -8535.275 0.824 0.096
Chain 1: 1000 -8265.666 0.745 0.096
Chain 1: 1100 -8511.673 0.648 0.064 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8147.001 0.103 0.057
Chain 1: 1300 -8381.110 0.069 0.057
Chain 1: 1400 -8368.193 0.059 0.045
Chain 1: 1500 -8261.950 0.036 0.036
Chain 1: 1600 -8367.764 0.032 0.033
Chain 1: 1700 -8455.076 0.029 0.029
Chain 1: 1800 -8046.728 0.028 0.029
Chain 1: 1900 -8143.125 0.023 0.028
Chain 1: 2000 -8115.435 0.020 0.013
Chain 1: 2100 -8236.437 0.019 0.013
Chain 1: 2200 -8086.062 0.016 0.013
Chain 1: 2300 -8183.504 0.015 0.013
Chain 1: 2400 -8194.151 0.015 0.013
Chain 1: 2500 -8153.133 0.014 0.012
Chain 1: 2600 -8153.303 0.013 0.012
Chain 1: 2700 -8068.749 0.013 0.012
Chain 1: 2800 -8033.300 0.008 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002602 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 26.02 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8440400.620 1.000 1.000
Chain 1: 200 -1588839.196 2.656 4.312
Chain 1: 300 -890162.307 2.032 1.000
Chain 1: 400 -457318.413 1.761 1.000
Chain 1: 500 -357114.701 1.465 0.946
Chain 1: 600 -232094.722 1.310 0.946
Chain 1: 700 -118605.876 1.260 0.946
Chain 1: 800 -85951.142 1.150 0.946
Chain 1: 900 -66356.190 1.055 0.785
Chain 1: 1000 -51211.402 0.979 0.785
Chain 1: 1100 -38749.590 0.911 0.539 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37930.698 0.482 0.380
Chain 1: 1300 -25945.234 0.450 0.380
Chain 1: 1400 -25669.827 0.356 0.322
Chain 1: 1500 -22273.565 0.343 0.322
Chain 1: 1600 -21495.322 0.293 0.296
Chain 1: 1700 -20375.693 0.203 0.295
Chain 1: 1800 -20321.569 0.165 0.152
Chain 1: 1900 -20647.590 0.137 0.055
Chain 1: 2000 -19162.824 0.116 0.055
Chain 1: 2100 -19400.784 0.085 0.036
Chain 1: 2200 -19626.769 0.084 0.036
Chain 1: 2300 -19244.455 0.039 0.020
Chain 1: 2400 -19016.655 0.040 0.020
Chain 1: 2500 -18818.631 0.025 0.016
Chain 1: 2600 -18448.980 0.024 0.016
Chain 1: 2700 -18406.011 0.018 0.012
Chain 1: 2800 -18122.935 0.020 0.016
Chain 1: 2900 -18404.043 0.020 0.015
Chain 1: 3000 -18390.196 0.012 0.012
Chain 1: 3100 -18475.203 0.011 0.012
Chain 1: 3200 -18165.967 0.012 0.015
Chain 1: 3300 -18370.640 0.011 0.012
Chain 1: 3400 -17845.731 0.013 0.015
Chain 1: 3500 -18457.282 0.015 0.016
Chain 1: 3600 -17764.318 0.017 0.016
Chain 1: 3700 -18150.834 0.019 0.017
Chain 1: 3800 -17111.110 0.023 0.021
Chain 1: 3900 -17107.257 0.022 0.021
Chain 1: 4000 -17224.575 0.022 0.021
Chain 1: 4100 -17138.372 0.022 0.021
Chain 1: 4200 -16954.741 0.022 0.021
Chain 1: 4300 -17093.048 0.021 0.021
Chain 1: 4400 -17049.944 0.019 0.011
Chain 1: 4500 -16952.501 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001294 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 12.94 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49294.244 1.000 1.000
Chain 1: 200 -22383.228 1.101 1.202
Chain 1: 300 -25754.050 0.778 1.000
Chain 1: 400 -18356.481 0.684 1.000
Chain 1: 500 -12237.379 0.647 0.500
Chain 1: 600 -12928.627 0.548 0.500
Chain 1: 700 -16989.516 0.504 0.403
Chain 1: 800 -13142.690 0.478 0.403
Chain 1: 900 -16986.904 0.450 0.293
Chain 1: 1000 -10962.569 0.460 0.403
Chain 1: 1100 -12237.319 0.370 0.293
Chain 1: 1200 -18332.799 0.283 0.293
Chain 1: 1300 -11198.931 0.334 0.332
Chain 1: 1400 -15053.699 0.319 0.293
Chain 1: 1500 -17978.533 0.285 0.256
Chain 1: 1600 -22263.495 0.299 0.256
Chain 1: 1700 -9774.083 0.403 0.293
Chain 1: 1800 -11035.769 0.385 0.256
Chain 1: 1900 -10107.079 0.372 0.256
Chain 1: 2000 -10105.741 0.317 0.192
Chain 1: 2100 -11103.524 0.315 0.192
Chain 1: 2200 -10634.908 0.287 0.163
Chain 1: 2300 -9523.634 0.235 0.117
Chain 1: 2400 -18555.915 0.258 0.117
Chain 1: 2500 -9438.775 0.338 0.117
Chain 1: 2600 -9426.105 0.319 0.114
Chain 1: 2700 -12484.090 0.216 0.114
Chain 1: 2800 -10150.525 0.227 0.117
Chain 1: 2900 -9135.091 0.229 0.117
Chain 1: 3000 -12383.301 0.255 0.230
Chain 1: 3100 -10142.463 0.268 0.230
Chain 1: 3200 -16323.905 0.302 0.245
Chain 1: 3300 -9265.245 0.366 0.262
Chain 1: 3400 -9789.809 0.323 0.245
Chain 1: 3500 -10783.483 0.236 0.230
Chain 1: 3600 -11041.666 0.238 0.230
Chain 1: 3700 -10295.062 0.221 0.221
Chain 1: 3800 -8755.242 0.215 0.176
Chain 1: 3900 -9345.157 0.210 0.176
Chain 1: 4000 -9887.763 0.190 0.092
Chain 1: 4100 -8746.282 0.181 0.092
Chain 1: 4200 -9929.883 0.155 0.092
Chain 1: 4300 -11030.024 0.088 0.092
Chain 1: 4400 -9555.437 0.099 0.100
Chain 1: 4500 -8664.963 0.100 0.103
Chain 1: 4600 -8592.672 0.098 0.103
Chain 1: 4700 -8765.782 0.093 0.103
Chain 1: 4800 -8337.834 0.080 0.100
Chain 1: 4900 -8769.459 0.079 0.100
Chain 1: 5000 -8937.152 0.075 0.100
Chain 1: 5100 -11706.425 0.086 0.100
Chain 1: 5200 -8944.762 0.105 0.100
Chain 1: 5300 -11259.311 0.116 0.103
Chain 1: 5400 -8591.477 0.131 0.103
Chain 1: 5500 -14421.115 0.161 0.206
Chain 1: 5600 -8857.144 0.223 0.237
Chain 1: 5700 -8803.665 0.222 0.237
Chain 1: 5800 -8540.596 0.220 0.237
Chain 1: 5900 -12387.502 0.246 0.309
Chain 1: 6000 -8478.689 0.290 0.311
Chain 1: 6100 -10526.112 0.286 0.311
Chain 1: 6200 -14151.761 0.281 0.311
Chain 1: 6300 -11896.004 0.279 0.311
Chain 1: 6400 -8590.574 0.287 0.311
Chain 1: 6500 -11308.259 0.270 0.256
Chain 1: 6600 -12164.503 0.214 0.240
Chain 1: 6700 -8933.644 0.250 0.256
Chain 1: 6800 -8222.359 0.256 0.256
Chain 1: 6900 -9538.152 0.238 0.240
Chain 1: 7000 -12021.807 0.213 0.207
Chain 1: 7100 -8534.953 0.234 0.240
Chain 1: 7200 -9736.367 0.221 0.207
Chain 1: 7300 -11345.082 0.216 0.207
Chain 1: 7400 -9023.703 0.203 0.207
Chain 1: 7500 -9902.214 0.188 0.142
Chain 1: 7600 -12028.252 0.199 0.177
Chain 1: 7700 -12129.014 0.164 0.142
Chain 1: 7800 -11105.822 0.164 0.142
Chain 1: 7900 -9218.881 0.171 0.177
Chain 1: 8000 -9850.875 0.157 0.142
Chain 1: 8100 -8454.930 0.132 0.142
Chain 1: 8200 -12399.777 0.152 0.165
Chain 1: 8300 -8611.610 0.182 0.177
Chain 1: 8400 -8315.409 0.159 0.165
Chain 1: 8500 -8291.946 0.151 0.165
Chain 1: 8600 -8923.162 0.140 0.092
Chain 1: 8700 -8281.067 0.147 0.092
Chain 1: 8800 -8437.587 0.140 0.078
Chain 1: 8900 -8459.767 0.120 0.071
Chain 1: 9000 -10619.465 0.133 0.078
Chain 1: 9100 -11680.540 0.126 0.078
Chain 1: 9200 -9149.802 0.122 0.078
Chain 1: 9300 -8118.559 0.091 0.078
Chain 1: 9400 -8554.420 0.092 0.078
Chain 1: 9500 -8413.583 0.093 0.078
Chain 1: 9600 -9874.848 0.101 0.091
Chain 1: 9700 -8586.507 0.108 0.127
Chain 1: 9800 -10419.973 0.124 0.148
Chain 1: 9900 -9056.199 0.139 0.150
Chain 1: 10000 -11425.369 0.139 0.150
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001919 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 19.19 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57063.373 1.000 1.000
Chain 1: 200 -17520.901 1.628 2.257
Chain 1: 300 -8789.365 1.417 1.000
Chain 1: 400 -8179.204 1.081 1.000
Chain 1: 500 -8926.839 0.882 0.993
Chain 1: 600 -8336.513 0.747 0.993
Chain 1: 700 -7990.368 0.646 0.084
Chain 1: 800 -8138.853 0.568 0.084
Chain 1: 900 -7817.420 0.509 0.075
Chain 1: 1000 -7716.496 0.460 0.075
Chain 1: 1100 -7682.814 0.360 0.071
Chain 1: 1200 -7770.096 0.135 0.043
Chain 1: 1300 -7531.327 0.039 0.041
Chain 1: 1400 -7868.836 0.036 0.041
Chain 1: 1500 -7573.385 0.032 0.039
Chain 1: 1600 -7588.250 0.025 0.032
Chain 1: 1700 -7556.497 0.021 0.018
Chain 1: 1800 -7594.001 0.019 0.013
Chain 1: 1900 -7596.730 0.015 0.011
Chain 1: 2000 -7652.320 0.015 0.007 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.006453 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 64.53 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86972.404 1.000 1.000
Chain 1: 200 -13645.477 3.187 5.374
Chain 1: 300 -9952.126 2.248 1.000
Chain 1: 400 -11117.795 1.712 1.000
Chain 1: 500 -8925.137 1.419 0.371
Chain 1: 600 -8388.284 1.193 0.371
Chain 1: 700 -8413.941 1.023 0.246
Chain 1: 800 -9126.151 0.905 0.246
Chain 1: 900 -8705.016 0.810 0.105
Chain 1: 1000 -8728.828 0.729 0.105
Chain 1: 1100 -8723.905 0.629 0.078 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8217.750 0.098 0.064
Chain 1: 1300 -8603.749 0.065 0.062
Chain 1: 1400 -8569.559 0.055 0.048
Chain 1: 1500 -8469.425 0.032 0.045
Chain 1: 1600 -8575.555 0.027 0.012
Chain 1: 1700 -8638.852 0.027 0.012
Chain 1: 1800 -8204.305 0.025 0.012
Chain 1: 1900 -8308.839 0.021 0.012
Chain 1: 2000 -8284.330 0.021 0.012
Chain 1: 2100 -8242.759 0.022 0.012
Chain 1: 2200 -8226.418 0.016 0.012
Chain 1: 2300 -8362.578 0.013 0.012
Chain 1: 2400 -8210.162 0.014 0.012
Chain 1: 2500 -8279.177 0.014 0.012
Chain 1: 2600 -8197.691 0.014 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003939 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 39.39 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8423292.900 1.000 1.000
Chain 1: 200 -1584259.289 2.658 4.317
Chain 1: 300 -889850.792 2.032 1.000
Chain 1: 400 -457649.755 1.760 1.000
Chain 1: 500 -357854.206 1.464 0.944
Chain 1: 600 -232777.647 1.310 0.944
Chain 1: 700 -119150.842 1.259 0.944
Chain 1: 800 -86455.309 1.149 0.944
Chain 1: 900 -66828.767 1.054 0.780
Chain 1: 1000 -51663.889 0.978 0.780
Chain 1: 1100 -39179.459 0.910 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38361.806 0.480 0.378
Chain 1: 1300 -26342.086 0.448 0.378
Chain 1: 1400 -26065.671 0.354 0.319
Chain 1: 1500 -22659.849 0.341 0.319
Chain 1: 1600 -21879.636 0.291 0.294
Chain 1: 1700 -20755.211 0.201 0.294
Chain 1: 1800 -20700.347 0.164 0.150
Chain 1: 1900 -21026.844 0.136 0.054
Chain 1: 2000 -19538.822 0.114 0.054
Chain 1: 2100 -19776.934 0.083 0.036
Chain 1: 2200 -20003.614 0.082 0.036
Chain 1: 2300 -19620.553 0.039 0.020
Chain 1: 2400 -19392.544 0.039 0.020
Chain 1: 2500 -19194.633 0.025 0.016
Chain 1: 2600 -18824.309 0.023 0.016
Chain 1: 2700 -18781.228 0.018 0.012
Chain 1: 2800 -18497.975 0.019 0.015
Chain 1: 2900 -18779.333 0.019 0.015
Chain 1: 3000 -18765.431 0.012 0.012
Chain 1: 3100 -18850.499 0.011 0.012
Chain 1: 3200 -18540.930 0.012 0.015
Chain 1: 3300 -18745.883 0.011 0.012
Chain 1: 3400 -18220.391 0.012 0.015
Chain 1: 3500 -18832.892 0.015 0.015
Chain 1: 3600 -18138.717 0.017 0.015
Chain 1: 3700 -18526.113 0.018 0.017
Chain 1: 3800 -17484.558 0.023 0.021
Chain 1: 3900 -17480.689 0.021 0.021
Chain 1: 4000 -17597.966 0.022 0.021
Chain 1: 4100 -17511.673 0.022 0.021
Chain 1: 4200 -17327.670 0.021 0.021
Chain 1: 4300 -17466.229 0.021 0.021
Chain 1: 4400 -17422.798 0.018 0.011
Chain 1: 4500 -17325.318 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001821 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 18.21 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -13547.548 1.000 1.000
Chain 1: 200 -10153.861 0.667 1.000
Chain 1: 300 -8905.107 0.491 0.334
Chain 1: 400 -8280.216 0.387 0.334
Chain 1: 500 -8569.257 0.317 0.140
Chain 1: 600 -8357.496 0.268 0.140
Chain 1: 700 -8268.119 0.231 0.075
Chain 1: 800 -8256.870 0.203 0.075
Chain 1: 900 -8402.105 0.182 0.034
Chain 1: 1000 -8296.820 0.165 0.034
Chain 1: 1100 -8382.439 0.066 0.025
Chain 1: 1200 -8310.338 0.034 0.017
Chain 1: 1300 -8233.777 0.020 0.013
Chain 1: 1400 -8242.028 0.013 0.011
Chain 1: 1500 -8348.987 0.011 0.011
Chain 1: 1600 -8285.551 0.009 0.010 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.0024 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 24 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -51555.567 1.000 1.000
Chain 1: 200 -17028.262 1.514 2.028
Chain 1: 300 -8925.550 1.312 1.000
Chain 1: 400 -8774.213 0.988 1.000
Chain 1: 500 -7953.604 0.811 0.908
Chain 1: 600 -9206.450 0.699 0.908
Chain 1: 700 -8640.040 0.608 0.136
Chain 1: 800 -8528.548 0.534 0.136
Chain 1: 900 -8123.135 0.480 0.103
Chain 1: 1000 -8073.043 0.433 0.103
Chain 1: 1100 -7795.009 0.336 0.066
Chain 1: 1200 -7851.563 0.134 0.050
Chain 1: 1300 -8022.657 0.046 0.036
Chain 1: 1400 -7761.671 0.047 0.036
Chain 1: 1500 -7659.029 0.038 0.034
Chain 1: 1600 -7875.084 0.027 0.027
Chain 1: 1700 -7731.148 0.023 0.021
Chain 1: 1800 -7807.115 0.022 0.021
Chain 1: 1900 -7910.201 0.019 0.019
Chain 1: 2000 -7719.469 0.020 0.021
Chain 1: 2100 -7694.645 0.017 0.019
Chain 1: 2200 -7857.548 0.019 0.021
Chain 1: 2300 -7659.726 0.019 0.021
Chain 1: 2400 -7745.454 0.017 0.019
Chain 1: 2500 -7713.142 0.016 0.019
Chain 1: 2600 -7659.743 0.014 0.013
Chain 1: 2700 -7572.498 0.013 0.012
Chain 1: 2800 -7755.917 0.014 0.013
Chain 1: 2900 -7511.273 0.016 0.021
Chain 1: 3000 -7669.503 0.016 0.021
Chain 1: 3100 -7653.715 0.016 0.021
Chain 1: 3200 -7870.151 0.017 0.021
Chain 1: 3300 -7578.908 0.018 0.021
Chain 1: 3400 -7824.943 0.020 0.024
Chain 1: 3500 -7568.324 0.023 0.028
Chain 1: 3600 -7633.510 0.023 0.028
Chain 1: 3700 -7584.869 0.023 0.028
Chain 1: 3800 -7584.756 0.020 0.028
Chain 1: 3900 -7543.643 0.017 0.021
Chain 1: 4000 -7535.674 0.015 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003776 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 37.76 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86300.163 1.000 1.000
Chain 1: 200 -14044.247 3.072 5.145
Chain 1: 300 -10342.356 2.168 1.000
Chain 1: 400 -11478.879 1.650 1.000
Chain 1: 500 -9325.762 1.367 0.358
Chain 1: 600 -8711.981 1.151 0.358
Chain 1: 700 -8832.335 0.988 0.231
Chain 1: 800 -9128.257 0.869 0.231
Chain 1: 900 -9160.950 0.773 0.099
Chain 1: 1000 -9133.949 0.696 0.099
Chain 1: 1100 -8946.868 0.598 0.070 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8737.394 0.086 0.032
Chain 1: 1300 -9011.281 0.053 0.030
Chain 1: 1400 -9004.125 0.043 0.024
Chain 1: 1500 -8853.820 0.022 0.021
Chain 1: 1600 -8969.782 0.016 0.017
Chain 1: 1700 -9035.824 0.015 0.017
Chain 1: 1800 -8601.208 0.017 0.017
Chain 1: 1900 -8705.339 0.018 0.017
Chain 1: 2000 -8681.066 0.018 0.017
Chain 1: 2100 -8626.975 0.016 0.013
Chain 1: 2200 -8623.830 0.014 0.012
Chain 1: 2300 -8761.223 0.013 0.012
Chain 1: 2400 -8605.808 0.014 0.013
Chain 1: 2500 -8675.338 0.013 0.012
Chain 1: 2600 -8593.210 0.013 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003366 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.66 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8401471.347 1.000 1.000
Chain 1: 200 -1586137.912 2.648 4.297
Chain 1: 300 -891750.540 2.025 1.000
Chain 1: 400 -458469.822 1.755 1.000
Chain 1: 500 -358726.659 1.460 0.945
Chain 1: 600 -233723.220 1.306 0.945
Chain 1: 700 -119902.522 1.255 0.945
Chain 1: 800 -87060.752 1.145 0.945
Chain 1: 900 -67393.878 1.050 0.779
Chain 1: 1000 -52184.451 0.974 0.779
Chain 1: 1100 -39648.068 0.906 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38828.845 0.478 0.377
Chain 1: 1300 -26771.471 0.446 0.377
Chain 1: 1400 -26491.325 0.352 0.316
Chain 1: 1500 -23073.954 0.339 0.316
Chain 1: 1600 -22289.343 0.289 0.292
Chain 1: 1700 -21161.381 0.200 0.291
Chain 1: 1800 -21105.421 0.162 0.148
Chain 1: 1900 -21431.898 0.134 0.053
Chain 1: 2000 -19941.346 0.113 0.053
Chain 1: 2100 -20179.968 0.082 0.035
Chain 1: 2200 -20406.661 0.081 0.035
Chain 1: 2300 -20023.575 0.038 0.019
Chain 1: 2400 -19795.510 0.038 0.019
Chain 1: 2500 -19597.445 0.024 0.015
Chain 1: 2600 -19227.355 0.023 0.015
Chain 1: 2700 -19184.256 0.018 0.012
Chain 1: 2800 -18900.834 0.019 0.015
Chain 1: 2900 -19182.356 0.019 0.015
Chain 1: 3000 -19168.571 0.012 0.012
Chain 1: 3100 -19253.554 0.011 0.012
Chain 1: 3200 -18944.024 0.011 0.015
Chain 1: 3300 -19148.936 0.010 0.012
Chain 1: 3400 -18623.382 0.012 0.015
Chain 1: 3500 -19235.890 0.014 0.015
Chain 1: 3600 -18541.834 0.016 0.015
Chain 1: 3700 -18929.168 0.018 0.016
Chain 1: 3800 -17887.599 0.022 0.020
Chain 1: 3900 -17883.698 0.021 0.020
Chain 1: 4000 -18001.050 0.021 0.020
Chain 1: 4100 -17914.667 0.021 0.020
Chain 1: 4200 -17730.675 0.021 0.020
Chain 1: 4300 -17869.261 0.021 0.020
Chain 1: 4400 -17825.876 0.018 0.010
Chain 1: 4500 -17728.360 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002364 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 23.64 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48855.982 1.000 1.000
Chain 1: 200 -17315.212 1.411 1.822
Chain 1: 300 -19252.992 0.974 1.000
Chain 1: 400 -14903.299 0.804 1.000
Chain 1: 500 -16967.728 0.667 0.292
Chain 1: 600 -11002.710 0.646 0.542
Chain 1: 700 -14598.562 0.589 0.292
Chain 1: 800 -14626.296 0.516 0.292
Chain 1: 900 -16818.582 0.473 0.246
Chain 1: 1000 -10811.453 0.481 0.292
Chain 1: 1100 -10751.570 0.382 0.246
Chain 1: 1200 -10399.146 0.203 0.130
Chain 1: 1300 -12043.467 0.207 0.137
Chain 1: 1400 -9672.311 0.202 0.137
Chain 1: 1500 -10998.341 0.202 0.137
Chain 1: 1600 -28717.963 0.209 0.137
Chain 1: 1700 -8916.011 0.407 0.137
Chain 1: 1800 -10410.823 0.421 0.144
Chain 1: 1900 -19199.650 0.454 0.245
Chain 1: 2000 -12327.361 0.454 0.245
Chain 1: 2100 -15658.934 0.475 0.245
Chain 1: 2200 -9596.991 0.534 0.458 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2300 -10752.902 0.531 0.458 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2400 -13177.668 0.525 0.458 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2500 -9913.294 0.546 0.458 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 2600 -9252.288 0.492 0.329
Chain 1: 2700 -9641.295 0.274 0.213
Chain 1: 2800 -9537.695 0.260 0.213
Chain 1: 2900 -8856.740 0.222 0.184
Chain 1: 3000 -8882.083 0.167 0.107
Chain 1: 3100 -8649.305 0.148 0.077
Chain 1: 3200 -11378.712 0.109 0.077
Chain 1: 3300 -9137.239 0.123 0.077
Chain 1: 3400 -11397.331 0.124 0.077
Chain 1: 3500 -8701.356 0.122 0.077
Chain 1: 3600 -11071.404 0.137 0.198
Chain 1: 3700 -8403.104 0.164 0.214
Chain 1: 3800 -9888.294 0.178 0.214
Chain 1: 3900 -8830.756 0.182 0.214
Chain 1: 4000 -10224.605 0.196 0.214
Chain 1: 4100 -9965.825 0.196 0.214
Chain 1: 4200 -9261.475 0.179 0.198
Chain 1: 4300 -9344.142 0.156 0.150
Chain 1: 4400 -10015.624 0.143 0.136
Chain 1: 4500 -8500.427 0.129 0.136
Chain 1: 4600 -10330.240 0.126 0.136
Chain 1: 4700 -9722.655 0.100 0.120
Chain 1: 4800 -8869.743 0.095 0.096
Chain 1: 4900 -10561.704 0.099 0.096
Chain 1: 5000 -9992.687 0.091 0.076
Chain 1: 5100 -8415.769 0.107 0.096
Chain 1: 5200 -8688.155 0.103 0.096
Chain 1: 5300 -11561.794 0.127 0.160
Chain 1: 5400 -9649.311 0.140 0.177
Chain 1: 5500 -14053.942 0.153 0.177
Chain 1: 5600 -8909.280 0.193 0.187
Chain 1: 5700 -8579.580 0.191 0.187
Chain 1: 5800 -9250.886 0.188 0.187
Chain 1: 5900 -11417.277 0.191 0.190
Chain 1: 6000 -8027.847 0.228 0.198
Chain 1: 6100 -8486.316 0.215 0.198
Chain 1: 6200 -8063.792 0.217 0.198
Chain 1: 6300 -9600.032 0.208 0.190
Chain 1: 6400 -10796.199 0.199 0.160
Chain 1: 6500 -8315.847 0.198 0.160
Chain 1: 6600 -8348.727 0.140 0.111
Chain 1: 6700 -12044.807 0.167 0.160
Chain 1: 6800 -9099.691 0.192 0.190
Chain 1: 6900 -11935.529 0.197 0.238
Chain 1: 7000 -9624.772 0.179 0.238
Chain 1: 7100 -8989.872 0.180 0.238
Chain 1: 7200 -8681.983 0.179 0.238
Chain 1: 7300 -8369.799 0.166 0.238
Chain 1: 7400 -10337.197 0.174 0.238
Chain 1: 7500 -8400.337 0.168 0.231
Chain 1: 7600 -8191.348 0.170 0.231
Chain 1: 7700 -9693.239 0.155 0.190
Chain 1: 7800 -9416.477 0.125 0.155
Chain 1: 7900 -7963.011 0.120 0.155
Chain 1: 8000 -8044.265 0.097 0.071
Chain 1: 8100 -10625.672 0.114 0.155
Chain 1: 8200 -10833.896 0.112 0.155
Chain 1: 8300 -7881.783 0.146 0.183
Chain 1: 8400 -9463.644 0.144 0.167
Chain 1: 8500 -11520.797 0.138 0.167
Chain 1: 8600 -8317.547 0.174 0.179
Chain 1: 8700 -8501.036 0.161 0.179
Chain 1: 8800 -8028.696 0.164 0.179
Chain 1: 8900 -8175.112 0.148 0.167
Chain 1: 9000 -10147.184 0.166 0.179
Chain 1: 9100 -8655.578 0.159 0.172
Chain 1: 9200 -7908.549 0.166 0.172
Chain 1: 9300 -8120.076 0.132 0.167
Chain 1: 9400 -8154.690 0.115 0.094
Chain 1: 9500 -10230.870 0.118 0.094
Chain 1: 9600 -8092.422 0.106 0.094
Chain 1: 9700 -8374.952 0.107 0.094
Chain 1: 9800 -8319.806 0.102 0.094
Chain 1: 9900 -9659.424 0.114 0.139
Chain 1: 10000 -8574.068 0.107 0.127
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001757 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 17.57 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57871.463 1.000 1.000
Chain 1: 200 -17267.829 1.676 2.351
Chain 1: 300 -8450.705 1.465 1.043
Chain 1: 400 -8022.045 1.112 1.043
Chain 1: 500 -8076.202 0.891 1.000
Chain 1: 600 -8219.064 0.745 1.000
Chain 1: 700 -7920.319 0.644 0.053
Chain 1: 800 -7897.344 0.564 0.053
Chain 1: 900 -7709.357 0.504 0.038
Chain 1: 1000 -7658.455 0.454 0.038
Chain 1: 1100 -7572.323 0.356 0.024
Chain 1: 1200 -7499.437 0.121 0.017
Chain 1: 1300 -7561.296 0.018 0.011
Chain 1: 1400 -7634.858 0.013 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003466 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.66 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85358.317 1.000 1.000
Chain 1: 200 -13103.704 3.257 5.514
Chain 1: 300 -9548.472 2.295 1.000
Chain 1: 400 -10475.139 1.744 1.000
Chain 1: 500 -8493.557 1.442 0.372
Chain 1: 600 -8034.250 1.211 0.372
Chain 1: 700 -8361.917 1.044 0.233
Chain 1: 800 -9033.938 0.922 0.233
Chain 1: 900 -8407.274 0.828 0.088
Chain 1: 1000 -8142.234 0.749 0.088
Chain 1: 1100 -8413.255 0.652 0.075 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -7977.708 0.106 0.074
Chain 1: 1300 -8449.736 0.074 0.057
Chain 1: 1400 -8270.080 0.068 0.056
Chain 1: 1500 -8139.684 0.046 0.055
Chain 1: 1600 -8251.979 0.041 0.039
Chain 1: 1700 -8335.710 0.039 0.033
Chain 1: 1800 -7940.669 0.036 0.033
Chain 1: 1900 -8042.315 0.030 0.032
Chain 1: 2000 -8012.840 0.027 0.022
Chain 1: 2100 -8135.044 0.025 0.016
Chain 1: 2200 -7915.915 0.023 0.016
Chain 1: 2300 -8071.004 0.019 0.016
Chain 1: 2400 -8084.669 0.017 0.015
Chain 1: 2500 -8054.333 0.016 0.014
Chain 1: 2600 -8057.043 0.014 0.013
Chain 1: 2700 -7963.241 0.015 0.013
Chain 1: 2800 -7934.212 0.010 0.012 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003964 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 39.64 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8398909.319 1.000 1.000
Chain 1: 200 -1583641.119 2.652 4.304
Chain 1: 300 -889530.043 2.028 1.000
Chain 1: 400 -457015.783 1.758 1.000
Chain 1: 500 -357404.041 1.462 0.946
Chain 1: 600 -232376.848 1.308 0.946
Chain 1: 700 -118693.698 1.258 0.946
Chain 1: 800 -85967.960 1.148 0.946
Chain 1: 900 -66322.790 1.054 0.780
Chain 1: 1000 -51134.047 0.978 0.780
Chain 1: 1100 -38630.635 0.910 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37804.673 0.482 0.381
Chain 1: 1300 -25779.781 0.451 0.381
Chain 1: 1400 -25499.401 0.357 0.324
Chain 1: 1500 -22092.483 0.345 0.324
Chain 1: 1600 -21310.743 0.295 0.297
Chain 1: 1700 -20186.639 0.204 0.296
Chain 1: 1800 -20131.288 0.167 0.154
Chain 1: 1900 -20457.127 0.139 0.056
Chain 1: 2000 -18970.302 0.117 0.056
Chain 1: 2100 -19208.379 0.086 0.037
Chain 1: 2200 -19434.565 0.085 0.037
Chain 1: 2300 -19052.108 0.040 0.020
Chain 1: 2400 -18824.346 0.040 0.020
Chain 1: 2500 -18626.470 0.026 0.016
Chain 1: 2600 -18256.916 0.024 0.016
Chain 1: 2700 -18213.999 0.019 0.012
Chain 1: 2800 -17931.065 0.020 0.016
Chain 1: 2900 -18212.094 0.020 0.015
Chain 1: 3000 -18198.280 0.012 0.012
Chain 1: 3100 -18283.245 0.011 0.012
Chain 1: 3200 -17974.163 0.012 0.015
Chain 1: 3300 -18178.706 0.011 0.012
Chain 1: 3400 -17654.085 0.013 0.015
Chain 1: 3500 -18265.325 0.015 0.016
Chain 1: 3600 -17572.796 0.017 0.016
Chain 1: 3700 -17958.992 0.019 0.017
Chain 1: 3800 -16920.021 0.023 0.022
Chain 1: 3900 -16916.221 0.022 0.022
Chain 1: 4000 -17033.489 0.023 0.022
Chain 1: 4100 -16947.352 0.023 0.022
Chain 1: 4200 -16763.890 0.022 0.022
Chain 1: 4300 -16902.058 0.022 0.022
Chain 1: 4400 -16859.091 0.019 0.011
Chain 1: 4500 -16761.707 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001192 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 11.92 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -12006.478 1.000 1.000
Chain 1: 200 -8933.627 0.672 1.000
Chain 1: 300 -7957.927 0.489 0.344
Chain 1: 400 -8053.876 0.370 0.344
Chain 1: 500 -7860.941 0.301 0.123
Chain 1: 600 -7778.450 0.252 0.123
Chain 1: 700 -7716.595 0.217 0.025
Chain 1: 800 -7733.117 0.190 0.025
Chain 1: 900 -7747.478 0.170 0.012
Chain 1: 1000 -7756.388 0.153 0.012
Chain 1: 1100 -7935.735 0.055 0.012
Chain 1: 1200 -7738.127 0.023 0.012
Chain 1: 1300 -7688.830 0.011 0.011
Chain 1: 1400 -7717.601 0.011 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001683 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 16.83 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -56661.801 1.000 1.000
Chain 1: 200 -17007.188 1.666 2.332
Chain 1: 300 -8543.562 1.441 1.000
Chain 1: 400 -8832.947 1.089 1.000
Chain 1: 500 -8455.736 0.880 0.991
Chain 1: 600 -9158.008 0.746 0.991
Chain 1: 700 -8222.620 0.656 0.114
Chain 1: 800 -8200.216 0.574 0.114
Chain 1: 900 -7932.327 0.514 0.077
Chain 1: 1000 -7759.560 0.465 0.077
Chain 1: 1100 -7736.835 0.365 0.045
Chain 1: 1200 -7559.494 0.134 0.034
Chain 1: 1300 -7720.836 0.037 0.033
Chain 1: 1400 -7814.389 0.035 0.023
Chain 1: 1500 -7617.454 0.033 0.023
Chain 1: 1600 -7513.408 0.027 0.022
Chain 1: 1700 -7486.386 0.016 0.021
Chain 1: 1800 -7525.326 0.016 0.021
Chain 1: 1900 -7578.161 0.014 0.014
Chain 1: 2000 -7569.049 0.012 0.012
Chain 1: 2100 -7581.870 0.011 0.012
Chain 1: 2200 -7652.989 0.010 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00384 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 38.4 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85719.413 1.000 1.000
Chain 1: 200 -13079.602 3.277 5.554
Chain 1: 300 -9567.241 2.307 1.000
Chain 1: 400 -10432.528 1.751 1.000
Chain 1: 500 -8442.766 1.448 0.367
Chain 1: 600 -8148.118 1.213 0.367
Chain 1: 700 -8415.533 1.044 0.236
Chain 1: 800 -8902.201 0.920 0.236
Chain 1: 900 -8425.984 0.824 0.083
Chain 1: 1000 -8183.767 0.745 0.083
Chain 1: 1100 -8404.421 0.647 0.057 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8168.306 0.095 0.055
Chain 1: 1300 -8330.318 0.060 0.036
Chain 1: 1400 -8263.838 0.053 0.032
Chain 1: 1500 -8208.007 0.030 0.030
Chain 1: 1600 -8204.051 0.026 0.029
Chain 1: 1700 -8142.548 0.024 0.026
Chain 1: 1800 -8023.005 0.020 0.019
Chain 1: 1900 -8136.685 0.016 0.015
Chain 1: 2000 -8098.181 0.013 0.014
Chain 1: 2100 -8238.548 0.012 0.014
Chain 1: 2200 -8023.533 0.012 0.014
Chain 1: 2300 -8165.320 0.012 0.014
Chain 1: 2400 -8170.253 0.011 0.014
Chain 1: 2500 -8139.627 0.011 0.014
Chain 1: 2600 -8133.772 0.011 0.014
Chain 1: 2700 -8045.283 0.011 0.014
Chain 1: 2800 -8027.729 0.010 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003215 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.15 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8405352.038 1.000 1.000
Chain 1: 200 -1586043.918 2.650 4.300
Chain 1: 300 -891267.439 2.026 1.000
Chain 1: 400 -458251.210 1.756 1.000
Chain 1: 500 -358417.658 1.461 0.945
Chain 1: 600 -233054.381 1.307 0.945
Chain 1: 700 -118954.757 1.257 0.945
Chain 1: 800 -86141.485 1.148 0.945
Chain 1: 900 -66423.126 1.053 0.780
Chain 1: 1000 -51181.588 0.978 0.780
Chain 1: 1100 -38635.911 0.910 0.538 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -37798.231 0.482 0.381
Chain 1: 1300 -25742.581 0.451 0.381
Chain 1: 1400 -25456.706 0.358 0.325
Chain 1: 1500 -22042.968 0.345 0.325
Chain 1: 1600 -21258.177 0.295 0.298
Chain 1: 1700 -20131.086 0.205 0.297
Chain 1: 1800 -20074.522 0.167 0.155
Chain 1: 1900 -20399.925 0.139 0.056
Chain 1: 2000 -18912.146 0.117 0.056
Chain 1: 2100 -19150.279 0.086 0.037
Chain 1: 2200 -19376.543 0.085 0.037
Chain 1: 2300 -18994.086 0.040 0.020
Chain 1: 2400 -18766.403 0.040 0.020
Chain 1: 2500 -18568.683 0.026 0.016
Chain 1: 2600 -18199.445 0.024 0.016
Chain 1: 2700 -18156.455 0.019 0.012
Chain 1: 2800 -17873.848 0.020 0.016
Chain 1: 2900 -18154.694 0.020 0.015
Chain 1: 3000 -18140.845 0.012 0.012
Chain 1: 3100 -18225.828 0.011 0.012
Chain 1: 3200 -17916.910 0.012 0.015
Chain 1: 3300 -18121.248 0.011 0.012
Chain 1: 3400 -17597.121 0.013 0.015
Chain 1: 3500 -18207.724 0.015 0.016
Chain 1: 3600 -17515.931 0.017 0.016
Chain 1: 3700 -17901.679 0.019 0.017
Chain 1: 3800 -16863.939 0.024 0.022
Chain 1: 3900 -16860.152 0.022 0.022
Chain 1: 4000 -16977.411 0.023 0.022
Chain 1: 4100 -16891.423 0.023 0.022
Chain 1: 4200 -16708.107 0.022 0.022
Chain 1: 4300 -16846.139 0.022 0.022
Chain 1: 4400 -16803.398 0.019 0.011
Chain 1: 4500 -16706.024 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001391 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 13.91 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49059.331 1.000 1.000
Chain 1: 200 -23412.592 1.048 1.095
Chain 1: 300 -18299.574 0.792 1.000
Chain 1: 400 -14098.977 0.668 1.000
Chain 1: 500 -14863.494 0.545 0.298
Chain 1: 600 -12280.166 0.489 0.298
Chain 1: 700 -11763.350 0.426 0.279
Chain 1: 800 -11663.001 0.373 0.279
Chain 1: 900 -11581.871 0.333 0.210
Chain 1: 1000 -13561.066 0.314 0.210
Chain 1: 1100 -14063.786 0.218 0.146
Chain 1: 1200 -14435.281 0.111 0.051
Chain 1: 1300 -11288.964 0.111 0.051
Chain 1: 1400 -10666.931 0.087 0.051
Chain 1: 1500 -10160.737 0.086 0.050
Chain 1: 1600 -10825.728 0.072 0.050
Chain 1: 1700 -12861.360 0.083 0.058
Chain 1: 1800 -12743.526 0.083 0.058
Chain 1: 1900 -10158.673 0.108 0.061
Chain 1: 2000 -10203.757 0.094 0.058
Chain 1: 2100 -12498.504 0.108 0.061
Chain 1: 2200 -10912.019 0.120 0.145
Chain 1: 2300 -10649.138 0.095 0.061
Chain 1: 2400 -9606.785 0.100 0.109
Chain 1: 2500 -10905.939 0.107 0.119
Chain 1: 2600 -9813.426 0.112 0.119
Chain 1: 2700 -17447.656 0.140 0.119
Chain 1: 2800 -9821.844 0.217 0.145
Chain 1: 2900 -9586.667 0.194 0.119
Chain 1: 3000 -9520.290 0.194 0.119
Chain 1: 3100 -10853.946 0.188 0.119
Chain 1: 3200 -9275.810 0.190 0.119
Chain 1: 3300 -14954.173 0.226 0.123
Chain 1: 3400 -12532.107 0.234 0.170
Chain 1: 3500 -9791.876 0.250 0.193
Chain 1: 3600 -10016.744 0.241 0.193
Chain 1: 3700 -20650.659 0.249 0.193
Chain 1: 3800 -11522.298 0.251 0.193
Chain 1: 3900 -9697.650 0.267 0.193
Chain 1: 4000 -8958.012 0.275 0.193
Chain 1: 4100 -9479.632 0.268 0.193
Chain 1: 4200 -10636.292 0.262 0.193
Chain 1: 4300 -13313.274 0.244 0.193
Chain 1: 4400 -9309.118 0.268 0.201
Chain 1: 4500 -16477.051 0.283 0.201
Chain 1: 4600 -9085.312 0.362 0.430
Chain 1: 4700 -9200.703 0.312 0.201
Chain 1: 4800 -9226.191 0.233 0.188
Chain 1: 4900 -9214.540 0.214 0.109
Chain 1: 5000 -9689.260 0.211 0.109
Chain 1: 5100 -8608.712 0.218 0.126
Chain 1: 5200 -14443.584 0.247 0.201
Chain 1: 5300 -10122.554 0.270 0.404
Chain 1: 5400 -14590.592 0.258 0.306
Chain 1: 5500 -10421.134 0.254 0.306
Chain 1: 5600 -14999.499 0.203 0.305
Chain 1: 5700 -12666.306 0.221 0.305
Chain 1: 5800 -8805.199 0.264 0.306
Chain 1: 5900 -9172.975 0.268 0.306
Chain 1: 6000 -11574.796 0.284 0.306
Chain 1: 6100 -8806.992 0.303 0.314
Chain 1: 6200 -9780.649 0.272 0.306
Chain 1: 6300 -8667.631 0.242 0.305
Chain 1: 6400 -10189.246 0.227 0.208
Chain 1: 6500 -12679.195 0.206 0.196
Chain 1: 6600 -9085.927 0.215 0.196
Chain 1: 6700 -9176.398 0.198 0.196
Chain 1: 6800 -9031.955 0.156 0.149
Chain 1: 6900 -12424.243 0.179 0.196
Chain 1: 7000 -8566.291 0.203 0.196
Chain 1: 7100 -10855.079 0.193 0.196
Chain 1: 7200 -9124.776 0.202 0.196
Chain 1: 7300 -11730.429 0.211 0.211
Chain 1: 7400 -9934.049 0.214 0.211
Chain 1: 7500 -8958.376 0.206 0.211
Chain 1: 7600 -8493.176 0.172 0.190
Chain 1: 7700 -8913.611 0.175 0.190
Chain 1: 7800 -9807.415 0.183 0.190
Chain 1: 7900 -11500.992 0.170 0.181
Chain 1: 8000 -8779.328 0.156 0.181
Chain 1: 8100 -9131.619 0.139 0.147
Chain 1: 8200 -8928.844 0.122 0.109
Chain 1: 8300 -9779.538 0.109 0.091
Chain 1: 8400 -14032.256 0.121 0.091
Chain 1: 8500 -9721.685 0.155 0.091
Chain 1: 8600 -8751.512 0.160 0.111
Chain 1: 8700 -8331.403 0.160 0.111
Chain 1: 8800 -8445.619 0.153 0.111
Chain 1: 8900 -10799.711 0.160 0.111
Chain 1: 9000 -9433.583 0.143 0.111
Chain 1: 9100 -8614.512 0.149 0.111
Chain 1: 9200 -8331.825 0.150 0.111
Chain 1: 9300 -9225.799 0.151 0.111
Chain 1: 9400 -8466.536 0.130 0.097
Chain 1: 9500 -12158.764 0.116 0.097
Chain 1: 9600 -9907.937 0.127 0.097
Chain 1: 9700 -8500.666 0.139 0.145
Chain 1: 9800 -8377.724 0.139 0.145
Chain 1: 9900 -10614.055 0.138 0.145
Chain 1: 10000 -9249.427 0.138 0.148
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001655 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 16.55 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -57522.037 1.000 1.000
Chain 1: 200 -17547.156 1.639 2.278
Chain 1: 300 -8846.776 1.421 1.000
Chain 1: 400 -8243.626 1.084 1.000
Chain 1: 500 -8581.053 0.875 0.983
Chain 1: 600 -8328.761 0.734 0.983
Chain 1: 700 -8467.219 0.632 0.073
Chain 1: 800 -8204.869 0.557 0.073
Chain 1: 900 -8071.224 0.497 0.039
Chain 1: 1000 -7885.897 0.449 0.039
Chain 1: 1100 -7878.359 0.349 0.032
Chain 1: 1200 -7651.350 0.125 0.030
Chain 1: 1300 -7864.330 0.029 0.030
Chain 1: 1400 -7924.166 0.022 0.027
Chain 1: 1500 -7667.563 0.022 0.027
Chain 1: 1600 -7848.669 0.021 0.024
Chain 1: 1700 -7592.711 0.023 0.027
Chain 1: 1800 -7642.838 0.020 0.024
Chain 1: 1900 -7679.348 0.019 0.024
Chain 1: 2000 -7692.593 0.017 0.023
Chain 1: 2100 -7672.009 0.017 0.023
Chain 1: 2200 -7790.950 0.016 0.015
Chain 1: 2300 -7651.557 0.015 0.015
Chain 1: 2400 -7708.721 0.015 0.015
Chain 1: 2500 -7649.705 0.012 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00346 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 34.6 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86243.961 1.000 1.000
Chain 1: 200 -13698.893 3.148 5.296
Chain 1: 300 -10045.296 2.220 1.000
Chain 1: 400 -10904.684 1.685 1.000
Chain 1: 500 -9030.815 1.389 0.364
Chain 1: 600 -8983.190 1.159 0.364
Chain 1: 700 -8573.774 1.000 0.207
Chain 1: 800 -8805.695 0.878 0.207
Chain 1: 900 -8819.369 0.781 0.079
Chain 1: 1000 -8526.977 0.706 0.079
Chain 1: 1100 -8866.009 0.610 0.048 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8452.735 0.085 0.048
Chain 1: 1300 -8722.487 0.052 0.038
Chain 1: 1400 -8690.268 0.044 0.034
Chain 1: 1500 -8583.810 0.025 0.031
Chain 1: 1600 -8692.983 0.026 0.031
Chain 1: 1700 -8773.173 0.022 0.026
Chain 1: 1800 -8349.405 0.024 0.031
Chain 1: 1900 -8450.510 0.025 0.031
Chain 1: 2000 -8424.948 0.022 0.013
Chain 1: 2100 -8550.560 0.020 0.013
Chain 1: 2200 -8353.411 0.017 0.013
Chain 1: 2300 -8445.314 0.015 0.012
Chain 1: 2400 -8514.070 0.016 0.012
Chain 1: 2500 -8460.325 0.015 0.012
Chain 1: 2600 -8461.714 0.014 0.011
Chain 1: 2700 -8378.422 0.014 0.011
Chain 1: 2800 -8338.261 0.009 0.010 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003772 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 37.72 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8380164.729 1.000 1.000
Chain 1: 200 -1584734.594 2.644 4.288
Chain 1: 300 -890593.117 2.022 1.000
Chain 1: 400 -457117.667 1.754 1.000
Chain 1: 500 -357414.093 1.459 0.948
Chain 1: 600 -232770.374 1.305 0.948
Chain 1: 700 -119239.959 1.255 0.948
Chain 1: 800 -86490.138 1.145 0.948
Chain 1: 900 -66892.673 1.050 0.779
Chain 1: 1000 -51723.732 0.975 0.779
Chain 1: 1100 -39223.769 0.907 0.535 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38409.111 0.480 0.379
Chain 1: 1300 -26384.657 0.448 0.379
Chain 1: 1400 -26107.587 0.354 0.319
Chain 1: 1500 -22698.246 0.341 0.319
Chain 1: 1600 -21916.088 0.291 0.293
Chain 1: 1700 -20791.839 0.201 0.293
Chain 1: 1800 -20736.671 0.164 0.150
Chain 1: 1900 -21062.948 0.136 0.054
Chain 1: 2000 -19574.674 0.114 0.054
Chain 1: 2100 -19813.298 0.083 0.036
Chain 1: 2200 -20039.499 0.082 0.036
Chain 1: 2300 -19656.836 0.039 0.019
Chain 1: 2400 -19428.871 0.039 0.019
Chain 1: 2500 -19230.739 0.025 0.015
Chain 1: 2600 -18861.051 0.023 0.015
Chain 1: 2700 -18818.055 0.018 0.012
Chain 1: 2800 -18534.741 0.019 0.015
Chain 1: 2900 -18816.038 0.019 0.015
Chain 1: 3000 -18802.366 0.012 0.012
Chain 1: 3100 -18887.302 0.011 0.012
Chain 1: 3200 -18577.971 0.012 0.015
Chain 1: 3300 -18782.725 0.011 0.012
Chain 1: 3400 -18257.514 0.012 0.015
Chain 1: 3500 -18869.514 0.015 0.015
Chain 1: 3600 -18176.065 0.016 0.015
Chain 1: 3700 -18562.910 0.018 0.017
Chain 1: 3800 -17522.358 0.023 0.021
Chain 1: 3900 -17518.460 0.021 0.021
Chain 1: 4000 -17635.815 0.022 0.021
Chain 1: 4100 -17549.492 0.022 0.021
Chain 1: 4200 -17365.725 0.021 0.021
Chain 1: 4300 -17504.175 0.021 0.021
Chain 1: 4400 -17460.958 0.018 0.011
Chain 1: 4500 -17363.446 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001468 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.68 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -48815.124 1.000 1.000
Chain 1: 200 -17246.689 1.415 1.830
Chain 1: 300 -18871.305 0.972 1.000
Chain 1: 400 -12251.442 0.864 1.000
Chain 1: 500 -14758.319 0.725 0.540
Chain 1: 600 -16334.075 0.621 0.540
Chain 1: 700 -14880.673 0.546 0.170
Chain 1: 800 -20090.411 0.510 0.259
Chain 1: 900 -11093.998 0.543 0.259
Chain 1: 1000 -14889.154 0.515 0.259
Chain 1: 1100 -21346.111 0.445 0.259
Chain 1: 1200 -10616.972 0.363 0.259
Chain 1: 1300 -12588.686 0.370 0.259
Chain 1: 1400 -21971.753 0.359 0.259
Chain 1: 1500 -10242.618 0.456 0.302
Chain 1: 1600 -9802.026 0.451 0.302
Chain 1: 1700 -9997.846 0.443 0.302
Chain 1: 1800 -10041.502 0.418 0.302
Chain 1: 1900 -10170.998 0.338 0.255
Chain 1: 2000 -19126.513 0.359 0.302
Chain 1: 2100 -19781.079 0.332 0.157
Chain 1: 2200 -10620.976 0.317 0.157
Chain 1: 2300 -11606.466 0.310 0.085
Chain 1: 2400 -9449.421 0.290 0.085
Chain 1: 2500 -9562.893 0.177 0.045
Chain 1: 2600 -9262.030 0.176 0.033
Chain 1: 2700 -9256.440 0.174 0.033
Chain 1: 2800 -13516.944 0.205 0.085
Chain 1: 2900 -8972.927 0.254 0.228
Chain 1: 3000 -19939.348 0.263 0.228
Chain 1: 3100 -8952.893 0.382 0.315
Chain 1: 3200 -9616.477 0.303 0.228
Chain 1: 3300 -17445.929 0.339 0.315
Chain 1: 3400 -9429.292 0.401 0.449
Chain 1: 3500 -9682.144 0.403 0.449
Chain 1: 3600 -8899.713 0.408 0.449
Chain 1: 3700 -9555.716 0.415 0.449
Chain 1: 3800 -10199.800 0.390 0.449
Chain 1: 3900 -9587.644 0.345 0.088
Chain 1: 4000 -13490.732 0.319 0.088
Chain 1: 4100 -9909.972 0.233 0.088
Chain 1: 4200 -10828.731 0.234 0.088
Chain 1: 4300 -14263.607 0.214 0.088
Chain 1: 4400 -10544.380 0.164 0.088
Chain 1: 4500 -9562.724 0.172 0.103
Chain 1: 4600 -13725.081 0.193 0.241
Chain 1: 4700 -16509.437 0.203 0.241
Chain 1: 4800 -10833.215 0.249 0.289
Chain 1: 4900 -12550.890 0.256 0.289
Chain 1: 5000 -14294.494 0.240 0.241
Chain 1: 5100 -9785.302 0.250 0.241
Chain 1: 5200 -11795.781 0.258 0.241
Chain 1: 5300 -9441.894 0.259 0.249
Chain 1: 5400 -10252.461 0.232 0.170
Chain 1: 5500 -10170.995 0.222 0.170
Chain 1: 5600 -11492.570 0.203 0.169
Chain 1: 5700 -8807.929 0.217 0.170
Chain 1: 5800 -9465.909 0.172 0.137
Chain 1: 5900 -9027.556 0.163 0.122
Chain 1: 6000 -9378.603 0.154 0.115
Chain 1: 6100 -8612.447 0.117 0.089
Chain 1: 6200 -8472.120 0.102 0.079
Chain 1: 6300 -8487.634 0.077 0.070
Chain 1: 6400 -13026.569 0.104 0.070
Chain 1: 6500 -9567.567 0.139 0.089
Chain 1: 6600 -8720.966 0.137 0.089
Chain 1: 6700 -11926.699 0.134 0.089
Chain 1: 6800 -9130.453 0.158 0.097
Chain 1: 6900 -10584.936 0.166 0.137
Chain 1: 7000 -8542.144 0.187 0.239
Chain 1: 7100 -9022.246 0.183 0.239
Chain 1: 7200 -8264.196 0.191 0.239
Chain 1: 7300 -8693.137 0.195 0.239
Chain 1: 7400 -11895.184 0.187 0.239
Chain 1: 7500 -9564.112 0.176 0.239
Chain 1: 7600 -8644.335 0.177 0.239
Chain 1: 7700 -8602.379 0.150 0.137
Chain 1: 7800 -10444.448 0.137 0.137
Chain 1: 7900 -8310.690 0.149 0.176
Chain 1: 8000 -10346.014 0.145 0.176
Chain 1: 8100 -8464.460 0.162 0.197
Chain 1: 8200 -12580.587 0.185 0.222
Chain 1: 8300 -8752.093 0.224 0.244
Chain 1: 8400 -8907.747 0.199 0.222
Chain 1: 8500 -9389.810 0.180 0.197
Chain 1: 8600 -11701.345 0.189 0.198
Chain 1: 8700 -10579.322 0.199 0.198
Chain 1: 8800 -8717.170 0.203 0.214
Chain 1: 8900 -10042.289 0.190 0.198
Chain 1: 9000 -10847.495 0.178 0.198
Chain 1: 9100 -11406.436 0.161 0.132
Chain 1: 9200 -8648.867 0.160 0.132
Chain 1: 9300 -8074.750 0.123 0.106
Chain 1: 9400 -8350.772 0.125 0.106
Chain 1: 9500 -10410.938 0.139 0.132
Chain 1: 9600 -10255.761 0.121 0.106
Chain 1: 9700 -10113.548 0.112 0.074
Chain 1: 9800 -8324.500 0.112 0.074
Chain 1: 9900 -8365.547 0.099 0.071
Chain 1: 10000 -8216.140 0.094 0.049
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003212 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 32.12 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -50960.257 1.000 1.000
Chain 1: 200 -16239.245 1.569 2.138
Chain 1: 300 -8714.239 1.334 1.000
Chain 1: 400 -8540.136 1.006 1.000
Chain 1: 500 -8299.465 0.810 0.864
Chain 1: 600 -8088.963 0.680 0.864
Chain 1: 700 -7745.738 0.589 0.044
Chain 1: 800 -8250.908 0.523 0.061
Chain 1: 900 -7558.240 0.475 0.061
Chain 1: 1000 -7924.716 0.432 0.061
Chain 1: 1100 -7602.673 0.336 0.046
Chain 1: 1200 -7527.162 0.123 0.044
Chain 1: 1300 -7740.698 0.040 0.042
Chain 1: 1400 -7662.748 0.039 0.042
Chain 1: 1500 -7574.747 0.037 0.042
Chain 1: 1600 -7787.560 0.037 0.042
Chain 1: 1700 -7524.848 0.036 0.035
Chain 1: 1800 -7600.227 0.031 0.028
Chain 1: 1900 -7586.352 0.022 0.027
Chain 1: 2000 -7548.791 0.018 0.012
Chain 1: 2100 -7575.979 0.014 0.010
Chain 1: 2200 -7680.328 0.015 0.012
Chain 1: 2300 -7579.444 0.013 0.012
Chain 1: 2400 -7606.509 0.012 0.012
Chain 1: 2500 -7558.837 0.012 0.010 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003585 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 35.85 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -86076.365 1.000 1.000
Chain 1: 200 -13472.440 3.195 5.389
Chain 1: 300 -9900.103 2.250 1.000
Chain 1: 400 -10638.218 1.705 1.000
Chain 1: 500 -8833.924 1.405 0.361
Chain 1: 600 -8717.928 1.173 0.361
Chain 1: 700 -8560.981 1.008 0.204
Chain 1: 800 -8652.955 0.883 0.204
Chain 1: 900 -8765.272 0.787 0.069
Chain 1: 1000 -8506.071 0.711 0.069
Chain 1: 1100 -8706.922 0.613 0.030 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -8422.179 0.078 0.030
Chain 1: 1300 -8631.333 0.044 0.024
Chain 1: 1400 -8612.364 0.037 0.023
Chain 1: 1500 -8511.861 0.018 0.018
Chain 1: 1600 -8613.397 0.018 0.018
Chain 1: 1700 -8700.651 0.017 0.013
Chain 1: 1800 -8307.209 0.021 0.023
Chain 1: 1900 -8408.825 0.021 0.023
Chain 1: 2000 -8379.351 0.018 0.012
Chain 1: 2100 -8503.629 0.017 0.012
Chain 1: 2200 -8287.545 0.016 0.012
Chain 1: 2300 -8437.604 0.016 0.012
Chain 1: 2400 -8452.435 0.016 0.012
Chain 1: 2500 -8420.371 0.015 0.012
Chain 1: 2600 -8422.605 0.014 0.012
Chain 1: 2700 -8329.132 0.014 0.012
Chain 1: 2800 -8301.250 0.009 0.011 MEAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003305 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 33.05 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8405352.307 1.000 1.000
Chain 1: 200 -1586659.477 2.649 4.298
Chain 1: 300 -892523.311 2.025 1.000
Chain 1: 400 -458306.263 1.756 1.000
Chain 1: 500 -358293.945 1.460 0.947
Chain 1: 600 -233131.874 1.306 0.947
Chain 1: 700 -119267.920 1.256 0.947
Chain 1: 800 -86418.785 1.147 0.947
Chain 1: 900 -66755.751 1.052 0.778
Chain 1: 1000 -51540.291 0.976 0.778
Chain 1: 1100 -39008.386 0.908 0.537 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38180.097 0.481 0.380
Chain 1: 1300 -26142.194 0.449 0.380
Chain 1: 1400 -25858.764 0.355 0.321
Chain 1: 1500 -22447.353 0.343 0.321
Chain 1: 1600 -21663.096 0.293 0.295
Chain 1: 1700 -20538.632 0.203 0.295
Chain 1: 1800 -20482.717 0.165 0.152
Chain 1: 1900 -20808.405 0.137 0.055
Chain 1: 2000 -19321.043 0.115 0.055
Chain 1: 2100 -19559.341 0.084 0.036
Chain 1: 2200 -19785.320 0.083 0.036
Chain 1: 2300 -19403.086 0.039 0.020
Chain 1: 2400 -19175.382 0.039 0.020
Chain 1: 2500 -18977.198 0.025 0.016
Chain 1: 2600 -18607.987 0.024 0.016
Chain 1: 2700 -18565.128 0.018 0.012
Chain 1: 2800 -18282.104 0.020 0.015
Chain 1: 2900 -18563.142 0.020 0.015
Chain 1: 3000 -18549.445 0.012 0.012
Chain 1: 3100 -18634.326 0.011 0.012
Chain 1: 3200 -18325.322 0.012 0.015
Chain 1: 3300 -18529.791 0.011 0.012
Chain 1: 3400 -18005.186 0.013 0.015
Chain 1: 3500 -18616.274 0.015 0.015
Chain 1: 3600 -17924.066 0.017 0.015
Chain 1: 3700 -18310.017 0.019 0.017
Chain 1: 3800 -17271.333 0.023 0.021
Chain 1: 3900 -17267.511 0.022 0.021
Chain 1: 4000 -17384.842 0.022 0.021
Chain 1: 4100 -17298.648 0.022 0.021
Chain 1: 4200 -17115.270 0.022 0.021
Chain 1: 4300 -17253.424 0.021 0.021
Chain 1: 4400 -17210.554 0.019 0.011
Chain 1: 4500 -17113.138 0.016 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.00147 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 14.7 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Success! Found best value [eta = 10] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -49379.661 1.000 1.000
Chain 1: 200 -18896.868 1.307 1.613
Chain 1: 300 -16303.767 0.924 1.000
Chain 1: 400 -14367.805 0.727 1.000
Chain 1: 500 -15241.111 0.593 0.159
Chain 1: 600 -16908.975 0.510 0.159
Chain 1: 700 -15864.597 0.447 0.135
Chain 1: 800 -15039.146 0.398 0.135
Chain 1: 900 -11337.170 0.390 0.135
Chain 1: 1000 -13927.884 0.370 0.159
Chain 1: 1100 -12432.211 0.282 0.135
Chain 1: 1200 -12466.110 0.121 0.120
Chain 1: 1300 -12310.132 0.106 0.099
Chain 1: 1400 -17946.076 0.124 0.099
Chain 1: 1500 -12789.458 0.158 0.120
Chain 1: 1600 -10458.158 0.171 0.186
Chain 1: 1700 -10161.175 0.167 0.186
Chain 1: 1800 -14328.807 0.191 0.223
Chain 1: 1900 -10084.517 0.200 0.223
Chain 1: 2000 -10709.781 0.188 0.223
Chain 1: 2100 -11173.714 0.180 0.223
Chain 1: 2200 -11054.354 0.180 0.223
Chain 1: 2300 -9795.397 0.192 0.223
Chain 1: 2400 -9957.789 0.162 0.129
Chain 1: 2500 -17233.371 0.164 0.129
Chain 1: 2600 -9637.838 0.221 0.129
Chain 1: 2700 -11066.207 0.231 0.129
Chain 1: 2800 -12080.980 0.210 0.129
Chain 1: 2900 -9984.758 0.189 0.129
Chain 1: 3000 -9720.816 0.186 0.129
Chain 1: 3100 -10801.880 0.192 0.129
Chain 1: 3200 -9278.013 0.207 0.129
Chain 1: 3300 -14753.109 0.231 0.164
Chain 1: 3400 -19065.587 0.252 0.210
Chain 1: 3500 -9265.432 0.316 0.210
Chain 1: 3600 -18935.737 0.288 0.210
Chain 1: 3700 -10149.384 0.362 0.226
Chain 1: 3800 -9074.816 0.365 0.226
Chain 1: 3900 -10469.585 0.357 0.226
Chain 1: 4000 -10297.134 0.356 0.226
Chain 1: 4100 -9305.987 0.357 0.226
Chain 1: 4200 -12847.417 0.368 0.276
Chain 1: 4300 -12170.113 0.337 0.226
Chain 1: 4400 -9528.071 0.342 0.276
Chain 1: 4500 -10978.052 0.249 0.133
Chain 1: 4600 -14601.847 0.223 0.133
Chain 1: 4700 -9296.301 0.193 0.133
Chain 1: 4800 -9116.368 0.184 0.133
Chain 1: 4900 -8955.596 0.172 0.132
Chain 1: 5000 -10431.496 0.185 0.141
Chain 1: 5100 -9763.933 0.181 0.141
Chain 1: 5200 -13011.584 0.178 0.141
Chain 1: 5300 -9544.834 0.209 0.248
Chain 1: 5400 -8740.622 0.190 0.141
Chain 1: 5500 -8841.042 0.178 0.141
Chain 1: 5600 -9793.009 0.163 0.097
Chain 1: 5700 -10742.877 0.115 0.092
Chain 1: 5800 -10908.877 0.114 0.092
Chain 1: 5900 -16404.658 0.146 0.097
Chain 1: 6000 -9216.843 0.210 0.097
Chain 1: 6100 -11451.119 0.223 0.195
Chain 1: 6200 -9553.605 0.218 0.195
Chain 1: 6300 -13316.111 0.210 0.195
Chain 1: 6400 -11162.787 0.220 0.195
Chain 1: 6500 -9611.566 0.235 0.195
Chain 1: 6600 -9423.745 0.227 0.195
Chain 1: 6700 -8737.979 0.226 0.195
Chain 1: 6800 -9271.512 0.230 0.195
Chain 1: 6900 -12053.157 0.220 0.195
Chain 1: 7000 -9056.151 0.175 0.195
Chain 1: 7100 -8756.386 0.159 0.193
Chain 1: 7200 -11748.882 0.164 0.193
Chain 1: 7300 -8636.937 0.172 0.193
Chain 1: 7400 -8535.564 0.154 0.161
Chain 1: 7500 -8572.894 0.138 0.078
Chain 1: 7600 -8751.236 0.138 0.078
Chain 1: 7700 -12101.093 0.158 0.231
Chain 1: 7800 -11031.340 0.162 0.231
Chain 1: 7900 -8745.660 0.165 0.255
Chain 1: 8000 -9108.509 0.136 0.097
Chain 1: 8100 -8714.034 0.137 0.097
Chain 1: 8200 -8508.578 0.114 0.045
Chain 1: 8300 -11410.821 0.104 0.045
Chain 1: 8400 -8793.690 0.132 0.097
Chain 1: 8500 -9411.270 0.138 0.097
Chain 1: 8600 -12829.230 0.163 0.254
Chain 1: 8700 -9840.535 0.166 0.254
Chain 1: 8800 -9087.733 0.164 0.254
Chain 1: 8900 -9165.649 0.139 0.083
Chain 1: 9000 -11964.431 0.158 0.234
Chain 1: 9100 -8649.179 0.192 0.254
Chain 1: 9200 -11925.584 0.217 0.266
Chain 1: 9300 -8729.299 0.228 0.275
Chain 1: 9400 -9198.859 0.204 0.266
Chain 1: 9500 -8875.349 0.201 0.266
Chain 1: 9600 -8668.548 0.176 0.234
Chain 1: 9700 -8410.101 0.149 0.083
Chain 1: 9800 -12471.160 0.173 0.234
Chain 1: 9900 -11054.299 0.185 0.234
Chain 1: 10000 -8384.370 0.194 0.275
Chain 1: Informational Message: The maximum number of iterations is reached! The algorithm may not have converged.
Chain 1: This variational approximation is not guaranteed to be meaningful.
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.001689 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 16.89 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -58710.319 1.000 1.000
Chain 1: 200 -18247.952 1.609 2.217
Chain 1: 300 -8918.545 1.421 1.046
Chain 1: 400 -8095.467 1.091 1.046
Chain 1: 500 -8345.525 0.879 1.000
Chain 1: 600 -8478.934 0.735 1.000
Chain 1: 700 -8117.253 0.636 0.102
Chain 1: 800 -8440.489 0.562 0.102
Chain 1: 900 -8041.977 0.505 0.050
Chain 1: 1000 -7821.795 0.457 0.050
Chain 1: 1100 -7787.819 0.358 0.045
Chain 1: 1200 -7645.242 0.138 0.038
Chain 1: 1300 -7810.197 0.035 0.030
Chain 1: 1400 -7835.490 0.025 0.028
Chain 1: 1500 -7559.588 0.026 0.028
Chain 1: 1600 -7774.976 0.027 0.028
Chain 1: 1700 -7684.125 0.024 0.028
Chain 1: 1800 -7597.954 0.021 0.021
Chain 1: 1900 -7616.253 0.017 0.019
Chain 1: 2000 -7643.450 0.014 0.012
Chain 1: 2100 -7566.947 0.015 0.012
Chain 1: 2200 -7838.244 0.016 0.012
Chain 1: 2300 -7588.932 0.017 0.012
Chain 1: 2400 -7588.677 0.017 0.012
Chain 1: 2500 -7623.363 0.014 0.011
Chain 1: 2600 -7536.573 0.012 0.011
Chain 1: 2700 -7450.376 0.012 0.011
Chain 1: 2800 -7637.254 0.014 0.012
Chain 1: 2900 -7384.159 0.017 0.012
Chain 1: 3000 -7548.043 0.019 0.022
Chain 1: 3100 -7527.664 0.018 0.022
Chain 1: 3200 -7750.472 0.017 0.022
Chain 1: 3300 -7453.428 0.018 0.022
Chain 1: 3400 -7705.392 0.021 0.024
Chain 1: 3500 -7447.359 0.024 0.029
Chain 1: 3600 -7507.425 0.024 0.029
Chain 1: 3700 -7462.838 0.023 0.029
Chain 1: 3800 -7531.414 0.022 0.029
Chain 1: 3900 -7425.353 0.020 0.022
Chain 1: 4000 -7409.373 0.018 0.014
Chain 1: 4100 -7425.150 0.018 0.014
Chain 1: 4200 -7464.676 0.015 0.009 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.003785 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 37.85 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Success! Found best value [eta = 1] earlier than expected.
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -85928.778 1.000 1.000
Chain 1: 200 -13991.578 3.071 5.141
Chain 1: 300 -10279.598 2.168 1.000
Chain 1: 400 -11614.235 1.654 1.000
Chain 1: 500 -9277.782 1.374 0.361
Chain 1: 600 -9317.406 1.146 0.361
Chain 1: 700 -9582.024 0.986 0.252
Chain 1: 800 -8801.061 0.874 0.252
Chain 1: 900 -8628.239 0.779 0.115
Chain 1: 1000 -9313.358 0.708 0.115
Chain 1: 1100 -8839.205 0.614 0.089 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -9177.911 0.103 0.074
Chain 1: 1300 -8710.506 0.073 0.054
Chain 1: 1400 -8798.073 0.062 0.054
Chain 1: 1500 -8721.299 0.038 0.037
Chain 1: 1600 -8729.201 0.037 0.037
Chain 1: 1700 -8617.514 0.036 0.037
Chain 1: 1800 -8675.096 0.028 0.020
Chain 1: 1900 -8551.363 0.027 0.014
Chain 1: 2000 -8613.579 0.021 0.013
Chain 1: 2100 -8756.949 0.017 0.013
Chain 1: 2200 -8552.581 0.015 0.013
Chain 1: 2300 -8705.882 0.012 0.013
Chain 1: 2400 -8544.767 0.013 0.014
Chain 1: 2500 -8615.353 0.013 0.014
Chain 1: 2600 -8527.153 0.014 0.014
Chain 1: 2700 -8561.345 0.013 0.014
Chain 1: 2800 -8521.209 0.013 0.014
Chain 1: 2900 -8614.617 0.012 0.011
Chain 1: 3000 -8447.807 0.013 0.016
Chain 1: 3100 -8603.782 0.014 0.018
Chain 1: 3200 -8475.757 0.013 0.015
Chain 1: 3300 -8483.456 0.011 0.011
Chain 1: 3400 -8643.882 0.011 0.011
Chain 1: 3500 -8652.640 0.010 0.011
Chain 1: 3600 -8432.256 0.012 0.015
Chain 1: 3700 -8578.356 0.013 0.017
Chain 1: 3800 -8438.748 0.014 0.017
Chain 1: 3900 -8373.249 0.014 0.017
Chain 1: 4000 -8449.143 0.013 0.017
Chain 1: 4100 -8445.574 0.011 0.015
Chain 1: 4200 -8428.535 0.010 0.009 MEAN ELBO CONVERGED MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
Chain 1: ------------------------------------------------------------
Chain 1: EXPERIMENTAL ALGORITHM:
Chain 1: This procedure has not been thoroughly tested and may be unstable
Chain 1: or buggy. The interface is subject to change.
Chain 1: ------------------------------------------------------------
Chain 1:
Chain 1:
Chain 1:
Chain 1: Gradient evaluation took 0.002497 seconds
Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 24.97 seconds.
Chain 1: Adjust your expectations accordingly!
Chain 1:
Chain 1:
Chain 1: Begin eta adaptation.
Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
Chain 1: Iteration: 250 / 250 [100%] (Adaptation)
Chain 1: Success! Found best value [eta = 0.1].
Chain 1:
Chain 1: Begin stochastic gradient ascent.
Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
Chain 1: 100 -8372609.774 1.000 1.000
Chain 1: 200 -1581496.207 2.647 4.294
Chain 1: 300 -892254.435 2.022 1.000
Chain 1: 400 -458569.949 1.753 1.000
Chain 1: 500 -359346.813 1.458 0.946
Chain 1: 600 -234331.944 1.304 0.946
Chain 1: 700 -120217.111 1.253 0.946
Chain 1: 800 -87284.548 1.144 0.946
Chain 1: 900 -67560.000 1.049 0.772
Chain 1: 1000 -52295.291 0.973 0.772
Chain 1: 1100 -39702.730 0.905 0.533 MAY BE DIVERGING... INSPECT ELBO
Chain 1: 1200 -38877.060 0.478 0.377
Chain 1: 1300 -26759.904 0.446 0.377
Chain 1: 1400 -26474.001 0.352 0.317
Chain 1: 1500 -23040.576 0.339 0.317
Chain 1: 1600 -22251.306 0.290 0.292
Chain 1: 1700 -21116.144 0.200 0.292
Chain 1: 1800 -21058.471 0.163 0.149
Chain 1: 1900 -21385.010 0.135 0.054
Chain 1: 2000 -19890.460 0.113 0.054
Chain 1: 2100 -20129.294 0.083 0.035
Chain 1: 2200 -20356.655 0.082 0.035
Chain 1: 2300 -19972.969 0.038 0.019
Chain 1: 2400 -19744.835 0.039 0.019
Chain 1: 2500 -19546.990 0.025 0.015
Chain 1: 2600 -19176.645 0.023 0.015
Chain 1: 2700 -19133.447 0.018 0.012
Chain 1: 2800 -18850.115 0.019 0.015
Chain 1: 2900 -19131.717 0.019 0.015
Chain 1: 3000 -19117.882 0.012 0.012
Chain 1: 3100 -19202.896 0.011 0.012
Chain 1: 3200 -18893.289 0.011 0.015
Chain 1: 3300 -19098.251 0.011 0.012
Chain 1: 3400 -18572.636 0.012 0.015
Chain 1: 3500 -19185.385 0.014 0.015
Chain 1: 3600 -18491.058 0.016 0.015
Chain 1: 3700 -18878.632 0.018 0.016
Chain 1: 3800 -17836.743 0.022 0.021
Chain 1: 3900 -17832.885 0.021 0.021
Chain 1: 4000 -17950.168 0.022 0.021
Chain 1: 4100 -17863.822 0.022 0.021
Chain 1: 4200 -17679.751 0.021 0.021
Chain 1: 4300 -17818.368 0.021 0.021
Chain 1: 4400 -17774.931 0.018 0.010
Chain 1: 4500 -17677.439 0.015 0.008 MEDIAN ELBO CONVERGED
Chain 1:
Chain 1: Drawing a sample of size 1000 from the approximate posterior...
Chain 1: COMPLETED.
Setting 'QR' to TRUE can often be helpful when using one of the variational inference algorithms. See the documentation for the 'QR' argument.
# average over all simulations to output final table
final_table <- as.data.frame(apply(summary_stats, c(2,3), FUN=mean))
row.names(final_table) <-c('Oracle', 'Unadjusted', 'PMF', 'DEF')
names(final_table) <- c('RMSE', 'All', 'Causal', 'Non-causal')
final_table <-round(final_table, 2)
final_table
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